Title: "Physics-aware Machine learning for Earth observation"

Abstract: Most problems in Earth sciences aim to do inferences about the system, where accurate predictions are just a tiny part of the whole problem. Inferences mean understanding variables relations, deriving models that are physically plausible, that are simple parsimonious, and mathematically tractable. Machine learning models alone are excellent approximators, but very often do not respect the most elementary laws of physics, like mass or energy conservation, so consistency and confidence are compromised. I will review the main challenges ahead in the field, and introduce several ways to live in the Physics and machine learning interplay. Physics-aware machine learning models are just a step towards understanding the data-generating process, for which learning causal representations promises great advances. I'll review some recent methodologies to cope with it too. This is a collective long-term AI agenda towards developing and applying algorithms capable of discovering knowledge in the Earth system.

Bio: Gustau Camps-Valls (born 1972 in València) is a Physicist and Full Professor in Electrical Engineering in the Universitat de València, Spain, where lectures on machine learning, remote sensing and signal processing. He is the Head of the Image and Signal Processing (ISP) group, an interdisciplinary group of 40 researchers working at the intersection of AI for Earth and Climate sciences.

Prof. Camps-Valls published over 250+ peer-reviewed international journal papers, 350+ international conference papers, 25 book chapters, and 5 international books on remote sensing, image processing and machine learning. He has an h-index of 78 with 29000+ citations in Google Scholar. He was listed as a Highly Cited Researcher in 2011, 2020 and 2021; currently has 13 «Highly Cited Papers» and 1 «Hot Paper», Thomson Reuters ScienceWatch identified his activities as a Fast Moving Front research (2011) and the most-cited paper in the area of Engineering in 2011, received the Google Classic paper award (2019), and Stanford Metrics includes him in the top 2% most cited researchers of 2017-2020. He publishes in both technical and scientific journals, from IEEE and PLOS One to Nature, Nature Communications, Science Advances, and PNAS.

He has been Program Committee member of international conferences (IEEE, SPIE, EGU, AGU), and Technical Program Chair at IEEE IGARSS 2018 (2400+ attendees) and general at AISTATS 2022. He served in technical committees of the IEEE GRSS & IEEE SPS, as Associate Editor of 5 top IEEE journals, and in the prestigious IEEE Distinguished Lecturer program of the GRSS (2017-2019) to promote «AI in Earth sciences» globally. He has given 100+ talks, keynote speaker in 10+ conferences, and (co)advised 10+ PhD theses.

He coordinated/participated in 60+ research projects, involving industry and academia at national and European levels. He assisted the aerospace industry in Advisory Boards; Fellow Consultant of the ESA PhiLab (2019) and member of the EUMETSAT MTG-IRS Science Team. He is compromised with open source/access in Science, and is habitual panel evaluator for H2020 (ERC, FET), NSF, China and Swiss Science Foundations.

He coordinates the ‘Machine Learning for Earth and Climate Sciences' research program of ELLIS, the top network of excellence on AI in Europe. He was elevated to IEEE Fellow member (2018) in two Societies (Geosciences and Signal Processing) and to ELLIS Fellow (2019). Prof. Camps-Valls is the only researcher receiving two European Research Council (ERC) grants in two different areas: an ERC Consolidator (2015, Computer Science) and ERC Synergy (2019, Physical Sciences) grants to advance AI for Earth and Climate Sciences. In 2021 he became a Member of the ESSC panel part of the European Science Foundation (ESF), and in 2022 was elevated to Fellow of the European Academy of Sciences (EurASc), Fellow of the Academia Europeae (AE), and Fellow of Asia-Pacific Artificial Intelligence Association (AAIA).

Uncertainty quantification of weather forecasts is a necessity for reliably planning for and responding to extreme weather events in a warming world. This motivates the need for well-calibrated ensembles in probabilistic weather forecasting. We present initial results for the calibration of large-scale deep neural weather models for data-driven probabilistic weather forecasting. By explicitly accounting for uncertainties about the forecast's initial condition and model parameters, we generate ensemble forecasts that show promising results on standard diagnostics for probabilistic forecasts. Specifically, we are approaching the Integrated Forecasting System (IFS), the gold standard on probabilistic weather forecasting, on: (i) the spread-error agreement; and (ii) the Continuous Ranked Probability Score (CRPS). Our approach scales to state-of-the-art data-driven weather models, enabling cheap post-hoc calibration of pretrained models with tens of millions of parameters and paving the way towards the next generation of well-calibrated data-driven weather models.

How can we learn from all available data when training machine-learnt climate models, without incurring any extra cost at simulation time? Typically, the training data comprises coarse-grained high-resolution data. But only keeping this coarse-grained data means the rest of the high-resolution data is thrown out. We use a transfer learning approach, which can be applied to a range of machine learning models, to leverage all the high-resolution data. We use three chaotic systems to show it stabilises training, gives improved generalisation performance and results in better forecasting skill. Our code is at https://github.com/raghul-parthipan/dontwastedata

Decadal climate predictions are key to inform adaptation strategies in a warming climate. Coupled climate models used for decadal predictions are, however, imperfect representations of the climate system leading to forecast biases. Biases can also result from a poor model initialization that, when combined with forecast drift, can produce errors depending non-linearly on lead time. We propose a deep learning-based bias correction approach for the post-processing of gridded forecasts to enhance the accuracy of decadal predictions.

In contrast to the rapid digitalization of several industries, agriculture suffers from low adoption of climate-smart farming tools. Even though AI-driven digital agriculture can offer high-performing predictive functionalities, they lack tangible quantitative evidence on their benefits to the farmers. Field experiments can derive such evidence, but are often costly and time consuming. To this end, we propose an observational causal inference framework for the empirical evaluation of the impact of digital tools on target farm performance indicators. This way, we can increase farmers' trust via enhancing the transparency of the digital agriculture market, and in turn accelerate the adoption of technologies that aim to increase productivity and secure a sustainable and resilient agriculture against a changing climate. As a case study, we perform an empirical evaluation of a recommendation system for optimal cotton sowing, which was used by a farmers' cooperative during the growing season of 2021. We leverage agricultural knowledge to develop the causal graph of the farm system, we use the back-door criterion to identify the impact of recommendations on the yield and subsequently we estimate it using several methods on observational data. The results showed that a field sown according to our recommendations enjoyed a significant increase in yield 12% to 17%.

To fight climate change and accommodate the increasing population, global crop production has to be strengthened. To achieve the "sustainable intensification" of agriculture, transforming it from carbon emitter to carbon sink is a priority, and understanding the environmental impact of agricultural management practices is a fundamental prerequisite to that. At the same time, the global agricultural landscape is deeply heterogeneous, with differences in climate, soil, and land use inducing variations in how agricultural systems respond to farmer actions. The "personalization" of sustainable agriculture with the provision of locally adapted management advice is thus a necessary condition for the efficient uplift of green metrics, and an integral development in imminent policies. Here, we formulate personalized sustainable agriculture as a Conditional Average Treatment Effect estimation task and use Causal Machine Learning for tackling it. Leveraging climate data, land use information and employing Double Machine Learning, we estimate the heterogeneous effect of sustainable practices on the field-level Soil Organic Carbon content in Lithuania. We thus provide a data-driven perspective for targeting sustainable practices and effectively expanding the global carbon sink.

Poster sessions take place in the following Topia space: https://topia.io/neurips-2022-workshop-tccml

The links below provide access to the video presentations, Rocket.chat, direct Topia links and further materials featured on the workshop website.

Papers Track:

1. Function Approximations for Reinforcement Learning Controller for Wave Energy Converters
2. Image-Based Soil Organic Carbon Estimation from Multispectral Satellite Images with Fourier Neural Operator and Structural Similarity
3. SolarDK: A high-resolution urban solar panel image classification and localisation dataset
5. Optimizing toward efficiency for SAR image ship detection
6. AutoML-based Almond Yield Prediction and Projection in California
7. Attention-Based Scattering Network for Satellite Imagery
9. Aboveground carbon biomass estimate with Physics-informed deep network
10. Improving the predictions of ML-corrected climate models with novelty detection
12. Scene-to-Patch Earth Observation: Multiple Instance Learning for Land Cover Classification
13. Deep learning for downscaling tropical cyclone rainfall
14. Short-term Prediction and Filtering of Solar Power Using State-Space Gaussian Processes
15. Identifying latent climate signals using sparse hierarchical Gaussian processes
16. Towards dynamical stability analysis of sustainable power grids using Graph Neural Networks
17. Detecting Methane Plumes using PRISMA: Deep Learning Model and Data Augmentation
18. Probabilistic forecasting of regional photovoltaic power production based on satellite-derived cloud motion
19. Robustifying machine-learned algorithms for efficient grid operation
20. Deep Hydrology: Hourly, Gap-Free Flood Maps Through Joint Satellite and Hydrologic Modelling
24. Convolutional Neural Processes for Inpainting Satellite Images: Application to Water Body Segmentation
25. A POMDP Model for Safe Geological Carbon Sequestration
27. Deep Climate Change: A Dataset and Adaptive domain pre-trained Language Models for Climate Change Related Tasks
28. Data-Driven Optimal Solver for Coordinating a Sustainable and Stable Power Grid
29. Don't Waste Data: Transfer Learning to Leverage All Data for Machine-Learnt Climate Model Emulation
30. Explainable Multi-Agent Recommendation System for Energy-Efficient Decision Support in Smart Homes
31. FIRO: A Deep-neural Network for Wildfire Forecast with Interpretable Hidden States
32. Towards a spatially transferable super resolution model for downscaling Antarctic surface melt
33. Forecasting European Ozone Air Pollution With Transformers
34. Stability Constrained Reinforcement Learning for Real-Time Voltage Control
35. Land Use Prediction using Electro-Optical to SAR Few-Shot Transfer Learning
36. Exploring Randomly Wired Neural Networks for Climate Model Emulation
38. SustainGym: A Benchmark Suite of Reinforcement Learning for Sustainability Applications
39. Remote estimation of geologic composition using interferometric synthetic-aperture radar in California’s Central Valley
41. Temperature impacts on hate speech online: evidence from four billion tweets
42. Cross Modal Distillation for Flood Extent Mapping
43. Transformer Neural Networks for Building Load Forecasting
44. Estimating Chicago’s tree cover and canopy height using multi-spectral satellite imagery
45. Reconstruction of Grid Measurements in the Presence of Adversarial Attacks
46. Heat Demand Forecasting with Multi-Resolutional Representation of Heterogeneous Temporal Ensemble
48. Identifying Compound Climate Drivers of Forest Mortality with β-VAE
49. TCFD-NLP: Assessing alignment of climate disclosures using NLP for the financial markets
50. Deep Learning for Rapid Landslide Detection using Synthetic Aperture Radar (SAR) Datacubes
51. Hybrid Recurrent Neural Network for Drought Monitoring
52. Deep Learning for Global Wildfire Forecasting
53. Causal Modeling of Soil Processes for Improved Generalization
54. Machine Learning for Activity-Based Road Transportation Emissions Estimation
56. Estimating Corporate Scope 1 Emissions Using Tree-Based Machine Learning Methods
57. Analyzing Micro-Level Rebound Effects of Energy Efficient Technologies
58. Comparing the carbon costs and benefits of low-resource solar nowcasting
59. Climate Policy Radar: Pipeline for automated analysis of public climate policies
60. Inferring signatures of reinforcing ideology underlying carbon tax opposition
63. Curriculum Based Reinforcement Learning to Avert Cascading Failures in the Electric Grid
64. Short-range forecasts of global precipitation using deep learning-augmented numerical weather prediction
65. A Multi-Scale Deep Learning Framework for Projecting Weather Extremes
66. A Global Classification Model for Cities using ML
67. EnhancedSD: Predicting Solar Power Reanalysis from Climate Projections via Image Super-Resolution
68. Positional Encoder Graph Neural Networks for Geographic Data
70. Towards Global Crop Maps with Transfer Learning
71. Pyrocast: a Machine Learning Pipeline to Forecast Pyrocumulonimbus (PyroCb) clouds
72. Evaluating Digital Tools for Sustainable Agriculture using Causal Inference
73. Generating physically-consistent high-resolution climate data with hard-constrained neural networks
75. Flood Prediction with Graph Neural Networks
76. Neural Representation of the Stratospheric Ozone Chemistry
78. Industry-scale CO2 Flow Simulations with Model-Parallel Fourier Neural Operators
79. Adaptive Bias Correction for Improved Subseasonal Forecast
82. An Interpretable Model of Climate Change Using Correlative Learning
83. Multimodal Wildland Fire Smoke Detection
84. Using uncertainty-aware machine learning models to study aerosol-cloud interactions
85. Accessible Large-Scale Plant Pathology Recognition
86. Dynamic weights enabled Physics-Informed Neural Network for simulating the mobility of Engineered Nano Particles in a contaminated aquifer
88. Learning to forecast vegetation greenness at fine resolution over Africa with ConvLSTMs
89. Generative Modeling of High-resolution Global Precipitation Forecasts
91. Continual VQA for Disaster Response Systems
92. Performance evaluation of deep segmentation models on Landsat-8 imagery

Proposals Track:

Tutorials Track:

113. Disaster Risk Monitoring Using Satellite Imagery
114. Machine Learning for Predicting Climate Extremes
115. FourCastNet: A practical introduction to a state-of-the-art deep learning global weather emulator
116. Automating the creation of LULC datasets

Title: Mitigating climate and air pollutions from the electricity and transportation sectors in the United States

Abstract: In this talk, I will cover 3 recent pieces: (1) A transition to sustainable, deeply decarbonized,, and equitable energy systems is needed in the United States, which will require changes in the way we provide electricity and transportation services. With an increasing interconnected system that encompasses variable energy sources and complex markets, the emissions embedded in electricity generation and consumption are becoming more difficult to estimate. Using flow tracing and consumption-based accounting, we have characterized the health damages from exposure to PM2.5 from electricity imports and find that that 8% of our estimated premature deaths from electricity consumption in the United States are due to electricity imports. There is large geographic heterogeneity, with the most impacts occurring in the Midwest. While the West Coast has much cleaner generation and lower impacts overall, in many West Coast Balancing Areas, more than 50% of the estimated premature mortality associated with electricity consumption is caused by electricity imports, with some groups experiencing larger impacts than others. (2) Vehicle electrification is very likely needed moving forward as we decarbonize the transportation sector We estimate net emissions from vehicle electrification depend on when vehicles are charged, and which types of plants are meeting that electricity demand. We define a new concept, the grid critical emissions factors (CEFs), as the emission intensity of the grid that needs to be achieved when electric vehicles are charging so that electric vehicles achieve lifecycle greenhouse gas emissions parity with some of the most efficient gasoline and hybrid vehicles across the US. We find that the Nissan Leaf and Chevy Bolt battery electric vehicles reduce lifecycle emissions relative to the Toyota Prius and the Honda Accord gasoline hybrids in most of United States. However, in rural counties of the Midwest and the South power grid marginal emissions reductions of up to 208 gCO2/kWh are still needed for these electric vehicles to have lower lifecycle emissions than the gasoline hybrids. With the exception of the Northeast and Florida, the longer-range Tesla Model S battery electric luxury sedan has higher emissions than the hybrids across the U.S., and the emissions intensity of the grid would need to decrease by up to 342 gCO2/kWh in some locations for this vehicle to be at emissions parity with the hybrid vehicles we studied. (3) Electric vehicles will contribute to emissions reductions in the United States, but their charging may challenge electricity grid operations. We present a data-driven, realistic model of charging demand that captures the diverse charging behaviours of future adopters in the US Western Interconnection. We study charging control and infrastructure build-out as critical factors shaping charging load and evaluate grid impact under rapid electric vehicle adoption with a detailed economic dispatch model of 2035 generation. We find that peak net electricity demand increases by up to 25% with forecast adoption and by 50% in a stress test with full electrification. Locally optimized controls and high home charging can strain the grid. Shifting instead to uncontrolled, daytime charging can reduce storage requirements, excess non-fossil fuel generation, ramping and emissions. Our results urge policymakers to reflect generation-level impacts in utility rates and deploy charging infrastructure that promotes a shift from home to daytime charging.

