Explainable Deep Learning for Probabilistic Nowcasting of Radar Reflectivity in Tornadic Storms
Abstract
Tornadoes pose substantial risk to human life and property in the United States, causing more than 50 fatalities and $100 million of property damage on average annually. When tornadoes are likely, weather radar provides critical information for forecasters by providing information on storm morphology, storm motion, and intensity trends. Additional tools such as satellite and numerical weather prediction model runs can provide useful short-term information for understanding changes in storm characteristics. This work demonstrates a U-Net deep-learning system for nowcasting the evolution of radar reflectivity following tornadogenesis, which can provide value to forecasters by synthesizing large amounts of input data (e.g., radar imagery, near-storm environment data) and generating predictions of radar reflectivity from its inputs. Inputs to the model are radar imagery from the Multi-Radar Multi-Sensor (MRMS) dataset and near-storm environment data from the High-Resolution Rapid Refresh (HRRR) numerical weather prediction model. The U-Net is trained on a dataset of tornadic storms to produce 30 minutes of probabilistic predictions of radar reflectivity following tornadogenesis, with probabilistic predictions obtained by predicting parameters of the SinhArcSinh, or SHASH, distribution. The model produces physically realistic predictions of radar evolution, achieves comparable skill to next-hour forecasts from the HRRR, demonstrates reasonable probabilistic calibration and is accompanied by a variety of explainability methods to improve understanding by end users. Additionally, predictions from the model can be obtained much more quickly than those from a numerical weather prediction model. With further development, this model could be extended to nowcast radar reflectivity evolution in an operational setting.
Related
- Advancing Near-Term Water Quality Forecasting with Explainable Machine Learning and Probabilistic Uncertainty Quantification
- Diffusion-Based Refinement for Kilometer-Scale Probabilistic Precipitation Nowcasting
- Physics‐Informed Fusion Learning for Nowcasting Fault Detection of Hydrological Multi Sensor Network
- A multi‐scale loss formulation for learning a probabilistic model with proper score optimisation
- Learning joint probabilistic weather forecasts from station observations alone
BibTeX
@misc{erickson2026explainable,
title = {{Explainable Deep Learning for Probabilistic Nowcasting of Radar Reflectivity in Tornadic Storms}},
author = {Nathan Erickson and Amy McGovern and Aaron Hill},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.35675},
eprint = {2609.35675},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.35675},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Explainable Deep Learning for Probabilistic Nowcasting of Radar Reflectivity in Tornadic Storms
AU - Nathan Erickson
AU - Amy McGovern
AU - Aaron Hill
PY - 2026
DA - 2026/09/28
PB - arXiv
DO - 10.48550/arxiv.2609.35675
UR - https://arxiv.org/abs/2609.35675
AB - Tornadoes pose substantial risk to human life and property in the United States, causing more than 50 fatalities and $100 million of property damage on average annually. When tornadoes are likely, weather radar provides critical information for forecasters by providing information on storm morphology, storm motion, and intensity trends. Additional tools such as satellite and numerical weather prediction model runs can provide useful short-term information for understanding changes in storm characteristics. This work demonstrates a U-Net deep-learning system for nowcasting the evolution of radar reflectivity following tornadogenesis, which can provide value to forecasters by synthesizing large amounts of input data (e.g., radar imagery, near-storm environment data) and generating predictions of radar reflectivity from its inputs. Inputs to the model are radar imagery from the Multi-Radar Multi-Sensor (MRMS) dataset and near-storm environment data from the High-Resolution Rapid Refresh (HRRR) numerical weather prediction model. The U-Net is trained on a dataset of tornadic storms to produce 30 minutes of probabilistic predictions of radar reflectivity following tornadogenesis, with probabilistic predictions obtained by predicting parameters of the SinhArcSinh, or SHASH, distribution. The model produces physically realistic predictions of radar evolution, achieves comparable skill to next-hour forecasts from the HRRR, demonstrates reasonable probabilistic calibration and is accompanied by a variety of explainability methods to improve understanding by end users. Additionally, predictions from the model can be obtained much more quickly than those from a numerical weather prediction model. With further development, this model could be extended to nowcast radar reflectivity evolution in an operational setting.
ER -