Poster
in
Workshop: Tackling Climate Change with Machine Learning

MS-nowcasting: Operational Precipitation Nowcasting with Convolutional LSTMs at Microsoft Weather

Sylwester Klocek · Haiyu Dong · Panashe Kanengoni · Najeeb Kazmi · Pete Luferenko · Zhongjian Lv · Shikhar Sharma · Jonathan Weyn · Siqi Xiang

Abstract:

We present the encoder-forecaster convolutional long short-term memory (LSTM) deep-learning model that powers Microsoft Weather's operational precipitation nowcasting product. This model takes as input a sequence of weather radar mosaics and deterministically predicts future radar reflectivity at lead times up to 6 hours. By stacking a large input receptive field along the feature dimension and conditioning the model's forecaster with predictions from the physics-based High Resolution Rapid Refresh (HRRR) model, we are able to outperform optical flow and HRRR baselines by 20-25% on multiple metrics averaged over all lead times.