Bio: Inês M.L. Azevedo is Associate Professor in the Department of Energy Resources Engineering at Stanford University. She also serves as Senior Fellow for the Woods Institute for the Environment at Stanford University and Fellow for the Precourt Institute for Energy (PIE) at Stanford University. She is the co-director of the Bits&Watts Initiative from PIE at Stanford University. Prof. Azevedo’s research interests focus on how to transition to a sustainable, low carbon, affordable, and equitable energy system. She is interested in sustainability and energy issues where a systems approach is needed, by combining engineering and technology analysis with economic and decision science approaches. Her current interest is to address energy issues with particular focus on distributional effects and equity. She has published 100+ peer-reviewed journal papers. She has participated as an author and committee member in several National Research Council reports from the U.S. National Academy of Sciences. She was one of the Lead Authors for IPCC AR6 report on Climate Mitigation for the Energy chapter, and she is now also participating as Lead Author for the upcoming U.S. National Climate Assessment chapter on climate change mitigation. Prof. Azevedo is also contributing as a chapter author to the upcoming U.S. National Climate Assessment report. Prof. Azevedo has received the World Economic Forum’s “Young Scientists under 40” award in 2014, and the C3E Women in Clean Energy Research Award in 2017.

Title: Physics-Guided Deep Learning for Climate Science

Abstract: While deep learning has shown tremendous success in many scientific domains, it remains a grand challenge to incorporate first principles in a systematic manner into such models. In this talk, I will demonstrate how to incorporate physical principles such as symmetry, conservation, and multi-scale into deep neural networks for forecasting and uncertainty quantification. I will showcase the applications of these models to challenging problems in climate science. Our methods demonstrate significant improvement in physical consistency, sample efficiency, and generalization in complex spatiotemporal dynamics.

Bio: Dr. Rose Yu is an assistant professor at the University of California San Diego, Department of Computer Science and Engineering. She earned her Ph.D. in Computer Sciences at USC in 2017. She was subsequently a Postdoctoral Fellow at Caltech. Her research focuses on advancing machine learning techniques for large-scale spatiotemporal data analysis, with applications to sustainability, health, and physical sciences. Among her awards, she has won NSF CAREER Award, Faculty Research Award from JP Morgan, Facebook, Google, Amazon, and Adobe, Several Best Paper Awards, Best Dissertation Award at USC, and was nominated as one of the ’MIT Rising Stars in EECS’.

Poster sessions take place in the following Topia space: https://topia.io/neurips-2022-workshop-tccml

The links below provide access to the video presentations, Rocket.chat, direct Topia links and further materials featured on the workshop website.

Papers Track:

1. Function Approximations for Reinforcement Learning Controller for Wave Energy Converters
4. Bayesian inference for aerosol vertical profiles
21. Machine learning emulation of a local-scale UK climate model
23. Learning evapotranspiration dataset corrections from water cycle closure supervision
26. Optimizing Japanese dam reservoir inflow forecast for efficient operation
28. Data-Driven Optimal Solver for Coordinating a Sustainable and Stable Power Grid
32. Towards a spatially transferable super resolution model for downscaling Antarctic surface melt
36. Exploring Randomly Wired Neural Networks for Climate Model Emulation
38. SustainGym: A Benchmark Suite of Reinforcement Learning for Sustainability Applications
40. AutoML for Climate Change: A Call to Action
45. Reconstruction of Grid Measurements in the Presence of Adversarial Attacks
59. Climate Policy Radar: Pipeline for automated analysis of public climate policies
61. Controllable Generation for Climate Modeling
68. Positional Encoder Graph Neural Networks for Geographic Data
69. Image-based Early Detection System for Wildfires
71. Pyrocast: a Machine Learning Pipeline to Forecast Pyrocumulonimbus (PyroCb) clouds
74. Transformers for Fast Emulation of Atmospheric Chemistry Box Models
76. Neural Representation of the Stratospheric Ozone Chemistry
77. DL-Corrector-Remapper: A grid-free bias-correction deep learning methodology for data-driven high-resolution global weather forecasting
81. An Unsupervised Learning Perspective on the Dynamic Contribution to Extreme Precipitation Changes
82. An Interpretable Model of Climate Change Using Correlative Learning
83. Multimodal Wildland Fire Smoke Detection
85. Accessible Large-Scale Plant Pathology Recognition
86. Dynamic weights enabled Physics-Informed Neural Network for simulating the mobility of Engineered Nano Particles in a contaminated aquifer
87. Calibration of Large Neural Weather Models
90. Learning Surrogates for Diverse Emission Models

Proposals Track:

94. Guided Transformer Network for Detecting Methane Emissions in Sentinel-2 Satellite Imagery
97. Deep-S2SWind: A data-driven approach for improving Sub-seasonal wind predictions
102. Deep learning-based bias adjustment of decadal climate predictions
104. Detecting Floods from Cloudy Scenes: A Fusion Approach Using Sentinel-1 and Sentinel-2 Imagery
108. Estimating Heating Loads in Alaska using Remote Sensing and Machine Learning Methods
109. Interpretable Spatiotemporal Forecasting of Arctic Sea Ice Concentration at Seasonal Lead Times
112. Personalizing Sustainable Agriculture with Causal Machine Learning

Tutorials Track:

116. Automating the creation of LULC datasets

Measuring and attributing greenhouse gas (GHG) emissions remains a challenging problem as the world strives towards meeting emissions reductions targets. As a significant portion of total global emissions, the road transportation sector represents an enormous challenge for estimating and tracking emissions at a global scale. To meet this challenge, we have developed a hybrid approach for estimating road transportation emissions that combines the strengths of machine learning and satellite imagery with localized emissions factors data to create an accurate, globally scalable, and easily configurable GHG monitoring framework.

In forest carbon credit, the concept of dynamic (or ex-post) baseline has been discussed to overcome the criticism of junk carbon credit, while an ex-ante baseline is still necessary in terms of project finance and risk assessment. We propose a Bayesian state-space SCM, which integrates both ex-ante and ex-post baseline estimation in a time-series causal inference framework. We apply the proposed model to a REDD+ project in Brazil, and show that it might have had a small, positive effect but had been over-credited and that the 90% predictive interval of the ex-ante baseline included the ex-post baseline, implying our ex-ante estimation can work effectively.

Natural disasters such as flood, wildfire, drought, and severe storms wreak havoc throughout the world, causing billions of dollars in damages, and uprooting communities, ecosystems, and economies. Unfortunately, flooding events are on the rise due to climate change and sea level rise. The ability to detect and quantify them can help us minimize their adverse impacts on the economy and human lives. Using satellites to study flood is advantageous since physical access to flooded areas is limited and deploying instruments in potential flood zones can be dangerous. We are proposing a hands-on tutorial to highlight the use of satellite imagery and computer vision to study natural disasters. Specifically, we aim to demonstrate the development and deployment of a flood detection model using Sentinel-1 satellite data. The tutorial will cover relevant fundamental concepts as well as the full development workflow of a deep learning-based application. We will include important considerations such as common pitfalls, data scarcity, augmentation, transfer learning, fine-tuning, and details of each step in the workflow. Importantly, the tutorial will also include a case study on how the application was used by authorities in response to a flood event. We believe this tutorial will enable machine learning practitioners of all levels to develop new technologies that tackle the risks posed by climate change. We expect to deliver the below learning outcomes:•Develop various deep learning-based computer vision solutions using hardware-accelerated open-source tools that are optimized for real-time deployment•Create an optimized pipeline for the machine learning development workflow•Understand different performance metrics for model evaluation that are relevant for real world datasets and data imbalances•Understand the public sector’s efforts to support climate action initiatives and point out where the audience can contribute

Climate change has led to a rapid increase in the occurrence of extreme weather events globally, including floods, droughts, and wildfires. In the longer term, some regions will experience aridification while others will risk sinking due to rising sea levels. Typically, such predictions are done via weather and climate models that simulate the physical interactions between the atmospheric, oceanic, and land surface processes that operate at different scales. Due to the inherent complexity, these climate models can be inaccurate or computationally expensive to run, especially for detecting climate extremes at high spatiotemporal resolutions. In this tutorial, we aim to introduce the participants to machine learning approaches for addressing two fundamental challenges. We will walk the participants through a hands-on tutorial for predicting climate extremes relating to temperature and precipitation in 2 setups: (1) temporal forecasting: the goal is to predict climate variables into the future (both direct single step approaches and iterative approaches that roll out the model for several timesteps), and (2) spatial downscaling: the goal is to learn a mapping that transforms low-resolution outputs of climate models into high-resolution regional forecasts. Through introductory presentations and colab notebooks, we aim to expose the participants to (a) APIs for accessing and navigating popular repositories that host global climate data, such as the Copernicus data store, (b) identifying relevant datasets, including auxiliary data (e.g., other climate variables such as geopotential), (c) scripts for downloading and preprocessing relevant datasets, (d) algorithms for training machine learning models, (d) metrics for evaluating model performance, and (e) visualization tools for both the dataset and predicted outputs. The coding notebooks will be in Python. No prior knowledge of climate science is required.

Subseasonal forecasting — predicting temperature and precipitation 2 to 6 weeks ahead — is critical for effective water allocation, wildfire management, and drought and flood mitigation. Recent international research efforts have advanced the subseasonal capabilities of operational dynamical models, yet temperature and precipitation prediction skills remains poor, partly due to stubborn errors in representing atmospheric dynamics and physics inside dynamical models. To counter these errors, we introduce an adaptive bias correction (ABC) method that combines state-of-the-art dynamical forecasts with observations using machine learning. When applied to the leading subseasonal model from the European Centre for Medium-Range Weather Forecasts (ECMWF), ABC improves temperature forecasting skill by 60-90% and precipitation forecasting skill by 40-69% in the contiguous U.S. We couple these performance improvements with a practical workflow, based on Cohort Shapley, for explaining ABC skill gains and identifying higher-skill windows of opportunity based on specific climate conditions.

Data-driven models, such as FourCastNet (FCN), have shown exemplary performance in high-resolution global weather forecasting. This performance, however, is based on supervision on mesh-gridded weather data without the utilization of raw climate observational data, the gold standard ground truth. In this work we develop a methodology to correct, remap, and fine-tune gridded uniform forecasts of FCN so it can be directly compared against observational ground truth, which is sparse and non-uniform in space and time. This is akin to bias-correction and post-processing of numerical weather prediction (NWP), a routine operation at meteorological and weather forecasting centers across the globe. The Adaptive Fourier Neural Operator (AFNO) architecture is used as the backbone to learn continuous representations of the atmosphere. The spatially and temporally non-uniform output is evaluated by the non-uniform discrete inverse Fourier transform (NUIDFT) given the output query locations. We call this network the Deep-Learning-Corrector-Remapper (DLCR). The improvement in DLCR’s performance against the gold standard ground truth over the baseline’s performance shows its potential to correct, remap, and fine-tune the mesh-gridded forecasts under the supervision of observations.

Weather extremes are a major societal and economic hazard, claiming thousands of lives and causing billions of dollars in damage every year. Under climate change, their impact and intensity are expected to worsen significantly. Unfortunately, general circulation models (GCMs), which are currently the primary tool for climate projections, cannot characterize weather extremes accurately. To address this, we present a multi-resolution deep-learning framework that, firstly, corrects a GCM's biases by matching low-order and tail statistics of its output with observations at coarse scales; and secondly, increases the level of detail of the debiased GCM output by reconstructing the finer scales as a function of the coarse scales. We use the proposed framework to generate statistically realistic realizations of the climate over Western Europe from a simple GCM corrected using observational atmospheric reanalysis. We also discuss implications for probabilistic risk assessment of natural disasters in a changing climate.

Several wildfire danger systems have emerged from decades of research. One such system is the National Fire-Danger Rating System (NFDRS), which is used widely across the United States and is a key predictor in the Global ECMWF Fire Forecasting (GEFF) model. The NFDRS is composed of over 100 equations relating wildfire risk to weather conditions, climate and land cover characteristics, and fuel. These equations and the corresponding 130+ parameters were developed via field and lab experiments. These parameters, which are fixed in the standard NFDRS and GEFF implementations, may not be the most appropriate for a climate-changing world. In order to adjust the NFDRS parameters to current climate conditions and specific geographical locations, we recast NFDRS in PyTorch to create a new deep learning-based Fire Index Risk Optimizer (FIRO). FIRO predicts the ignition component, or the probability a wildfire would require suppression in the presence of a firebrand, and calibrates the uncertain parameters for a specific region and climate conditions by training on observed fires. Given the rare occurrence of wildfires, we employed the extremal dependency index (EDI) as the loss function. Using ERA5 reanalysis and MODIS burned area data, we trained FIRO models for California, Texas, Italy, and Madagascar. Across these four geographies, the average EDI improvement was 175% above the standard NFDRS implementation

The increasing intensity and frequency of floods is one of the many consequences of our changing climate. In this work, we explore ML techniques that improve the flood detection module of an operational early flood warning system. Our method exploits an unlabelled dataset of paired multi-spectral and Synthetic Aperture Radar (SAR) imagery to reduce the labeling requirements of a purely supervised learning method. Past attempts have used such unlabelled data by creating weak labels out of them, but end up learning the label mistakes in those weak labels. Motivated by knowledge distillation and semi supervised learning, we explore the use of a teacher to train a student with the help of a small hand labeled dataset and a large unlabelled dataset. Unlike the conventional self distillation setup, we propose a cross modal distillation framework that transfers supervision from a teacher trained on richer modality (multi-spectral images) to a student model trained on SAR imagery. The trained models are then tested on the Sen1Floods11 dataset. Our model outperforms the Sen1Floods11 SAR baselines by an absolute margin of 4.15% mean Intersection-over-Union (mIoU) on the test split.

Poster sessions take place in the following Topia space: https://topia.io/neurips-2022-workshop-tccml

The links below provide access to the video presentations, Rocket.chat, direct Topia links and further materials featured on the workshop website.

Papers Track:

1. Function Approximations for Reinforcement Learning Controller for Wave Energy Converters
4. Bayesian inference for aerosol vertical profiles
11. Levee protected area detection for improved flood risk assessment in global hydrology models
22. Bridging the Microwave Data Gap; Using Bayesian Deep Learning to “See” the Unseen
28. Data-Driven Optimal Solver for Coordinating a Sustainable and Stable Power Grid
36. Exploring Randomly Wired Neural Networks for Climate Model Emulation
37. Closing the Domain Gap -- Blended Synthetic Imagery for Climate Object Detection
45. Reconstruction of Grid Measurements in the Presence of Adversarial Attacks
55. Bayesian State-Space SCM for Deforestation Baseline Estimation for Forest Carbon Credit
62. Learn to Bid: Deep Reinforcement Learning with Transformer for Energy Storage Bidding in Energy and Contingency Reserve Markets
80. Nowformer : A Locally Enhanced Temporal Learner for Precipitation Nowcasting
83. Multimodal Wildland Fire Smoke Detection
86. Dynamic weights enabled Physics-Informed Neural Network for simulating the mobility of Engineered Nano Particles in a contaminated aquifer
93. Synthesis of Realistic Load Data: Adversarial Networks for Learning and Generating Residential Load Patterns

Proposals Track:

94. Guided Transformer Network for Detecting Methane Emissions in Sentinel-2 Satellite Imagery
95. Identification of medical devices using machine learning on distribution feeder data for informing power outage response
98. Towards Low Cost Automated Monitoring of Life Below Water to De-risk Ocean-Based Carbon Dioxide Removal and Clean Power
108. Estimating Heating Loads in Alaska using Remote Sensing and Machine Learning Methods

Soil organic carbon (SOC) sequestration is the transfer and storage of atmospheric carbon dioxide in soils, which plays an important role in climate change mitigation. SOC concentration can be improved by proper land use, thus it is beneficial if SOC can be estimated at a regional or global scale. As multispectral satellite data can provide SOC-related information such as vegetation and soil properties at a global scale, estimation of SOC through satellite data has been explored as an alternative to manual soil sampling. Although existing works show promising results, most studies are based on pixel-based approaches with traditional machine learning methods, and convolutional neural networks (CNNs) are seldom used. To study the advantages of using CNNs on SOC remote sensing, in this paper, we propose the FNO-DenseNet based on the state-of-the-art Fourier neural operator (FNO). By combining the advantages of the FNO and DenseNet, the FNO-DenseNet outperformed the FNO in our experiments with hundreds of times fewer parameters. The FNO-DenseNet also outperformed a pixel-based random forest by 18% in the mean absolute percentage error. To the best of our knowledge, this is the first work of applying the FNO on SOC remote sensing.

