Temporal Context Conditioning for Seasonality-Aware Precipitation Nowcasting of High-Intensity Rainfall

2026-06-08

AI Summary

This study addresses the challenge that existing radar-based nowcasting models struggle to capture the influence of meteorological context—such as seasonality and diurnal cycles—on intense precipitation events. To overcome this limitation, the authors propose a lightweight, time-aware deep learning approach that integrates physically inspired temporal periodic encodings (specifically, time of day and time of year) into an enhanced SmaAt-UNet architecture. These temporal cues modulate intermediate features through time-conditioned layers coupled with a compact attention mechanism. The method introduces minimal additional parameters yet significantly improves forecast accuracy for rare, high-intensity rainfall events and more faithfully reproduces the seasonal variability and intensity distribution of precipitation. Its effectiveness is validated on KNMI radar data.

Abstract

Precipitation nowcasting is increasingly being approached with deep learning models that learn directly from recent radar observations. Although such models can efficiently capture short-term precipitation motion, they often lack broader contextual information about the meteorological conditions under which rainfall develops. This paper investigates whether lightweight temporal context can improve radar-based nowcasting, particularly for high-intensity rainfall. We propose the Time-Aware Small-Attention U-Net (TA-SmaAt-UNet), which extends the core SmaAt-UNet model with temporal conditioning layers that use cyclical encodings of time-of-day and time-of-year to modulate intermediate feature representations. Experiments on KNMI radar precipitation data show that temporal conditioning is most beneficial for rare, high-intensity precipitation events, while also improving the representation of seasonal variability and predicted rainfall-intensity distributions. A layer conductance analysis further indicates that the added temporal conditioning layers are actively used by the model despite their small parameter cost. These findings suggest that simple, physically motivated temporal context can improve the realism and reliability of deep learning-based precipitation nowcasts. The implementation of our models and training setup is available on \href{https://github.com/gijsvn/TA-SmaAt-UNet}{GitHub}.

Problem

• precipitation nowcasting
• high-intensity rainfall
• temporal context
• seasonality
• radar-based forecasting

Innovation

• temporal context conditioning
• precipitation nowcasting
• high-intensity rainfall
• seasonality-aware modeling
• lightweight neural architecture