SmaAt-UNet: Precipitation Nowcasting using a Small Attention-UNet Architecture

TL;DR

This paper introduces SmaAt-UNet, a lightweight neural network architecture with attention modules for accurate short-term precipitation forecasting, outperforming traditional models in efficiency while maintaining comparable accuracy.

Contribution

It presents a novel, efficient UNet-based architecture with attention and depthwise-separable convolutions for precipitation nowcasting, reducing model complexity significantly.

Findings

Achieves comparable accuracy to larger models with only a quarter of parameters.

Effective on real-world datasets from the Netherlands and France.

Demonstrates potential for real-time weather prediction applications.

Abstract

Weather forecasting is dominated by numerical weather prediction that tries to model accurately the physical properties of the atmosphere. A downside of numerical weather prediction is that it is lacking the ability for short-term forecasts using the latest available information. By using a data-driven neural network approach we show that it is possible to produce an accurate precipitation nowcast. To this end, we propose SmaAt-UNet, an efficient convolutional neural networks-based on the well known UNet architecture equipped with attention modules and depthwise-separable convolutions. We evaluate our approaches on a real-life datasets using precipitation maps from the region of the Netherlands and binary images of cloud coverage of France. The experimental results show that in terms of prediction performance, the proposed model is comparable to other examined models while only using a quarter of the trainable parameters.