The body of research on classification of solar panel arrays from aerial imagery is increasing, yet there are still not many public benchmark datasets. This paper introduces two novel benchmark datasets for classifying and localizing solar panel arrays in Denmark: A human annotated dataset for classification and segmentation, as well as a classification dataset acquired using self-reported data from the Danish national building registry. We explore the performance of prior works on the new benchmark dataset, and present results after fine-tuning models using a similar approach as recent works. Furthermore, we train models of newer architectures and provide benchmark baselines to our datasets in several scenarios. We believe the release of these datasets may improve future research in both local and global geospatial domains for identifying and mapping of solar panel arrays from aerial imagery. The data is accessible at https://osf.io/aj539/.

Aerosol-cloud interactions constitute the largest source of uncertainty in assessments of the anthropogenic climate change. This uncertainty arises in part from the difficulty in measuring the vertical distributions of aerosols. We often have to settle for less informative vertically aggregated proxies such as aerosol optical depth (AOD). In this work, we develop a framework to infer vertical aerosol profiles using AOD and readily available vertically resolved meteorological predictors such as temperature or relative humidity. We devise a simple Gaussian process prior over aerosol vertical profiles and update it with AOD observations. We validate our approach using ECHAM-HAM aerosol-climate model data. Our results show that, while simple, our model is able to reconstruct realistic extinction profiles with well-calibrated uncertainty. In particular, the model demonstrates a faithful reconstruction of extinction patterns arising from aerosol water uptake in the boundary layer.

The detection and prevention of illegal fishing is critical to maintaining a healthy and functional ecosystem. Recent research on ship detection in satellite imagery has focused exclusively on performance improvements, disregarding detection efficiency. However, the speed and compute cost of vessel detection are essential for a timely intervention to prevent illegal fishing. Therefore, we investigated optimization methods that lower detection time and cost with minimal performance loss. We trained an object detection model based on a convolutional neural network (CNN) using a dataset of satellite images. Then, we designed two efficiency optimizations that can be applied to the base CNN or any other base model. The optimizations consist of a fast, cheap classification model and a statistical algorithm. The integration of the optimizations with the object detection model leads to a trade-off between speed and performance. We studied the trade-off using metrics that give different weight to execution time and performance. We show that by using a classification model the average precision of the detection model can be approximated to 0.5% in 44% of the time or to 7.3% in 25% of the time.

Multi-channel satellite imagery, from stacked spectral bands or spatiotemporal data, have meaningful representations for various atmospheric properties. Combining these features in an effective manner to create a performant and trustworthy model is of utmost importance to forecasters. Neural networks show promise, yet suffer from unintuitive computations, fusion of high-level features, and may be limited by the quantity of available data. In this work, we leverage the scattering transform to extract high-level features without additional trainable parameters and introduce a separation scheme to bring attention to independent input channels. Experiments show promising results on estimating tropical cyclone intensity and predicting the occurrence of lightning from satellite imagery.

Inaccuracies in the models of the Earth system, i.e., structural and parametric model errors, lead to inaccurate climate change projections.Errors in the model can originate from unresolved phenomena due to a low numerical resolution, as well as misrepresentations of physical phenomena or boundaries (e.g., orography). Therefore, such models lead to inaccurate short--term forecasts of weather and extreme events, and more importantly, long term climate projections. While calibration methods have been introduced to address for parametric uncertainties, e.g., by better estimation of system parameters from observations, addressing structural uncertainties, especially in an interpretable manner, remains a major challenge.Therefore, with increases in both the amount and frequency of observations of the Earth system, algorithmic innovations are required to identify interpretable representations of the model errors from observations. We introduce a flexible, general-purpose framework to discover interpretable model errors, and show its performance on a canonical prototype of geophysical turbulence, the two--level quasi--geostrophic system. Accordingly, a Bayesian sparsity--promoting regression framework is proposed, that uses a library of kernels for discovery of model errors. As calculating the library from noisy and sparse data (e.g., from observations) using convectional techniques leads to interpolation errors, here we use a coordinate-based multi--layer embedding to impute the sparse observations. We demonstrate the importance of alleviating spectral bias, and propose a random Fourier feature layer to reduce it in the proposed embeddings, and subsequently enable an accurate discovery. Our framework is demonstrated to successfully identify structural model errors due to linear and nonlinear processes (e.g., radiation, surface friction, advection), as well as misrepresented orography.

While previous works have shown that machine learning (ML) can improve the prediction accuracy of coarse-grid climate models, these ML-augmented methods are more vulnerable to irregular inputs than the traditional physics-based models they rely on. Because ML-predicted corrections feed back into the climate model’s base physics, the ML-corrected model regularly produces out of sample data, which can cause model instability and frequent crashes. This work shows that adding semi-supervised novelty detection to identify out-of-sample data and disable the ML-correction accordingly stabilizes simulations and sharply improves the quality of predictions. We design an augmented climate model with a one-class support vector machine (OCSVM) novelty detector that provides better temperature and precipitation forecasts in a year-long simulation than either a baseline (no-ML) or a standard ML-corrected run. By improving the accuracy of coarse-grid climate models, this work helps make accurate climate models accessible to researchers without massive computational resources.

Power outages caused by extreme weather events due to climate change have doubled in the United States in the last two decades. Outages pose severe health risks to over 4.4 million individuals dependent on in-home medical devices. Data on the number of such individuals residing in a given area is limited. This study proposes a load disaggregation model to predict the number of medical devices behind an electric distribution feeder. This data can be used to inform planning and response to power outages and other emergencies. The proposed solution serves as a measure for climate change adaptation.

Precise flood risk assessment is needed to reduce human societies vulnerability as climate change increases hazard risk and exposure related to floods. Levees are built to protect people and goods from flood, which alters river hydrology, but are still not accounted for by global hydrological model. Detecting and integrating levee structures to global hydrological simulations is thus expected to enable more precise flood simulation and risk assessment, with important consequences for flood risk mitigation. In this work, we propose a new formulation to the problem of identifying levee structures: instead of detecting levees themselves, we focus on segmenting the region of the floodplain they protect. This formulation allows to better identify protected areas, to leverage the structure of hydrological data, and to simplify the integration of levee information to global hydrological models.

Flooding is often the leading cause of mortality and damages from tropical cyclones. With rainfall from tropical cyclones set to rise under global warming, better estimates of extreme rainfall are required to better support resilience efforts. While high resolution climate models capture tropical cyclone statistics well, they are computationally expensive leading to a trade-off between accuracy and generating enough ensemble members to generate sufficient high impact, low probability events. Often, downscaling models are used as a computationally cheaper alternative. Here, we develop and evaluate a set of deep learning models for downscaling tropical cyclone rainfall for more robust risk analysis.

To transform our economy towards net-zero emissions, industrial development of clean energy technologies (CETs) to replace fossil energy technologies (FETs) is crucial. Although the media has great power in influencing consumer behavior and decision making in business and politics, its role in the energy transformation is still underexplored. In this paper, we analyze the global energy discourse via machine learning. For this, we collect a large-scale dataset with ~5 million news articles from seven of the world’s major CO2 emitting countries, covering eight CETs and four FETs. Using machine learning, we then analyze the content of news articles on a highly granular level and along several dimensions, namely relevance (for the energy discourse), context (e.g., costs, regulation, investment), and connotations (e.g., high/increasing vs. low/decreasing costs). By linking empirical discourse patterns to investment and deployment data of CETs and FETs, this study advances the current understanding about the role of the media in the energy transformation. Thereby, it enables businesses, investors, and policy makers to respond more effectively to sensitive topics in the media discourse and leverage windows of opportunity for scaling CETs.

Short-term forecasting of solar photovoltaic energy (PV) production is important for powerplant management. Ideally these forecasts are equipped with error bars, so that downstream decisions can account for uncertainty. To produce predictions with error bars in this setting, we consider Gaussian processes (GPs) for modelling and predicting solar photovoltaic energy production in the UK. A standard application of GP regression on the PV timeseries data is infeasible due to the large data size and non-Gaussianity of PV readings. However, this is made possible by leveraging recent advances in scalable GP inference, in particular, by using the state-space form of GPs, combined with modern variational inference techniques. The resulting model is not only scalable to large datasets but can also handle continuous data streams via Kalman filtering.

Extracting latent climate signals from multiple climate model simulations is important to estimate future climate change. To tackle this we develop a sparse hierarchical Gaussian process (SHGP), which probabilistically learns a latent distribution from a set of vectors. We use this to predict the latent surface temperature change globally and for central England from an ensemble of climate models, in a scalable manner and with robust uncertainty propagation.

To mitigate climate change, the share of renewable needs to be increased. Renewable energies introduce new challenges to power grids due to decentralization, reduced inertia and volatility in production. The operation of sustainable power grids with a high penetration of renewable energies requires new methods to analyze the dynamical stability. We provide new datasets of dynamical stability of synthetic power grids, and find that graph neural networks (GNNs) are surprisingly effective at predicting the highly non-linear target from topological information only. To illustrate the potential to scale to real-sized power grids, we demonstrate the successful prediction on a Texan power grid model.

The new generation of hyperspectral imagers, such as PRISMA, has improvedsignificantly our detection capability of methane (CH4) plumes from space at highspatial resolution (∼30m). We present here a complete framework to identifyCH4 plumes using images from the PRISMA satellite mission and a deep learningtechnique able to automatically detect plumes over large areas. To compensatefor the sparse database of PRISMA images, we trained our model by transposinghigh resolution plumes from Sentinel-2 to PRISMA. Our methodology avoidscomputationally expensive synthetic plume from Large Eddy Simulations whilegenerating a broad and realistic training database, and paves the way for large-scale detection of methane plumes using future hyperspectral sensors (EnMAP, EMIT, CarbonMapper).

Solar energy generation drastically increased in the last years, and it is expected to grow even more in the next decades. So, accurate intra-day forecasts are needed to improve the predictability of the photovoltaic power production and associated balancing measures to increase the shares of renewable energy in the power grid. Most forecasting methods require numerical weather predictions, which are slow to compute, or long-term datasets to run the forecast. These issues make the models difficult to implement in an operational setting. To overcome these problems, we propose a novel regional intraday probabilistic PV power forecasting model able to exploit only 2 hours of satellite-derived cloudiness maps to produce the ensemble forecast. The model is easy to implement in an operational setting as it is based on Pysteps, an already-operational Python library for precipitation nowcasting. With few adaptations of the Steps algorithm, we reached state-of-the-art performance, reaching a 71% lower RMSE than the Persistence model and a 50% lower CRPS than the Persistence Ensemble model for forecast lead times up to 4 hours.

We propose a learning-augmented algorithm, RobustML, for operation of dispatchable generation that exploits the good performance of a machine-learned algorithm while providing worst-case guarantees on cost. We evaluate the algorithm on a realistic two-generator system, where it exhibits robustness to distribution shift while enabling improved efficiency as renewable penetration increases.

Climate change is causing the intensification of rainfall extremes. Precipitation projections with high spatial resolution are important for society to prepare for these changes, e.g. to model flooding impacts. Physics-based simulations for creating such projections are very computationally expensive. This work demonstrates the effectiveness of diffusion models, a form of deep generative models, for generating much more cheaply realistic high resolution rainfall samples for the UK conditioned on data from a low resolution simulation. We show for the first time a machine learning model that is able to produce realistic high-resolution rainfall predictions based on a physical model that resolves atmospheric convection, a key process behind extreme rainfall. By adding self-learnt, location-specific information to low resolution relative vorticity, quantiles and time-mean of the samples match well their counterparts from the high-resolution simulation.

Having microwave data with the spatial and temporal resolution of infrared data would provide a large positive impact on many climate and weather applications. We demonstrate that Bayesian deep learning is a promising technique for both creating and improving synthetic microwave data from infrared data. We report 0.7% mean absolute percentage error for 183+/-3 GHz microwave brightness temperature and uncertainty metrics and find that more training data is needed to achieve improved performance at 166 GHz, 37 GHz, and 23 GHz. Analysis of the spatial distribution of uncertainty reveals that additional cloud data will provide the greatest increase in skill, which will potentially allow for generation of many secondary products derived from microwave data in the future.

Oceans will play a crucial role in our efforts to combat the growing climate emergency. Researchers have proposed several strategies to harness greener energy through oceans and use oceans as carbon sinks. However, the risks these strategies might pose to the ocean and marine ecosystem are not well understood. It is imperative that we quickly develop a range of tools to monitor ocean processes and marine ecosystems alongside the technology to deploy these solutions on a large scale into the oceans. Large arrays of inexpensive cameras placed deep underwater coupled with machine learning pipelines to automatically detect, classify, count and estimate fish populations have the potential to continuously monitor marine ecosystems and help study the impacts of these solutions on the ocean. In this proposal, we discuss the challenges presented by a dark artificially lit underwater video dataset captured 500m below the surface, propose potential solutions to address these challenges, and present preliminary results from detecting and classifying 6 species of fish in deep underwater camera data.

Evapotranspiration (ET) is one of the most uncertain components of the global water cycle.Improving global ET estimates is needed to better our understanding of the global water cycle so as to forecast the consequences of climate change on the future of global water resource distribution.This work presents a methodology to derive monthly corrections of global ET datasets at 0.25 degree resolution. We use ML to generalize sparse catchment-level water cycle closure residual information to global and dense pixel-level residuals. Our model takes a probabilistic view on ET datasets and their correction that we use to regress catchment-level residuals using a sum-aggregated supervision. Using four global ET datasets, we show that our learned model has learned ET corrections that accurately generalize its water cycle-closure results to unseen catchments.

The widespread availability of satellite images has allowed researchers to monitor the impact of climate on socio-economic and environmental issues through examples like crop and water body classification to measure food scarcity and risk of flooding. However, a common issue of satellite images is missing values due to measurement defects, which render them unusable by existing methods without data imputation. To repair the data, inpainting methods can be employed, which are based on classical PDEs or interpolation methods. Recently, deep learning approaches have shown promise in this realm, however many of these methods do not explicitly take into account the inherent spatio-temporal structure of satellite images. In this work, we cast satellite image inpainting as a meta-learning problem, and implement Convolutional Neural Processes (ConvNPs) in which we frame each satellite image as its own task or 2D regression problem. We show that ConvNPs outperform classical methods and state-of-the-art deep learning inpainting models on a scanline problem for LANDSAT 7 satellite images, assessed on a variety of in- and out-of-distribution images. Our results successfully match the performance of clean images on a downstream water body segmentation task in Canada.

