05 Jun 2023
 | 05 Jun 2023
Status: this preprint is open for discussion.

DEUCE v1.0: A neural network for probabilistic precipitation nowcasting with aleatoric and epistemic uncertainties

Bent Harnist, Seppo Pulkkinen, and Terhi Mäkinen

Abstract. Precipitation nowcasting (forecasting locally for 0–6 h) serves both public security and industries, facilitating the mitigation of losses incurred due to e.g. flash floods, and is usually done by predicting weather radar echoes, which provides better performance than NWP at that scale. Probabilistic nowcasts are especially useful as they provide a desirable framework for operational decision-making. Many extrapolation-based statistical nowcasting methods exist, but they all suffer from a limited ability to capture the nonlinear growth and decay of precipitation, leading to a recent paradigm shift towards deep learning methods, more capable of representing these patterns.

Despite of its potential advantages, the application of deep learning in probabilistic nowcasting has only recently started to be explored. Here we develop a novel probabilistic precipitation nowcasting method, based on Bayesian neural networks with variational inference and the U-Net architecture, named DEUCE. The method estimates the total predictive uncertainty of precipitation by combining estimates of the epistemic (knowledge-related, reducible) and heteroscedastic aleatoric (data-dependent, irreducible) uncertainties, and produces an ensemble of development scenarios for the following 60 minutes.

DEUCE is trained and verified using Finnish Meteorological Institute radar composites against established classical models. Our model is found to produce both skillful and reliable probabilistic nowcasts based on various evaluation criteria. It improves ROC Area Under the Curve scores 1–5 % over STEPS and LINDA-P baselines, and comes close to the best-performer STEPS on a CRPS metric. The reliability of DEUCE is demonstrated with, e.g., having the lowest Expected Calibration Error at 20 and 25 dBZ reflectivity thresholds, and coming second at 35 dBZ. On the other hand, deterministic performance of ensemble means is found to be worse than that of extrapolation and LINDA-D baselines. Lastly, the composition of the predictive uncertainty is analysed and described, with the conclusion that aleatoric uncertainty is more significant and informative than epistemic uncertainty in the DEUCE model.

Bent Harnist et al.

Status: open (extended)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2023-1100', Anonymous Referee #1, 25 Sep 2023 reply

Bent Harnist et al.

Data sets

Data for the manuscript "DEUCE v1.0: A neural network for probabilistic precipitation nowcasting with aleatoric and epistemic uncertainties" by Harnist et al. (2023) Bent Harnist, Seppo Pulkkinen, Terhi Mäkinen

Model code and software

fmidev/deuce-nowcasting: Initial release of the source code for the manuscript Bent Harnist

Bent Harnist et al.


Total article views: 326 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
198 116 12 326 5 3
  • HTML: 198
  • PDF: 116
  • XML: 12
  • Total: 326
  • BibTeX: 5
  • EndNote: 3
Views and downloads (calculated since 05 Jun 2023)
Cumulative views and downloads (calculated since 05 Jun 2023)

Viewed (geographical distribution)

Total article views: 325 (including HTML, PDF, and XML) Thereof 325 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 30 Sep 2023
Short summary
Probabilistic precipitation nowcasting (local forecasting for 0–6 hours) is crucial for reducing damage from events like flash floods. For this goal, we propose the DEUCE neural network-based model, which uses data and model uncertainties to generate an ensemble of potential precipitation development scenarios for the next hour. Trained and evaluated with Finnish precipitation composites, DEUCE was found to produce more skillful and reliable nowcasts than established models.