Geological carbon capture and sequestration (CCS), where CO2 is stored in subsurface formations, is a promising and scalable approach for reducing global emissions.However, if done incorrectly, it may lead to earthquakes and leakage of CO2 back to the surface, harming both humans and the environment. These risks are exacerbated by the large amount of uncertainty in the structure of the storage formation. For these reasons, we propose that CCS operations be modeled as a partially observable Markov decision process (POMDP) and decisions be informed using automated planning algorithms. To this end, we develop a simplified model of CCS operations based on a 2D spillpoint analysis that retains many of the challenges and safety considerations of the real-world problem. We show how off-the-shelf POMDP solvers outperform expert baselines for safe CCS planning. This POMDP model can be used as a test bed to drive the development of novel decision-making algorithms for CCS operations.

Despite a climate and topology favorable to hydropower (HP) generation, HP only accounts for 4% of today’s Japanese primary energy consumption mix. In recent years, calls for improving the efficiency of Japanese HP towards achieving a more sustainable energy mix have emerged from prominent voices in the Ministry of Land, Infrastructure, Transport and Tourism (MILT). Among potential optimizations, data-driven dam operation policies using accurate river discharge forecasts have been advocated for. In the meantime, Machine Learning (ML) has recently made important strides in hydrological modeling, with forecast accuracy improvements demonstrated on both precipitation nowcasting and river discharge prediction. We are motivated by the convergence of these societal and technological contexts: our final goal is to provide scientific evidence and actionable insights for dam infrastructure managers and policy makers to implement more energy-efficient and flood-resistant dam operation policies on a national scale. Towards this goal this work presents a preliminary study of ML-based dam inflow forecasts on a dataset of 127 Japanese public dams we assembled. We discuss our preliminary results and lay out a path for future studies.

The quantity and quality of literature around climate change (CC) and its impacts are increasing yearly. Yet, this field has received limited attention in the Natural Language Processing (NLP) community. With the help of large Language Models (LMs) and transfer learning, NLP can support policymakers, researchers, and climate activists in making sense of large-scale and complex CC-related texts. CC-related texts include specific language that general language models cannot represent accurately. Therefore we collected a climate change corpus consisting of over 360 thousand abstracts of top climate scientists' articles from trustable sources covering large temporal and spatial scales. Comparison of the performance of GPT2 LM and our 'climateGPT2 models', fine-tuned on the CC-related corpus, on claim generation (text generation) and fact-checking, downstream tasks show the better performance of the climateGPT2 models compared to the GPT2. The climateGPT2 models decrease the validation loss to 1.08 for claim generation from 43.4 obtained by GPT2. We found that climateGPT2 models improved the masked language model objective for the fact-checking task by increasing the F1 score from 0.67 to 0.72.

With today's pressing climate change concerns, the widespread integration of low-carbon technologies such as sustainable generation systems (e.g. photovoltaics, wind turbines, etc.) and flexible consumer devices (e.g. storage, electric vehicles, smart appliances, etc.) into the electric grid is vital. Although these power entities can be deployed at large, these are highly variable in nature and must interact with the existing grid infrastructure without violating electrical limits so that the system continues to operate in a stable manner at all times. In order to ensure the integrity of grid operations while also being economical, system operators will need to rapidly solve the optimal power flow (OPF) problem in order to adapt to these fluctuations. Inherent non-convexities in the OPF problem do not allow traditional model-based optimization techniques to offer guarantees on optimality, feasibility and convergence. In this paper, we propose a data-driven OPF solver built on information-theoretic and semi-supervised machine learning constructs. We show that this solver is able to rapidly compute solutions (i.e. in sub-second range) that are within 3% of optimality with guarantees on feasibility on a benchmark IEEE 118-bus system.

Transparent, understandable, and persuasive recommendations support the electricity consumers’ behavioral change to tackle the energy efficiency problem. This paper proposes an explainable multi-agent recommendation system for load shifting for household appliances. First, we extend a novel multi-agent approach by designing an Explainability Agent that provides explainable recommendations for optimal appliance scheduling in a textual and visual manner. Second, we enhance the predictive capacity of other agents by including weather data and applying state-of-the-art models (i.e., k-nearest-neighbours, extreme gradient boosting, adaptive boosting, random forest, logistic regression, and explainable boosting machines). Since we want to help the user understand a single recommendation, we focus on local explainability approaches. In particular, we apply post-model approaches LIME (local, interpretable, model-agnostic explanation) and SHAP (Shapley additive explanations) as model-agnostic tools that can explain the predictions of the chosen classifiers. We further provide an overview of the predictive and explainability performance. Our results show a substantial improvement in the performance of the multi-agent system while at the same time opening up the “black box” of recommendations. To show the pathway to positive impact regarding climate change, we provide a discussion on the potential impact of the suggested approach.

Surface melt on the Antarctic Ice Sheet is an important climate indicator, yet the spatial scale of modeling and observing surface melt is insufficient to capture crucial details and understand local processes. High-resolution climate models could provide a solution, but they are computationally expensive and require finetuning for some model parameters. An alternative method, pioneering in geophysics, is single-image super resolution (SR) applied on lower-resolution model output. However, often input and output of such SR models are available on the same, fixed spatial domain. High-resolution model simulations over Antarctica are available only in some regions. To be able to apply an SR model elsewhere, we propose to make the single-image SR model physics-aware, using surface albedo and elevation as additional input. Our results show a great improvement in the spatial transferability of the conventional SR model. Although issues with the input satellite-derived albedo remain, adding physics awareness paves a way toward a spatially transferable SR model for downscaling Antarctic surface melt.

Surface ozone is an air pollutant that contributes to hundreds of thousands of premature deaths annually. Accurate short-term ozone forecasts may allow improved policy to reduce the risk to health, such as air quality warnings. However, forecasting ozone is a difficult problem, as surface ozone concentrations are controlled by a number of physical and chemical processes which act on varying timescales. Accounting for these temporal dependencies appropriately is likely to provide more accurate ozone forecasts. We therefore deploy a state-of-the-art transformer-based model, the Temporal Fusion Transformer, trained on observational station data from three European countries. In four-day test forecasts of daily maximum 8-hour ozone, the novel approach is highly skilful (MAE = 4.6 ppb, R2 = 0.82), and generalises well to two European countries unseen during training (MAE = 4.9 ppb, R2 = 0.79). The model outperforms standard machine learning models on our data, and compares favourably to the published performance of other deep learning architectures tested on different data. We illustrate that the model pays attention to physical variables known to control ozone concentrations, and that the attention mechanism allows the model to use relevant days of past ozone concentrations to make accurate forecasts.

This paper is a summary of a recently submitted work. Deep Reinforcement Learning (DRL) has been recognized as a promising tool to address the challenges in real-time control of power systems. However, its deployment in real-world power systems has been hindered by a lack of explicit stability and safety guarantees. In this paper, we propose a stability constrained reinforcement learning method for real-time voltage control in both single-phase and three-phase distribution grids. The key idea underlying our approach is an explicitly constructed Lyapunov function that certifies stability. We demonstrate the effectiveness of our approach with IEEE test feeders, where the proposed method achieves the best overall performance, while always achieving voltage stability. In contrast, standard RL methods often fail to achieve voltage stability.

Satellite image analysis has important implications for land use, urbanization, and ecosystem monitoring. Deep learning methods can facilitate the analysis of different satellite modalities, such as electro-optical (EO) and synthetic aperture radar (SAR) imagery, by supporting knowledge transfer between the modalities to compensate for individual shortcomings. Recent progress has shown how distributional alignment of neural network embeddings can produce powerful transfer learning models by employing a sliced Wasserstein distance (SWD) loss. We analyze how this method can be applied to Sentinel-1 and -2 satellite imagery and develop several extensions toward making it effective in practice. In an application to few-shot Local Climate Zone (LCZ) prediction, we show that these networks outperform multiple common baselines on datasets with a large number of classes. Further, we provide evidence that instance normalization can significantly stabilize the training process and that explicitly shaping the embedding space using supervised contrastive learning can lead to improved performance.

Exploring the climate impacts of various anthropogenic emissions scenarios is key to making informed decisions for climate change mitigation and adaptation. State-of-the-art Earth system models can provide detailed insight into these impacts, but have a large associated computational cost on a per-scenario basis. This large computational burden has driven recent interest in developing cheap machine learning models for the task of climate model emulation. In this manuscript, we explore the efficacy of randomly wired neural networks for this task. We describe how they can be constructed and compare them to their standard feedforward counterparts using the ClimateBench dataset. Specifically, we replace the dense layers in multilayer perceptrons, convolutional neural networks, and convolutional long short-term memory networks with randomly wired ones and assess the impact on model performance for models with 1 million and 10 million parameters. We find average performance improvements of 4.2% across model complexities and prediction tasks, with substantial performance improvements of up to 16.4% in some cases. Furthermore, we find no significant difference in prediction speed between networks with standard feedforward dense layers and those with randomly wired layers. These findings indicate that randomly wired neural networks may be suitable direct replacements for traditional dense layers in many standard models.

Object detection models have great potential to increase both the frequency and cost-efficiency of assessing climate-relevant infrastructure in satellite imagery. However, model performance can suffer when models are applied to stylistically different geographies. We propose a technique to generate synthetic imagery using minimal labeled examples of the target object at a low computational cost. Our technique blends example objects onto unlabeled images of the target domain. We show that including these synthetic images improves the average precision of a YOLOv3 object detection model when compared to a baseline and other popular domain adaptation techniques.

The lack of standardized benchmarks for reinforcement learning (RL) in sustainability applications has made it difficult to both track progress on specific domains and identify bottlenecks for researchers to focus their efforts on. In this paper, we present SustainGym, a suite of two environments designed to test the performance of RL algorithms on realistic sustainability tasks. The first environment simulates the problem of scheduling decisions for a fleet of electric vehicle (EV) charging stations, and the second environment simulates decisions for a battery storage system bidding in an electricity market. We describe the structure and features of the environments and show that standard RL algorithms have significant room for improving performance. We discuss current challenges in introducing RL to real-world sustainability tasks, including physical constraints and distribution shift.

California's Central Valley is the national agricultural center, producing 1/4 of the nation’s food. However, land in the Central Valley is sinking at a rapid rate (as much as 20 cm per year) due to continued groundwater pumping. Land subsidence has a significant impact on infrastructure resilience and groundwater sustainability. In this study, we aim to identify specific regions with different temporal dynamics of land displacement and find relationships with underlying geological composition. Then, we aim to remotely estimate geologic composition using interferometric synthetic aperture radar (InSAR)-based land deformation temporal changes using machine learning techniques. We identified regions with different temporal characteristics of land displacement in that some areas (e.g., Helm) with coarser grain geologic compositions exhibited potentially reversible land deformation (elastic land compaction). We found a significant correlation between InSAR-based land deformation and geologic composition using random forest and deep neural network regression models. We also achieved significant accuracy with 1/4 sparse sampling to reduce any spatial correlations among data, suggesting that the model has the potential to be generalized to other regions for indirect estimation of geologic composition. Our results indicate that geologic composition can be estimated using InSAR-based land deformation data. In-situ measurements of geologic composition can be expensive and time consuming and may be impractical in some areas. The generalizability of the model sheds light on high spatial resolution geologic composition estimation utilizing existing measurements.

The challenge that climate change poses to humanity has spurred a rapidly developing field of artificial intelligence research focused on climate change applications. The climate change ML (CCML) community works on a diverse, challenging set of problems which often involve physics-constrained ML or heterogeneous spatiotemporal data. It would be desirable to use automated machine learning (AutoML) techniques to automatically find high-performing architectures and hyperparameters for a given dataset. In this work, we benchmark popular Auto ML libraries on three high-leverage CCML applications: climate modeling, wind power forecasting, and catalyst discovery. We find that out-of-the-box AutoML libraries currently fail to meaningfully surpass the performance of human-designed CCML models. However, we also identify a few key weaknesses, which stem from the fact that most AutoML techniques are tailored to computer vision and NLP applications. For example, while dozens of search spaces have been designed for image and language data, none have been designed for spatiotemporal data. Addressing these key weaknesses can lead to the discovery of novel architectures that yield substantial performance gains across numerous CCML applications. Therefore, we present a call to action to the AutoML community, since there are a number of concrete, promising directions for future work in the space of AutoML for CCML. We release our code and a list of resources at https://github.com/climate-change-automl/climate-change-automl.

Human aggression is no longer limited to the physical space but exists in the form of hate speech on social media. Here, we examine the effect of temperature on the occurrence of hate speech on Twitter and interpret the results in the context of climate change, human behavior and mental health. Employing supervised machine learning models, we identify hate speech in a data set of four billion geolocated tweets from over 750 US cities (2014 – 2020). We statistically evaluate the changes in daily hate tweets against changes in local temperature, isolating the temperature influence from confounding factors using binned panel-regression models. We find a low prevalence of hate tweets in moderate temperatures and observe sharp increases of up to 12% for colder and up to 22% for hotter temperatures, indicating that not only hot but also cold temperatures increase aggressive tendencies. Further, we observe that for extreme temperatures hate speech also increases as a percentage of total tweeting activity, crowding out non-hate speech. The quasi-quadratic shape of the temperature-hate tweet curve is robust across varying climate zones, income groups, religious and political beliefs. The prevalence of the results across climatic and socioeconomic splits points to limits in adaptation. Our results illuminate hate speech online as an impact channel through which temperature alters societal aggression.

Accurate electrical load forecasts of buildings are needed to optimize local energystorage and make use of demand-side flexibility. We study the usage of Trans-former neural networks for short-term electrical load forecasting of 370 buildingsfrom a public dataset. On some buildings Transformer neural networks give thebest forecasts, and on others multi-layer perceptrons are better. In addition, westudy whether models trained on a subset of the buildings generalize to unseenbuildings, and find that Transformer neural networks generalize better than multi-layer perceptrons and our statistical baselines.

Information on urban tree canopies is fundamental to mitigating climate change as well as improving quality of life. Urban tree planting initiatives face a lack of up-to-date data about the horizontal and vertical dimensions of the tree canopy in cities. We present a pipeline that utilizes LiDAR data as ground-truthand then trains a multi-task machine learning model to generate reliable estimatesof tree cover and canopy height in urban areas using multi-source multi-spectralsatellite imagery for the case study of Chicago.

In efforts to mitigate the adverse effects of climate change, policymakers have set ambitious goals to reduce the carbon footprint of all sectors - including the electric grid. To facilitate this, sustainable energy systems like renewable generation must { be} deployed at high numbers throughout the grid. As these are highly variable in nature, the grid must be closely monitored so that system operators will have elevated situational awareness and can execute timely actions to maintain stable grid operations. With the widespread deployment of sensors like phasor measurement units (PMUs), an abundance of data is available for conducting accurate state estimation. However, due to the cyber-physical nature of the power grid, measurement data can be perturbed in an adversarial manner to enforce incorrect decision-making. In this paper, we propose a novel reconstruction method that leverages on machine learning constructs like CGAN and gradient search to recover the original states when subjected to adversarial perturbations. Experimental studies conducted on the practical IEEE 118-bus benchmark power system show that the proposed method can reduce errors due to perturbation by large margins (i.e. up to 100%).

One of the primal challenges faced by utility companies is ensuring efficient supply with minimal greenhouse gas emissions. The advent of smart meters and smart grids provide an unprecedented advantage in realizing an optimised supply of thermal energies through proactive techniques such as load forecasting. In this paper, we propose a forecasting framework for heat demand based on neural networks where the time series are encoded as scalograms equipped with the capacity of embedding exogenous variables such as weather, and holiday/non-holiday. Subsequently, CNNs are utilized to predict the heat load multi-step ahead. Finally, the proposed framework is compared with other state-of-the-art methods, such as SARIMAX and LSTM. The quantitative results from retrospective experiments show that the proposed framework consistently outperforms the state-of-the-art baseline method with real-world data acquired from Denmark. A minimal mean error of 7.54% for MAPE and 417kW for RMSE is achieved with the proposed framework in comparison to all other methods.

Ceilometers use a laser beam to capture certain phenomena in the atmosphere like clouds, precipitation, or aerosol layers. These measurements can be visualized in so-called quick looks that at the moment are mostly analyzed manually by meteorology experts. In this work, we illustrate the path towards the automatic analysis of quick looks by using a hybrid approach combining an image segmentation algorithm with unsupervised representation learning and clustering. We present a first proof of concept and give an outlook on possible future work.

Wind energy’s ability to liberate the world of conventional sources of energy relies on lowering the significant costs associated with the maintenance of wind turbines. Since icing events on turbine rotor blades are a leading cause of operational failures, identifying icing in advance is critical. Some recent studies have utilised deep learning techniques to predict icing events with high accuracy by leveraging rotor blade images, but these studies only focus on specific wind parks and fail to generalize to unseen scenarios (e.g. new rotor blade designs). We propose the utilisation of synthetic data augmentation via neural style transfer to improve the generalization of existing ice prediction models. We show that training models with augmented data that captures domain-invariant icing characteristics can help improve predictive performance across multiple wind parks. Through efficient identification of icing, this study can support preventive maintenance of wind energy sources by making them more reliable towards tackling climate change.

Light goods vehicles (LGV) used extensively in the last mile of delivery are one of the leading polluters in cities. Cargo-bike logistics has been put forward as a high impact candidate for replacing LGVs, with experts estimating over half of urban van deliveries being replaceable by cargo bikes, due to their faster speeds, shorter parking times and more efficient routes across cities. By modelling the relative delivery performance of different vehicle types across urban micro-regions, machine learning can help operators evaluate the business and environmental impact of adding cargo-bikes to their fleets. In this paper, we introduce two datasets, and present initial progress in modelling urban delivery service time (e.g. cruising for parking, unloading, walking). Using Uber’s H3 index to divide the cities into hexagonal cells, and aggregating OpenStreetMap tags for each cell, we show that urban context is a critical predictor of delivery performance.

With the advent of climate change and global warming, the Greenland Ice Sheet (GrIS) has been melting at an alarming rate, losing over 215 Gt per yr, and accounting for 10% of mean global sea level rise since the 1990s. It is imperative to understand what dynamics are causing ice loss and influencing ice flow in order to successfully project mass changes of ice sheets and associated sea level rise. This work applies machine learning, ice thickness data, and horizontal ice velocity measurements from satellite radar data to quantify the magnitudes and distributions of the basal traction forces that are holding the GrIS back from flowing into the ocean. Our approach uses a hybrid model: InSAR velocity data trains a linear regression model, and these model coefficients are fed into a geophysical algorithm to estimate basal tractions that capture relationships between the ice motion and physical variables. Results indicate promising model performance and reveal significant presence of large basal traction forces around the coastline of the GrIS.

With climate change predicted to increase the likelihood of landslide events, there is a growing need for rapid landslide detection technologies that help inform emergency responses. Synthetic Aperture Radar (SAR) is a remote sensing technique that can provide measurements of affected areas independent from weather or lighting conditions. Usage of SAR, however, is hindered by domain knowledge that is necessary for the pre-processing steps and its interpretation requires expert knowledge. We provide simplified, pre-processed, machine-learning ready SAR datacubes for four globally located landslide events obtained from several Sentinel-1 satellite passes before and after a landslide triggering event together with segmentation maps of the landslides. From this dataset, using the Hokkaido, Japan datacube, we study the feasibility of SAR-based landslide detection with supervised deep learning (DL). Our results demonstrate that DL models can be used to detect landslides from SAR data, achieving an Area under the Precision-Recall curve exceeding 0.7. We find that additional satellite visits enhance detection performance, but that early detection is possible when SAR data is combined with terrain information from a digital elevation model. This can be especially useful for time-critical emergency interventions.

Droughts are pervasive hydrometeorological phenomena and global hazards, whose frequency and intensity are expected to increase in the context of climate change. Drought monitoring is of paramount relevance. Here we propose a hybrid model for drought detection that integrates both climatic indices and data-driven models in a hybrid deep learning approach. We exploit time-series of multi-scale Standardized Precipitation Evapotranspiration Index together with precipitation and temperature as inputs. We introduce a dual-branch recurrent neural network with convolutional lateral connections for blending the data. Experimental and ablative results show that the proposed system outperforms both the considered drought index and purely data-driven deep learning models. Our results suggest the potential of hybrid models for drought monitoring and open the door to synergistic systems that learn from data and domain knowledge altogether.

Measuring and monitoring soil organic carbon is critical for agricultural productivity and for addressing critical environmental problems. Soil organic carbon not only enriches nutrition in soil, but also has a gamut of co-benefits such as improving water storage and limiting physical erosion. Despite a litany of work in soil organic carbon estimation, current approaches do not generalize well across soil conditions and management practices. We empirically show that explicit modeling of cause-and-effect relationships among the soil processes improves the out-of-distribution generalizability of prediction models. We provide a comparative analysis of soil organic carbon estimation models where the skeleton is estimated using causal discovery methods. Our framework provide an average improvement of 81% in test mean squared error and 52% in test mean absolute error.

Methane leak detection and remediation are critical for tackling climate change, where methane dispersion simulations play an important role in emission source attribution. As 3D modeling of methane dispersion is often costly and time-consuming, we train a deep-learning-based surrogate model using the Fourier Neural Operator to learn the PDE solver in our study. Our preliminary result shows that our surrogate modeling provides a fast, accurate and cost-effective solution to methane dispersion simulations, thus reducing the cycle time of methane leak detection.

Companies worldwide contribute to climate change, emitting significant amounts of greenhouse gases (GHGs). Yet, most do not report their direct or Scope 1 emissions, resulting in a large data gap in corporate emissions. This study aims to fill this gap by training several decision-tree machine learning models to predict company-level Scope 1 emissions. Our results demonstrate that the Extreme Gradient Boosting and LightGBM models perform best, where the former shows a 19% improvement in prediction error over a benchmark model. Our model is also of reduced complexity and greater computational efficiency; it does not require meta-learners and is trained on a smaller number of features, for which data is more common and accessible compared to prior works. Our features are uniquely chosen based on concepts of environmental pollution in economic theory. Predicting corporate emissions with machine learning can be used as a gap-filling approach, which would allow for better GHG accounting and tracking, thus facilitating corporate decarbonization efforts in the long term. It can also impact representations of a company’s carbon performance and carbon risks, thereby helping to funnel investments towards companies with lower emissions and those making true efforts to decarbonize.

Energy preservation is central to prevent resource depletion, climate change and environment degradation. Investment in raising efficiency of appliances is among the most significant attempts to save energy. Ironically, introduction of many such energy saving appliances increased the total energy consumption instead of reducing it. This effect in literature is attributed to the inherent Jevons paradox (JP) and optimism bias (OB) in consumer behavior. However, the magnitude of these instincts vary among different people. Identification of this magnitude for each household can enable the development of appropriate policies that induce desired energy saving behaviour. Using the RECS 2015 dataset, the paper uses machine learning for each electrical appliance to determine the dependence of their total energy consumption on their energy star rating. This shows that only substitutable appliances register increase in energy demand upon boosted efficiency. Lastly, an index is noted to indicate the varying influence of JP and OB on different households.

Mitigating emissions in line with climate goals requires the rapid integration of low carbon energy sources, such as solar photovoltaics (PV) into the electricity grid. However, the energy produced from solar PV fluctuates due to clouds obscuring the sun's energy. Solar PV yield nowcasting is used to help anticipate peaks and troughs in demand to support grid integration. This paper compares multiple low-resource approaches to nowcasting solar PV yield. To do so, we use a dataset of UK satellite imagery and solar PV energy readings over a 1 to 4-hour time range. Our work investigates the performance of multiple nowcasting models. The paper also estimates the carbon emissions generated and averted by deploying models such as these, and finds that short-term PV forecasting may have a benefit several orders of magnitude greater than its carbon cost and that this benefit holds for small models that could be deployable in low-resource settings around the globe.

The number of standardized policy documents regarding climate policy and their publication frequency is significantly increasing. The documents are long and tedious for manual analysis, especially for policy experts, lawmakers, and citizens who lack access or domain expertise to utilize data analytics tools. Potential consequences of such a situation include reduced citizen governance and involvement in climate policies and an overall surge in analytics costs, rendering less accessibility for the public. In this work, we use a Latent Dirichlet Allocation-based pipeline for the automatic summarization and analysis of 10-years of national energy and climate plans (NECPs) for the period from 2021 to 2030, established by 27 Member States of the European Union. We focus on analyzing policy framing, the language used to describe specific issues, to detect essential nuances in the way governments frame their climate policies and achieve climate goals. The methods leverage topic modeling and clustering for the comparative analysis of policy documents across different countries. It allows for easier integration in potential user-friendly applications for the development of theories and processes of climate policy. This would further lead to better citizen governance and engagement over climate policies and public policy research.

To be effective, good policy interventions often need to be popular among citizens. The success of a given sustainability transition policy can then rest on what citizens think about it. Setting a price on carbon is a timely example. In Canada for example, despite the federal government's estimate that 80\% of households are receiving a cash surplus as a result of this policy, its popularity is evenly split. Previous work in various countries shows that public support is strongly influenced by ideology, with Canadian conservative leadership campaigning on abolishing the policy. Here, we ask what semantic structure underlies carbon tax opposition, with the hope of informing more effective messaging to increase support. To address this question, we use a large dataset of open-ended responses of Canadians elaborating on their support of or opposition to the tax. To capture the underlying semantics, we use the highly expressive structural topic model (STM), a standard implementation of which has been used to study carbon tax opinion in other countries. Unlike previous work, we focus on STM's ability to flexibly allow for topic mixtures. We have fit these models and find topics learned from oppose responses have higher semantic coherence. Focussing the inferred topic mixture weights, we find that oppose responses mix topics using weights that are less heterogeneous, lower-dimensional, and more (and more densely) correlated than those inferred from support responses. As a result, dislodging conservatives' opposition to carbon pricing, if possible at all, may require addressing multiple beliefs in tandem to overcome the putatively stabilizing, cooperative effect of the highly correlated topics raised when justifying that opposition.

Recent years have seen increased interest in modeling future climate trends, especially from the point of view of accurately predicting, understanding and mitigating downstream impacts. For instance, current state-of-the-art process-based agriculture models rely on high-resolution climate data during the growing season for accurate estimation of crop yields. However, high-resolution climate data for future climates is unavailable and needs to be simulated, and that too for multiple possible climate scenarios, which becomes prohibitively expensive via traditional methods. Meanwhile, deep generative models leveraging the expressivity of neural networks have shown immense promise in modeling distributions in high dimensions. Here, we cast the problem of simulation of climate scenarios in a generative modeling framework. Specifically, we leverage the GAN (Generative Adversarial Network) framework for simulating synthetic climate scenarios. We condition the model by quantifying the degree of ``extremeness" of the observed sample, which allows us to sample from different parts of the distribution. We demonstrate the efficacy of the proposed method on the CHIRPS precipitation dataset.

As part of efforts to tackle climate change, grid-scale battery energy storage systems (BESS) play an essential role in facilitating reliable and secure power system operation with variable renewable energy (VRE). BESS can balance time-varying electricity demand and supply in the spot market through energy arbitrage and in the frequency control ancillary services (FCAS) market through service enablement or delivery. Effective algorithms are needed for the optimal participation of BESS in multiple markets. Using deep reinforcement learning (DRL), we present a BESS bidding strategy in the joint spot and contingency FCAS markets, leveraging a transformer-based temporal feature extractor to exploit the temporal trends of volatile energy prices. We validate our strategy on real-world historical energy prices in the Australian National Electricity Market (NEM). We demonstrate that the novel DRL-based bidding strategy significantly outperforms benchmarks. The simulation also reveals that the joint bidding in both the spot and contingency FCAS markets can yield a much higher profit than in individual markets and even profits combined in some jurisdictions of the Australian NEM. Our work provides a viable use case for the BESS, contributing to the power system operation with high penetration of renewables.

As the result of climate change, extreme flood events are becoming more frequent. To better respond to such disasters, and to test and calibrate flood models, we need accurate real-world data on flooding extent. Detection of floods from remote sensed imagery suffers from a widespread problem: clouds block flood scenes in images, leading to degraded and fragmented flood datasets. To address this challenge, we propose a workflow based on U-Net, and a dataset that detects flood in cloud-prone areas by fusing information from the Sentinel-1 and Sentinel-2 satellites. The expected result will be a reliable and detailed catalogue of flood extents and how they change through time, allowing us to better understand flooding in different morphological settings and climates.

Precipitation drives the hydroclimate of Earth and its spatiotemporal changes on a day to day basis have one of the most notable socioeconomic impacts. The success of numerical weather prediction (NWP) is measured by the improvement of forecasts for various physical fields such as temperature and pressure. Large biases however exist in the precipitation predictions. Pure deep learning based approaches lack the advancements acheived by NWP in the past two to three decades. Hybrid methodology using NWP outputs as inputs to the deep learning based refinement tool offer an attractive means taking advantage of both NWP and state of the art deep learning algorithms. Augmenting the output from a well-known NWP model: Coupled Forecast System ver.2 (CFSv2) with deep learning for the first time, we demonstrate a hybrid model capability (DeepNWP) which shows substantial skill improvements for short-range global precipitation at 1-, 2- and 3-days lead time. To achieve this hybridization, we address the sphericity of the global data by using modified DLWP-CS architecture which transforms all the fields to cubed-sphere projection. The dynamical model outputs corresponding to precipitation and surface temperature are ingested to a UNET for predicting the target ground truth precipitation. While the dynamical model CFSv2 shows a bias in the range of +5 to +7 mm/day over land, the multivariate deep learning model reduces it to -1 to +1 mm/day over global land areas. We validate the results by taking examples from Hurricane Katrina in 2005, Hurricane Ivan in 2004, Central European floods in 2010, China floods in 2010, India floods in 2005 and the Myanmar cyclone Nargis in 2008.

Compared to rural areas, urban areas experience higher temperatures for longer periods of time because of the urban heat island (UHI) effect. This increased heat stress leads to greater mortality, increased energy demand, regional changes to precipitation patterns, and increased air pollution. Urban developers can minimize the UHI effect by incorporating features that promote air flow and heat dispersion (e.g., increasing green space). However, understanding which urban features to implement is complex, as local meteorology strongly dictates how the environment responds to changes in urban form. In this proposal we describe a methodology for estimating the causal relationship between changes in urban form and changes in the UHI effect. Changes in urban form and temperature changes are measured using convolutional neural networks, and a causal inference matching approach is proposed to estimate causal relationships. The success of this methodology will enable urban developers to implement city-specific interventions to mitigate the warming planet's impact on cities.

Renewable energy-based electricity systems are seen as a keystone of future decarbonization efforts. However, power system planning does not currently consider the impacts of climate change on renewable energy resources such as solar energy, chiefly due to a paucity of climate-impacted high-resolution solar power data. Existing statistical downscaling (SD) methods that learn to map coarse-resolution versions of historical reanalysis to generate finer resolution outputs are of limited use when applied to future climate model projections due to the domain gap between climate models and reanalysis. In contrast, we present EnhancedSD, a deep learning-based framework for downscaling coarse-scale climate model outputs to high-resolution observational (reanalysis) data. Our proposed approach toward SD allows for future reanalysis projections, which can be pivotal for mitigating climate change’s impacts on power systems planning.

Modeling spatial dependencies in geographic data is of crucial importance for the modeling of our planet. Graph neural networks (GNNs) provide a powerful and scalable solution for modeling continuous spatial data. However, in the absence of further context on the geometric structure of the data, they often rely on Euclidean distances to construct the input graphs. This assumption can be improbable in many real-world settings, where the spatial structure is more complex and explicitly non-Euclidean (e.g., road networks). In this paper, we propose PE-GNN, a new framework that incorporates spatial context and correlation explicitly into the models. Building on recent advances in geospatial auxiliary task learning and semantic spatial embeddings, our proposed method (1) learns a context-aware vector encoding of the geographic coordinates and (2) predicts spatial autocorrelation in the data in parallel with the main task. We show the effectiveness of our approach on two climate-relevant regression tasks: 3d spatial interpolation and air temperature prediction. The code for this study can be accessed via: https://bit.ly/3xDpfyV.

Wildfires are a disastrous phenomenon which cause damage to land, loss of property, air pollution, and even loss of human life. Due to the warmer and drier conditions created by climate change, more severe and uncontrollable wildfires are expected to occur in the coming years. This could lead to a global wildfire crisis and have dire consequences on our planet. Hence, it has become imperative to use technology to help prevent the spread of wildfires. One way to prevent the spread of wildfires before they become too large is to perform early detection i.e, detecting the smoke before the actual fire starts. In this paper, we present our Wildfire Detection and Alert System which use machine learning to detect wildfire smoke with a high degree of accuracy and can send immediate alerts to users. Our technology is currently being used in the USA to monitor data coming in from hundreds of cameras daily. We show that our system has a high true detection rate and a low false detection rate. Our performance evaluation study also shows that on an average our system detects wildfire smoke faster than an actual person.

The continuous increase in global population and the impact of climate change on crop production are expected to affect the food sector significantly. In this context, there is need for timely, large-scale and precise mapping of crops for evidence-based decision making. A key enabler towards this direction are new satellite missions that freely offer big remote sensing data of high spatio-temporal resolution and global coverage. During the previous decade and because of this surge of big Earth observations, deep learning methods have dominated the remote sensing and crop mapping literature. Nevertheless, deep learning models require large amounts of annotated data that are scarce and hard-to-acquire. To address this problem, transfer learning methods can be used to exploit available annotations and enable crop mapping for other regions, crop types and years of inspection. In this work, we have developed and trained a deep learning model for paddy rice detection in South Korea using Sentinel-1 VH time-series. We then fine-tune the model for i) paddy rice detection in France and Spain and ii) barley detection in the Netherlands. Additionally, we propose a modification in the pre-trained weights in order to incorporate extra input features (Sentinel-1 VV). Our approach shows excellent performance when transferring in different areas for the same crop type and rather promising results when transferring in a different area and crop type.

Pyrocumulonimbus (pyroCbs) are storm clouds generated by extreme wildfires. PyroCbs are associated with unpredictable wildfire spread. They can also inject smoke particles and trace gases into the upper troposphere and lower stratosphere. As the climate warms, these previously rare events are becoming more common. This paper presents Pyrocast, a pipeline for pyroCb analysis and forecasting. This paper presents the pipeline's two first components: a pyroCb database and a pyroCb forecasting model. The database brings together geostationary imagery and environmental data for over 148 pyroCb events across North America and Australia between 2018 and 2022. Random Forests, Convolutional Neural Networks (CNNs), and CNNs pretrained with Auto-Encoders were tested to predict the generation of pyroCb for a given fire 6 hours in advance. The best model predicted pyroCb with an AUC of 0.90±0.04.

Drought detection and prediction is challenging due to the slow onset of the event and varying degrees of dependence on numerous physical and socio-economic factors that differentiate droughts from other natural disasters. In this work, we propose DeepXD (Deep learning for Droughts), a deep learning model with 26 physics-informed input features for SPI (Standardised Precipitation Index) forecasting to identify and classify droughts using monthly oceanic indices, global meteorological and vegetation data, location (latitude, longitude) and land cover for the years 1982 to 2018. In our work, we propose extracting features by considering the atmosphere and land moisture and energy budgets and forecasting global droughts on a seasonal and an annual scale at 1, 3, 6, 9, 12 and 24 months lead times. SPI helps us to identify the severity and the duration of the drought to classify them as meteorological, agricultural and hydrological.

The availability of reliable, high-resolution climate and weather data is important to inform long-term decisions on climate adaptation and mitigation and to guide rapid responses to extreme events. Forecasting models are limited by computational costs and therefore often can only make coarse resolution predictions. Statistical downscaling can provide an efficient method of upsampling low-resolution data. In this field, deep learning has been applied successfully, often using image super-resolution methods from computer vision. Despite achieving visually compelling results in some cases, such models often violate conservation laws when predicting physical variables. In order to conserve important physical quantities, we developed a deep downscaling method that guarantees physical constraints are satisfied, by adding a renormalization layer at the end of the neural network. Furthermore, the constrained model also improves the performance according to standard metrics. We show the applicability of our methods across different popular architectures and upsampling factors using ERA5 reanalysis data.

When modeling atmospheric chemistry, concentrations are determined by numerically solving large systems of ordinary differential equations that represent a set of chemical reactions. These solvers can be very computationally intensive, particularly those with the thousands or tens of thousands of chemical species and reactions that make up the most accurate models. We demonstrate the application of a deep learning transformer architecture to emulate an atmospheric chemistry box model, and show that this attention-based model outperforms LSTM and autoencoder baselines while providing interpretable predictions that are more than 2 orders of magnitude faster than a numerical solver. This work is part of a larger study to replace the numerical solver in a 3D global chemical model with a machine learned emulator and achieve significant speedups for global climate simulations.

Restoring ecosystems and reducing deforestation are necessary tools to mitigate the anthropogenic climate crisis. Current measurements of forest carbon stock can be inaccurate, in particular for underrepresented and small-scale forests in the Global South, hindering transparency and accountability in the Monitoring, Reporting, and Verification (MRV) of these ecosystems. There is thus need for high quality datasets to properly validate ML-based solutions. To this end, we present ForestBench, which aims to collect and curate geographically-balanced gold-standard datasets of small-scale forest plots in the Global South, by collecting ground-level measurements and visual drone imagery of individual trees. These equitable validation datasets for ML-based MRV of nature-based solutions shall enable assessing the progress of ML models for estimating above-ground biomass, ground cover, and tree species diversity.

Climate change is increasing the frequency of flooding around the world. As a consequence, there is an increasing demand for effective flood prediction. Machine learning is a promising alternative to hydrodynamic models for flood prediction. However, existing approaches focus on capturing either the spatial or temporal flood patterns using CNNs or RNNs, respectively. In this work, we propose FloodGNN, which is a graph neural network (GNN) for flood prediction. Compared to existing approaches, FloodGNN (i) employs a graph-based model (GNN); (ii) operates on both spatial and temporal dimensions; and (iii) processes the water flow velocities as vector features, instead of scalar features. Experiments show that FloodGNN achieves promising results, outperforming an RNN-based baseline.

Common representations of the stratospheric ozone layer in climate modeling are widely considered only in a highly simplified form that falls short of the current understanding of the stratospheric ozone chemistry. For climate projections, it would be of advantage to include the mutual interactions between stratospheric ozone, temperature, and atmospheric dynamics to accurately represent radiative forcing.Our aim with this paper was to replace the ozone layer in climate models with a machine-learned implicit neural representation that provides a particularly fast, yet accurate and stable, simulation.First, we explored correlations and causalities to create a comprehensive benchmark data set from simulations of the ATLAS chemistry and transport model. Given this data set, we experimented with different variants of multilayer perceptrons suitable for physical problems to learn an implicit neural representation of our latent function. We optimised our model by an extensive Bayesian hyperparameter search. For validation, we coupled this model into the ATLAS chemistry and transport model and benchmarked computation time, accuracy, and stability against the full chemistry model.The resulting implicit neural representation (Neural-SWIFT) allowed us to perform long-term simulations with high accuracy, without significant error accumulation, and a factor of 700 faster than the baseline model. The accuracy of our model surpassed the previous polynomial approach and is able to accurately reproduce the regimes of chemical production and loss as well as seasonality in both hemispheres when compared to the full chemistry model.Neural-SWIFT enables mutual interactions between stratospheric ozone, temperature, and atmospheric dynamics and performs comparably to a full chemistry model, but in a much faster and simpler way, and is thus intended for use in climate models.

Carbon capture and storage (CCS) is one of the most promising technologies for reducing greenhouse gas emissions and relies on numerical reservoir simulations for identifying and monitoring CO2 storage sites. In many commercial settings however, numerical reservoir simulations are too computationally expensive for important downstream application such as optimization or uncertainty quantification. Deep learning-based surrogate models offer the possibility to solve PDEs many orders of magnitudes faster than conventional simulators, but they are difficult to scale to industrial-scale problem settings. Using model-parallel deep learning, we train the largest CO2 surrogate model to date on a 3D simulation grid with two million grid points. To train the 3D simulator, we generate a new training dataset based on a real-world CCS simulation benchmark. Once trained, each simulation with the network is five orders of magnitude faster than a numerical reservoir simulator and 4,500 times cheaper. This paves the way to applications that require thousands of (sequential) simulations, such as optimizing the location of CO2 injection wells to maximize storage capacity and minimize risk of leakage.

There are many benefits from the accurate forecasting of Arctic sea ice, however existing models struggle to reliably predict sea ice concentration at long lead times. Many numerical models exist but can be sensitive to initial conditions, and while recent deep learning-based methods improve overall robustness, they either do not utilize temporal trends or rely on architectures that are not performant at learning long-term sequential dependencies. We propose a method of forecasting sea ice concentration using neural circuit policies, a form of continuous time recurrent neural architecture, which improve the learning of long-term sequential dependencies compared to existing techniques and offer the added benefits of adaptability to irregular sequence intervals and high interpretability.

Climate change is threatening human health in unprecedented orders and many ways. These threats are expected to grow unless effective and evidence-based policies are developed and acted upon to minimize or eliminate them. Attaining such a task requires the highest degree of the flow of knowledge from science into policy. The multidisciplinary, location-specific, and vastness of published science makes it challenging to keep track of novel work in this area, as well as making the traditional knowledge synthesis methods inefficient in infusing science into policy. To this end, we consider developing multiple domain-specific language models (LMs) with different variations from Climate- and Health-related information, which can serve as a foundational step toward capturing available knowledge to enable solving different tasks, such as detecting similarities between climate- and health-related concepts, fact-checking, relation extraction, evidence of health effects to policy text generation, and more. To our knowledge, this is the first work that proposes developing multiple domain-specific language models for the considered domains. We will make the developed models, resources, and codebase available for the researchers.

The precipitation video datasets have distinctive meteorological patterns where a mass of fluid moves in a particular direction across the entire frames, and each local area of the fluid has an individual life cycle from initiation to maturation to decay. This paper proposes a novel transformer-based model for precipitation nowcasting that can extract global and local dynamics within meteorological characteristics. The experimental results show our model achieves state-of-the-art performances on the precipitation nowcasting benchmark.

This proposal aims to develop more accurate federated learning (FL) methods with faster convergence properties and lower communication requirements, specifically for forecasting distributed energy resources (DER) such as renewables, energy storage, and loads in modern, low-carbon power grids. This will be achieved by (i) leveraging recently developed extensions of FL such as hierarchical and iterative clustering to improve performance with non-IID data, (ii) experimenting with different types of FL global models well-suited to time-series data, and (iii) incorporating domain-specific knowledge from power systems to build more general FL frameworks and architectures that can be applied to diverse types of DERs beyond just load forecasting, and with heterogeneous clients.

Despite the importance of quantifying how the spatial patterns of extreme precipitation will change with warming, we lack tools to objectively analyze the storm-scale outputs of modern climate models. To address this gap, we develop an unsupervised machine learning framework to quantify how storm dynamics affect precipitation extremes and their changes without sacrificing spatial information. Over a wide range of precipitation quantiles, we find that the spatial patterns of extreme precipitation changes are dominated by spatial shifts in storm regimes rather than intrinsic changes in how these storm regimes produce precipitation.

Determining changes in global temperature and precipitation that may indicate climate change is complicated by annual variations. One approach for finding potential climate change indicators is to train a model that predicts the year from annual means of global temperatures and precipitations. Such data is available from the CMIP6 ensemble of simulations. Here a two-hidden-layer neural network trained on this data successfully predicts the year. Differences among temperature and precipitation patterns for which the model predicts specific years reveal changes through time. To find these optimal patterns, a new way of interpreting what the neural network has learned is explored. Alopex, a stochastic correlative learning algorithm, is used to find optimal temperature and precipitation maps that best predict a given year. These maps are compared over multiple years to show how temperature and precipitations patterns indicative of each year change over time.

Research has shown that climate change creates warmer temperatures and drier conditions, leading to longer wildfire seasons and increased wildfire risks in the United States. These factors have in turn led to increases in the frequency, extent, and severity of wildfires in recent years. Given the danger posed by wildland fires to people, property, wildlife, and the environment, there is an urgency to provide tools for effective wildfire management. Early detection of wildfires is essential to minimizing potentially catastrophic destruction. In this paper, we present our work on integrating multiple data sources in SmokeyNet, a deep learning model using spatio-temporal information to detect smoke from wildland fires. Camera image data is integrated with weather sensor measurements and processed by SmokeyNet to create a multimodal wildland fire smoke detection system. Our results show that incorporating multimodal data in SmokeyNet improves performance in terms of both F1 and time-to-detection over the baseline with a single data source. With a time-to-detection of only a few minutes, SmokeyNet can serve as an automated early notification system, providing a useful tool in the fight against destructive wildfires.

Aerosol-cloud interactions (ACI) include various effects that result from aerosols entering a cloud, and affecting cloud properties. In general, an increase in aerosol concentration results in smaller droplet sizes which leads to larger, brighter, longer-lasting clouds that reflect more sunlight and cool the Earth. The strength of the effect is however heterogeneous, meaning it depends on the surrounding environment, making ACIs one of the most uncertain effects in our current climate models. In our work, we use causal machine learning to estimate ACI from satellite observations by reframing the problem as a treatment (aerosol) and outcome (change in droplet radius). We predict the causal effect of aerosols on clouds with bounds of uncertainty depending on the unknown factors that may be influencing the impact of aerosols. Of the three climate models evaluated, we find that only one plausibly recreates the trend, lending more credence to its estimate cooling due to ACI.

Numerous polluted groundwater sites across the globe require an active remediation strategy for the restoration of natural environmental conditions and local ecosystem. The Engineered Nanoparticles (ENPs) has emerged as an efficient reactive agent for the in-situ degradation of groundwater contaminants. While the performance of these ENPs has been highly promising on the laboratory scale, their application in a real field case conditions is still limited. The optimized injection of the ENPs in the contaminated aquifer and its subsequent monitoring are hindered by the complex transport and retention mechanisms of ENPs. Therefore, a predictive tool for understanding the transport and retention behavior of ENPs becomes highly important. The existing tools in the literature are dominated with numerical simulators, which have limited flexibility and accuracy in the presence of sparse dataset. In this work, a dynamic weights enabled Physics-Informed Neural network (dw-PINN) framework is applied to model the nano-particle´s behavior within an aquifer. The result from the forward model demonstrates the effective capability of dw-PINN in accurately predicting the ENPs mobility. The model verification step shows that the mean squared error of the predicted ENPs concentration using dw-PINN converges to a minimum value of 1.3e-5. In the subsequent step, the result from the inverse model estimates the governing parameters of ENPs mobility with reasonable accuracy. The research work demonstrates the tool´s capability in providing predictive insights for the development of an efficient groundwater remediation strategy.

Forecasting the state of vegetation in response to climate and weather events is a major challenge. Its implementation will prove crucial in predicting crop yield, forest damage, or more generally the impact on ecosystems services relevant for socio-economic functioning, which if absent can lead to humanitarian disasters. Vegetation status depends on weather and environmental conditions that modulate complex ecological processes taking place at several timescales. Interactions between vegetation and different environmental drivers express responses at instantaneous but also time-lagged effects, often showing an emerging spatial context at landscape and regional scales. We formulate the land surface forecasting task as a strongly guided video prediction task where the objective is to forecast the vegetation developing at very fine resolution using topography and weather variables to guide the prediction. We use a Convolutional LSTM (ConvLSTM) architecture to address this task and predict changes in the vegetation state in Africa using Sentinel-2 satellite NDVI, having ERA5 weather reanalysis, SMAP satellite measurements, and topography (DEM of SRTMv4.1) as variables to guide the prediction. Ours results highlight how ConvLSTM models can not only forecast the seasonal evolution of NDVI at high resolution, but also the differential impacts of weather anomalies over the baselines. The model is able to predict different vegetation types, even those with very high NDVI variability during target length.

Forecasting global precipitation patterns and, in particular, extreme precipitation events is of critical importance to preparing for and adapting to climate change. Making accurate high-resolution precipitation forecasts using traditional physical models remains a major challenge in operational weather forecasting as they incur substantial computational costs and struggle to achieve sufficient forecast skill. Recently, deep-learning-based models have shown great promise in closing the gap with numerical weather prediction (NWP) models in terms of precipitation forecast skill, opening up exciting new avenues for precipitation modeling. However, it is challenging for these deep learning models to fully resolve the fine-scale structures of precipitation phenomena and adequately characterize the extremes of the long-tailed precipitation distribution. In this work, we present several improvements to the architecture and training process of a current state-of-the art deep learning precipitation model (FourCastNet) using a novel generative adversarial network (GAN) to better capture fine scales and extremes. Our improvements achieve superior performance in capturing the extreme percentiles of global precipitation, while comparable to state-of-the-art NWP models in terms of forecast skill at 1--2 day lead times. Together, these improvements set a new state-of-the-art in global precipitation forecasting.

Transportation plays a major role in global CO2 emission levels, a factor that directly connects with climate change. Roadway interventions that reduce CO2 emission levels have thus become a timely requirement. An integral need in assessing the impact of such roadway interventions is access to industry-standard programmatic and instantaneous emission models with various emission conditions such as fuel types, vehicle types, cities of interest, etc. However, currently, there is a lack of well-calibrated emission models with all these properties. Addressing these limitations, this paper presents 1100 programmatic and instantaneous vehicular CO2 emission models with varying fuel types, vehicle types, road grades, vehicle ages, and cities of interest. We hope the presented emission models will facilitate future research in tackling transportation-related climate impact. The released version of the emission models can be found here.

Visual Question Answering (VQA) is a multi-modal task that involves answering questions from an input image, semantically understanding the contents of the image and answering it in natural language. Using VQA for disaster management is an important line of research due to the scope of problems that are answered by the VQA system. However, the main challenge is the delay caused by the generation of labels in the assessment of the affected areas. To tackle this, we deployed pre-trained CLIP model, which is trained on visual-image pairs. however, we empirically see that the model has poor zero-shot performance. Thus, we instead use pre-trained embeddings of text and image from this model for our supervised training and surpass previous state-of-the-art results on the FloodNet dataset. We expand this to a continual setting, which is a more real-life scenario. We tackle the problem of catastrophic forgetting using various experience replay methods.

Responsible energy consumption plays a key role in reducing carbon footprint and CO2 emissions to tackle climate change. A better understanding of the residential consumption behavior using smart meter data is at the heart of the mission, which can inform residential demand flexibility, appliance scheduling, and home energy management. However, access to high-quality residential load data is still limited due to the cost-intensive data collection process and privacy concerns of data sharing. In this paper, we develop a Generative Adversarial Network (GAN)-based method to model the complex and diverse residential load patterns and generate synthetic yet realistic load data. We adopt a generation-focused weight selection method to select model weights to address the mode collapse problem and generate diverse load patterns. We evaluate our method using real-world data and demonstrate that it outperforms three representative state-of-the-art benchmark models in better preserving the sequence level temporal dependencies and aggregated level distributions of load patterns.

Methane (CH_4) is the chief contributor to global climate change and its mitigation is targeted by the EU, US and jurisdictions worldwide~\cite{methane-reduction}. Recent studies have shown that imagery from the multi-spectral instrument on Sentinel-2 satellites is capable of detecting and estimating large methane emissions. However, most of the current methods rely on temporal relations between a ratio of shortwave-infrared spectra and assume relatively constant ground conditions, and availability of ground information on when there was no methane emission on site. To address such limitations we propose a guided query-based transformer neural network architecture, that will detect and quantify methane emissions without dependence on temporal information. The guided query aspect of our architecture is driven by a Sentinel Enhanced Matched Filter (SEMF) approach, also discussed in this paper. Our network uses all 12 spectral channels of Sentinel-2 imagery to estimate ground terrain and detect methane emissions. No dependence on temporal data makes it more robust to changing ground and terrain conditions and more computationally efficient as it reduces the need to process historical time-series imagery to compute a single date emissions analysis.

Climate-related disclosure is increasing in importance as companies and stakeholders alike aim to reduce their environmental impact and exposure to climate-induced risk. Companies primarily disclose this information in annual or other lengthy documents where climate information is not the sole focus. To assess the quality of a company's climate-related disclosure, these documents, often hundreds of pages long, must be reviewed manually by climate experts. We propose a more efficient approach to assessing climate-related financial information. We construct a model leveraging TF-IDF, sentence transformers and multi-label k nearest neighbors (kNN). The developed model is capable of assessing alignment of climate disclosures at scale, with a level of granularity and transparency that will support decision-making in the financial markets with relevant climate information. In this paper, we discuss the data that enabled this project, the methodology, and how the resulting model can drive climate impact.

This paper develops a novel data set for three key resources use; namely, food, water, and energy, for 9000 cities globally. The data set is then utilized to develop a clustering approach as a starting point towards a global classification model. This novel clustering approach aims to contribute to developing an inclusive view of resource efficiency for all urban centers globally. The proposed clustering algorithm is comprised of three steps: first, outlier detection to address specific city characteristics, then a Variational Autoencoder (VAE), and finally, Agglomerative Clustering (AC) to improve the classification results. Our results show that this approach is more robust and yields better results in creating delimited clusters with high Calinski-Harabasz Index scores and Silhouette Coefficient than other baseline clustering methods.

How can we learn from all available data when training machine-learnt climate models, without incurring any extra cost at simulation time? Typically, the training data comprises coarse-grained high-resolution data. But only keeping this coarse-grained data means the rest of the high-resolution data is thrown out. We use a transfer learning approach, which can be applied to a range of machine learning models, to leverage all the high-resolution data. We use three chaotic systems to show it stabilises training, gives improved generalisation performance and results in better forecasting skill. Our code is at https://github.com/raghul-parthipan/dontwastedata

Several wildfire danger systems have emerged from decades of research. One such system is the National Fire-Danger Rating System (NFDRS), which is used widely across the United States and is a key predictor in the Global ECMWF Fire Forecasting (GEFF) model. The NFDRS is composed of over 100 equations relating wildfire risk to weather conditions, climate and land cover characteristics, and fuel. These equations and the corresponding 130+ parameters were developed via field and lab experiments. These parameters, which are fixed in the standard NFDRS and GEFF implementations, may not be the most appropriate for a climate-changing world. In order to adjust the NFDRS parameters to current climate conditions and specific geographical locations, we recast NFDRS in PyTorch to create a new deep learning-based Fire Index Risk Optimizer (FIRO). FIRO predicts the ignition component, or the probability a wildfire would require suppression in the presence of a firebrand, and calibrates the uncertain parameters for a specific region and climate conditions by training on observed fires. Given the rare occurrence of wildfires, we employed the extremal dependency index (EDI) as the loss function. Using ERA5 reanalysis and MODIS burned area data, we trained FIRO models for California, Texas, Italy, and Madagascar. Across these four geographies, the average EDI improvement was 175% above the standard NFDRS implementation

The increasing intensity and frequency of floods is one of the many consequences of our changing climate. In this work, we explore ML techniques that improve the flood detection module of an operational early flood warning system. Our method exploits an unlabelled dataset of paired multi-spectral and Synthetic Aperture Radar (SAR) imagery to reduce the labeling requirements of a purely supervised learning method. Past attempts have used such unlabelled data by creating weak labels out of them, but end up learning the label mistakes in those weak labels. Motivated by knowledge distillation and semi supervised learning, we explore the use of a teacher to train a student with the help of a small hand labeled dataset and a large unlabelled dataset. Unlike the conventional self distillation setup, we propose a cross modal distillation framework that transfers supervision from a teacher trained on richer modality (multi-spectral images) to a student model trained on SAR imagery. The trained models are then tested on the Sen1Floods11 dataset. Our model outperforms the Sen1Floods11 SAR baselines by an absolute margin of 4.15% mean Intersection-over-Union (mIoU) on the test split.

Decadal climate predictions are key to inform adaptation strategies in a warming climate. Coupled climate models used for decadal predictions are, however, imperfect representations of the climate system leading to forecast biases. Biases can also result from a poor model initialization that, when combined with forecast drift, can produce errors depending non-linearly on lead time. We propose a deep learning-based bias correction approach for the post-processing of gridded forecasts to enhance the accuracy of decadal predictions.

Measuring and attributing greenhouse gas (GHG) emissions remains a challenging problem as the world strives towards meeting emissions reductions targets. As a significant portion of total global emissions, the road transportation sector represents an enormous challenge for estimating and tracking emissions at a global scale. To meet this challenge, we have developed a hybrid approach for estimating road transportation emissions that combines the strengths of machine learning and satellite imagery with localized emissions factors data to create an accurate, globally scalable, and easily configurable GHG monitoring framework.

In forest carbon credit, the concept of dynamic (or ex-post) baseline has been discussed to overcome the criticism of junk carbon credit, while an ex-ante baseline is still necessary in terms of project finance and risk assessment. We propose a Bayesian state-space SCM, which integrates both ex-ante and ex-post baseline estimation in a time-series causal inference framework. We apply the proposed model to a REDD+ project in Brazil, and show that it might have had a small, positive effect but had been over-credited and that the 90% predictive interval of the ex-ante baseline included the ex-post baseline, implying our ex-ante estimation can work effectively.

Weather extremes are a major societal and economic hazard, claiming thousands of lives and causing billions of dollars in damage every year. Under climate change, their impact and intensity are expected to worsen significantly. Unfortunately, general circulation models (GCMs), which are currently the primary tool for climate projections, cannot characterize weather extremes accurately. To address this, we present a multi-resolution deep-learning framework that, firstly, corrects a GCM's biases by matching low-order and tail statistics of its output with observations at coarse scales; and secondly, increases the level of detail of the debiased GCM output by reconstructing the finer scales as a function of the coarse scales. We use the proposed framework to generate statistically realistic realizations of the climate over Western Europe from a simple GCM corrected using observational atmospheric reanalysis. We also discuss implications for probabilistic risk assessment of natural disasters in a changing climate.

In contrast to the rapid digitalization of several industries, agriculture suffers from low adoption of climate-smart farming tools. Even though AI-driven digital agriculture can offer high-performing predictive functionalities, they lack tangible quantitative evidence on their benefits to the farmers. Field experiments can derive such evidence, but are often costly and time consuming. To this end, we propose an observational causal inference framework for the empirical evaluation of the impact of digital tools on target farm performance indicators. This way, we can increase farmers' trust via enhancing the transparency of the digital agriculture market, and in turn accelerate the adoption of technologies that aim to increase productivity and secure a sustainable and resilient agriculture against a changing climate. As a case study, we perform an empirical evaluation of a recommendation system for optimal cotton sowing, which was used by a farmers' cooperative during the growing season of 2021. We leverage agricultural knowledge to develop the causal graph of the farm system, we use the back-door criterion to identify the impact of recommendations on the yield and subsequently we estimate it using several methods on observational data. The results showed that a field sown according to our recommendations enjoyed a significant increase in yield 12% to 17%.

Data-driven models, such as FourCastNet (FCN), have shown exemplary performance in high-resolution global weather forecasting. This performance, however, is based on supervision on mesh-gridded weather data without the utilization of raw climate observational data, the gold standard ground truth. In this work we develop a methodology to correct, remap, and fine-tune gridded uniform forecasts of FCN so it can be directly compared against observational ground truth, which is sparse and non-uniform in space and time. This is akin to bias-correction and post-processing of numerical weather prediction (NWP), a routine operation at meteorological and weather forecasting centers across the globe. The Adaptive Fourier Neural Operator (AFNO) architecture is used as the backbone to learn continuous representations of the atmosphere. The spatially and temporally non-uniform output is evaluated by the non-uniform discrete inverse Fourier transform (NUIDFT) given the output query locations. We call this network the Deep-Learning-Corrector-Remapper (DLCR). The improvement in DLCR’s performance against the gold standard ground truth over the baseline’s performance shows its potential to correct, remap, and fine-tune the mesh-gridded forecasts under the supervision of observations.

Subseasonal forecasting — predicting temperature and precipitation 2 to 6 weeks ahead — is critical for effective water allocation, wildfire management, and drought and flood mitigation. Recent international research efforts have advanced the subseasonal capabilities of operational dynamical models, yet temperature and precipitation prediction skills remains poor, partly due to stubborn errors in representing atmospheric dynamics and physics inside dynamical models. To counter these errors, we introduce an adaptive bias correction (ABC) method that combines state-of-the-art dynamical forecasts with observations using machine learning. When applied to the leading subseasonal model from the European Centre for Medium-Range Weather Forecasts (ECMWF), ABC improves temperature forecasting skill by 60-90% and precipitation forecasting skill by 40-69% in the contiguous U.S. We couple these performance improvements with a practical workflow, based on Cohort Shapley, for explaining ABC skill gains and identifying higher-skill windows of opportunity based on specific climate conditions.

Uncertainty quantification of weather forecasts is a necessity for reliably planning for and responding to extreme weather events in a warming world. This motivates the need for well-calibrated ensembles in probabilistic weather forecasting. We present initial results for the calibration of large-scale deep neural weather models for data-driven probabilistic weather forecasting. By explicitly accounting for uncertainties about the forecast's initial condition and model parameters, we generate ensemble forecasts that show promising results on standard diagnostics for probabilistic forecasts. Specifically, we are approaching the Integrated Forecasting System (IFS), the gold standard on probabilistic weather forecasting, on: (i) the spread-error agreement; and (ii) the Continuous Ranked Probability Score (CRPS). Our approach scales to state-of-the-art data-driven weather models, enabling cheap post-hoc calibration of pretrained models with tens of millions of parameters and paving the way towards the next generation of well-calibrated data-driven weather models.

To fight climate change and accommodate the increasing population, global crop production has to be strengthened. To achieve the "sustainable intensification" of agriculture, transforming it from carbon emitter to carbon sink is a priority, and understanding the environmental impact of agricultural management practices is a fundamental prerequisite to that. At the same time, the global agricultural landscape is deeply heterogeneous, with differences in climate, soil, and land use inducing variations in how agricultural systems respond to farmer actions. The "personalization" of sustainable agriculture with the provision of locally adapted management advice is thus a necessary condition for the efficient uplift of green metrics, and an integral development in imminent policies. Here, we formulate personalized sustainable agriculture as a Conditional Average Treatment Effect estimation task and use Causal Machine Learning for tackling it. Leveraging climate data, land use information and employing Double Machine Learning, we estimate the heterogeneous effect of sustainable practices on the field-level Soil Organic Carbon content in Lithuania. We thus provide a data-driven perspective for targeting sustainable practices and effectively expanding the global carbon sink.

Waves are a more consistent form of clean energy than wind and solar and the latest Wave Energy Converters (WEC) platforms like CETO 6 have evolved into complex multi-generator designs with a high energy capture potential for financial viability. Multi-Agent Reinforcement Learning (MARL) controller can handle these complexities and control the WEC optimally unlike the default engineering controllers like Spring Damper which suffer from lower energy capture and mechanical stress from the spinning yaw motion. In this paper, we look beyond the normal hyper-parameter and MARL agent tuning and explored the most suitable architecture for the neural network function approximators for the policy and critic networks of MARL which act as its brain. We found that unlike the commonly used fully connected network (FCN) for MARL, the sequential models like transformers and LSTMs can model the WEC system dynamics better. Our novel transformer architecture, Skip Transformer-XL (STrXL), with several gated residual connections in and around the transformer block performed better than the state-of-the-art with faster training convergence. STrXL boosts energy efficiency by an average of 25% to 28% over the existing spring damper (SD) controller for waves at different angles and almost eliminated the mechanical stress from the rotational yaw motion, saving costly maintenance on open seas, and thus reducing the Levelized Cost of wave energy (LCOE). Demo: https://tinyurl.com/4s4mmb9v

Almonds are one of the most lucrative products of California, but are also among the most sensitive to climate change. In order to better understand the relationship between climatic factors and almond yield, an automated machine learning framework is used to build a collection of machine learning models. The prediction skill is assessed using historical records. Future projections are derived using 17 downscaled climate outputs. The ensemble mean projection displays almond yield changes under two different climate scenarios, along with two technology development scenarios, where the role of technology development is highlighted. The mean projections and distributions provide insightful results to stakeholders and can be utilized by policymakers for climate adaptation.

The global carbon cycle is a key process to understand how our climate is changing. However, monitoring the dynamics is difficult because a high-resolution robust measurement of key state parameters including the aboveground carbon biomass (AGB) is required. We use deep neural network to generate a wall-to-wall map of AGB within the Continental USA (CONUS) with 30-meter spatial resolution for the year 2021. We combine radar and optical hyperspectral imagery, with a physical climate parameter of solar-induced chlorophyll fluorescence (SIF)-based gross primary productivity (GPP). Validation results show that a masked variation of UNet has the lowest validation RMSE of 37.93 ± 1.36 Mg C/ha, as compared to 52.30 ± 0.03 Mg C/ha for random forest algorithm. Furthermore, models that learn from SIF-based GPP in addition to radar and optical imagery reduce validation RMSE by almost 10% and the standard deviation by 40%. Finally, we apply our model to measure losses in AGB from the recent 2021 Caldor wildfire in California, and validate our analysis with Sentinel-based burn index.

Land cover classification (LCC), and monitoring how land use changes over time, is an important process in climate change mitigation and adaptation. Existing approaches that use machine learning with Earth observation data for LCC rely on fully-annotated and segmented datasets. Creating these datasets requires a large amount of effort, and a lack of suitable datasets has become an obstacle in scaling the use of LCC. In this study, we propose Scene-to-Patch models: an alternative LCC approach utilising Multiple Instance Learning (MIL) that requires only high-level scene labels. This enables much faster development of new datasets whilst still providing segmentation through patch-level predictions, ultimately increasing the accessibility of using LCC for different scenarios. On the DeepGlobe-LCC dataset, our approach outperforms a non-MIL baseline on both scene- and patch-level performance. This work provides the foundation for expanding the use of LCC in climate change mitigation methods for technology, government, and academia.

A major transformation to mitigate climate change implies a rapid decarbonisation of the energy system and thus increasing the use of renewable energy sources, such as wind power. However, renewable resources are strongly dependent on local and large-scale weather conditions, which might be influenced by climate change. Thus, weather-related risk assessments are essential for the energy sector, in particular, for power system management decisions for which forecasts of climatic conditions from several weeks to months (i.e. sub-seasonal scales) are of key importance. Here, we propose a data-driven approach to predict wind speed at longer lead-times that can benefit the energy sector. Furthermore, we aim to explore the potential of machine learning algorithms, particularly deep learning methods, to predict periods of low wind speed conditions that have a strong impact on the energy sector.

Climate change-driven weather disasters are rapidly increasing in both frequency and magnitude. Floods are the most damaging of these disasters, with approximately 1.46 billion people exposed to inundation depths of over 0.15m, a significant life and livelihood risk. Accurate knowledge of flood-extent for ongoing and historical events facilitates climate adaptation in flood-prone communities by enabling near real-time disaster monitoring to support planning, response, and relief during these extreme events. Satellite observations can be used to derive flood-extent maps directly; however, these observations are impeded by cloud and canopy cover, and can be very infrequent and hence miss the flood completely. In contrast, physically-based inundation models can produce spatially complete event maps but suffer from high uncertainty if not frequently calibrated with expensive land and infrastructure surveys. In this study, we propose a deep learning approach to reproduce satellite-observed fractional flood-extent maps given dynamic state variables from hydrologic models, fusing information contained within the states with direct observations from satellites. Our model has an hourly temporal resolution, contains no cloud-gaps, and generalizes to watersheds across the continental United States with a 6% error on held-out areas that never flooded before. We further demonstrate through a case study in Houston, Texas that our model can distinguish tropical cyclones that caused flooding from those that did not within two days of landfall, thereby providing a reliable source for flood-extent maps that can be used by disaster monitoring services.

Climate change is expected to lead to higher rates of forest mortality. Forest mortality is a complex phenomenon driven by the interaction of multiple climatic variables at multiple temporal scales, further modulated by the current state of the forest (e.g. age, stem diameter, and leaf area index). Identifying the compound climate drivers of forest mortality would greatly improve understanding and projections of future forest mortality risk. Observation data are, however, limited in accuracy and sample size, particularly in regard to forest state variables and mortality events. In contrast, simulations with state-of-the-art forest models enable the exploration of novel machine learning techniques for associating forest mortality with driving climate conditions. Here we simulate 160,000 years of beech forest dynamics with the forest model FORMIND. We then apply β-VAE to learn disentangled latent representations of weather conditions and identify those that are most likely to cause high forest mortality. The learned model successfully identifies three characteristic climate representations that can be interpreted as different compound drivers of forest mortality.

Climate change is expected to aggravate wildfire activity through the exacerbation of fire weather. Improving our capabilities to anticipate wildfires on a global scale is of uttermost importance for mitigating their negative effects.In this work, we use deep learning to forecast the presence of global burned areas on a sub-seasonal scale. We present an open-access global analysis-ready datacube, which contains a variety of variables related to the seasonal and sub-seasonal fire drivers (climate, vegetation, oceanic indices, human-related variables), as well as the historical burned areas and wildfire emissions for 2000-2021. We train a deep learning model, which treats global wildfire forecasting as an image segmentation task and skillfully predicts the presence of burned areas 8, 16, 32 and 64 days ahead of time. Our work paves the way towards improved anticipation of global wildfire patterns.

Plant diseases are costly and threaten agricultural production and food security worldwide. Climate change is increasing the frequency and severity of plant diseases and pests. Therefore, detection and early remediation can have a significant impact, especially in developing countries. However, AI solutions are yet far from being in production.

The current process for plant disease diagnostic consists of manual identification and scoring by humans, which is time-consuming, low-supply, and expensive. Although computer vision models have shown promise for efficient and automated plant disease identification, there are limitations for real-world applications: a notable variation in visual symptoms of a single disease, different light and weather conditions, and the complexity of the models. In this work, we study the performance of efficient classification models and training "tricks" to solve this problem. Our analysis represents a plausible solution for these ecological disasters and might help to assist producers worldwide. More information available at: https://github.com/mv-lab/mlplants

Contrails, short for condensation trails, are line-shaped ice clouds produced by aircraft engine exhaust when they fly through the cold and humid air. They generate a greenhouse effect by absorbing or directing back to Earth approximately 33% of emitted outgoing longwave radiation. They account for over half of the climate change resulting from aviation activities. Avoiding contrails and adjusting flight routes could be an inexpensive and effective way to reduce their impact. An accurate, automated, and reliable detection algorithm is required to develop and evaluate contrail avoidance strategies. Advancement in contrail detection has been severely limited due to several factors, primarily due to a lack of quality-labelled data. Recently, McCloskey et al. proposed a large human-labelled Landsat-8. Each contrail is carefully labelled with various inputs in various scenes of Landsat-8 satellite imagery. In this work, we benchmark several popular segmentation models with combinations of different loss functions and encoder backbones. This work is the first to apply state-of-the-art segmentation techniques to detect contrails in low-orbit satellite imagery. Our work can also be used as an open benchmark for contrail segmentation.

Natural disasters such as flood, wildfire, drought, and severe storms wreak havoc throughout the world, causing billions of dollars in damages, and uprooting communities, ecosystems, and economies. Unfortunately, flooding events are on the rise due to climate change and sea level rise. The ability to detect and quantify them can help us minimize their adverse impacts on the economy and human lives. Using satellites to study flood is advantageous since physical access to flooded areas is limited and deploying instruments in potential flood zones can be dangerous. We are proposing a hands-on tutorial to highlight the use of satellite imagery and computer vision to study natural disasters. Specifically, we aim to demonstrate the development and deployment of a flood detection model using Sentinel-1 satellite data. The tutorial will cover relevant fundamental concepts as well as the full development workflow of a deep learning-based application. We will include important considerations such as common pitfalls, data scarcity, augmentation, transfer learning, fine-tuning, and details of each step in the workflow. Importantly, the tutorial will also include a case study on how the application was used by authorities in response to a flood event. We believe this tutorial will enable machine learning practitioners of all levels to develop new technologies that tackle the risks posed by climate change. We expect to deliver the below learning outcomes:•Develop various deep learning-based computer vision solutions using hardware-accelerated open-source tools that are optimized for real-time deployment•Create an optimized pipeline for the machine learning development workflow•Understand different performance metrics for model evaluation that are relevant for real world datasets and data imbalances•Understand the public sector’s efforts to support climate action initiatives and point out where the audience can contribute

Climate change has led to a rapid increase in the occurrence of extreme weather events globally, including floods, droughts, and wildfires. In the longer term, some regions will experience aridification while others will risk sinking due to rising sea levels. Typically, such predictions are done via weather and climate models that simulate the physical interactions between the atmospheric, oceanic, and land surface processes that operate at different scales. Due to the inherent complexity, these climate models can be inaccurate or computationally expensive to run, especially for detecting climate extremes at high spatiotemporal resolutions. In this tutorial, we aim to introduce the participants to machine learning approaches for addressing two fundamental challenges. We will walk the participants through a hands-on tutorial for predicting climate extremes relating to temperature and precipitation in 2 setups: (1) temporal forecasting: the goal is to predict climate variables into the future (both direct single step approaches and iterative approaches that roll out the model for several timesteps), and (2) spatial downscaling: the goal is to learn a mapping that transforms low-resolution outputs of climate models into high-resolution regional forecasts. Through introductory presentations and colab notebooks, we aim to expose the participants to (a) APIs for accessing and navigating popular repositories that host global climate data, such as the Copernicus data store, (b) identifying relevant datasets, including auxiliary data (e.g., other climate variables such as geopotential), (c) scripts for downloading and preprocessing relevant datasets, (d) algorithms for training machine learning models, (d) metrics for evaluating model performance, and (e) visualization tools for both the dataset and predicted outputs. The coding notebooks will be in Python. No prior knowledge of climate science is required.

Accurate, reliable, and efficient means of forecasting global weather patterns are of paramount importance to our ability to mitigate and adapt to climate change. Currently, real-time weather forecasting requires repeated numerical simulation and data assimilation cycles on dedicated supercomputers, which restricts the ability to make reliable, high-resolution forecasts to a handful of organizations. However, recent advances in deep learning, specifically the FourCastNet model, have shown that data-driven approaches can forecast important atmospheric variables with excellent skill and comparable accuracy to standard numerical methods, but at orders-of-magnitude lower computational and energy cost during inference, enabling larger ensembles for better probabilistic forecasts. In this tutorial, we demonstrate various applications of FourCastNet for high-resolution global weather forecasting, with examples including real-time forecasts, uncertainty quantification for extreme events, and adaptation to specific variables or localized regions of interest. The tutorial will provide examples that will demonstrate the general workflow for formatting and working with global atmospheric data, running autoregressive inference to obtain daily global forecasts, saving/visualizing model predictions of atmospheric events such as hurricanes and atmospheric rivers, and computing quantitative evaluation metrics for weather models. The exercises will primarily use PyTorch and do not require detailed understanding of the climate and weather system. With this tutorial, we hope to equip attendees with basic knowledge about building deep learning-based weather model surrogates and obtaining forecasts of crucial atmospheric variables using these models.

High resolution and accurate Land Use and Land Cover mapping (LULC) datasets are increasingly important and can be widely used in monitoring climate change impacts in agriculture, deforestation, and the carbon cycle. These datasets represent physical classifications of land types and spatial information over the surface of the Earth. These LULC datasets can be leveraged in a plethora of research topics and industries to mitigate and adapt to environmental changes. High resolution urban classifications can be used to better monitor and estimate building albedo and urban heat island impacts, and accurate representation of forest and vegetation can even be leveraged to better monitor the carbon cycle and climate change through improve land surface modelling. The advent of machine learning (ML) based CV techniques over the past decade provide a viable option to automate LULC mapping. One impediment to this has been the lack of large ML datasets. Large vector datasets for LULC are available, but can’t be used directly by ML practitioners due to a knowledge gap in transforming the input to a dataset of paired satellite images and segmentation masks. We demonstrate a novel end-to-end pipeline for LULC dataset creation which takes vector land cover data and provides a training-ready dataset. We will use Sentinel-2 satellite imagery and the European Urban Atlas LULC data. The pipeline manages everything from downloading satellite data, to creating and storing encoded segmentation masks and automating data checks. We then use the resulting dataset to train a semantic segmentation model. The aim of the pipeline is to provide a way for users to create their own custom datasets using various combinations of multispectral satellite and vector data. In addition to presenting the pipeline, we aim to provide an introduction to multispectral imagery, challenges in using it for ML, and give a brief perspective of the future work in ML for Earth observation.

Alaska and the larger Arctic region are in much greater need of decarbonization than the rest of the globe as a result of the accelerated consequences of climate change over the past ten years. Heating for homes and businesses accounts for over 75% of the energy used in the Arctic region. However, the lack of thorough and precise heating load estimations in these regions poses a significant obstacle to the transition to renewable energy. In order to accurately measure the massive heating demands in Alaska, this research pioneers a geospatial-first methodology that integrates remote sensing and machine learning techniques. Building characteristics such as height, size, year of construction, thawing degree days, and freezing degree days are extracted using open-source geospatial information in Google Earth Engine (GEE). These variables coupled with heating load forecasts from the AK Warm simulation program are used to train models that forecast heating loads on Alaska’s Railbelt utility grid. Our research greatly advances geospatial capability in this area and considerably informs the decarbonization activities currently in progress in Alaska.

We present an approach to integrate the domain knowledge of the electric power grid operations into reinforcement learning (RL) frameworks for effectively learning RL agents to prevent cascading failures. A curriculum-based approach with reward tuning is incorporated into the training procedure by modifying the environment using the network physics. Our procedure is tested on an actor-critic-based agent on the IEEE 14-bus test system using the RL environment developed by RTE, the French transmission system operator (TSO). We observed that naively training the RL agent without the curriculum approach failed to prevent cascading for most test scenarios, while the curriculum based RL agents succeeded in most test scenarios, illustrating the importance of properly integrating domain knowledge of physical systems for real-world RL applications.