MAUNet-Light: A Concise MAUNet Architecture for Bias Correction and Downscaling of Precipitation Estimates

TL;DR

The paper tackles biases in precipitation estimates from satellite and climate models by developing MAUNet and a lightweight variant MAUNet-Light for bias correction and downscaling. It leverages a teacher-student framework with knowledge refinement to produce compact models that retain accuracy close to the full MAUNet. Empirical results show MAUNet outperforms baselines in bias correction and downscaling, while MAUNet-Light_KR achieves near-teacher performance with a substantially smaller parameter count, demonstrating a practical path to efficient operational products. The work highlights the potential of model compression and knowledge refinement for climate variable upscaling and bias adjustment, with implications for real-time, resource-constrained deployments, particularly for precipitation fields and extreme-event representation.

Abstract

Satellite-derived data products and climate model simulations of geophysical variables like precipitation, often exhibit systematic biases compared to in-situ measurements. Bias correction and spatial downscaling are fundamental components to develop operational weather forecast systems, as they seek to improve the consistency between coarse-resolution climate model simulations or satellite-based estimates and ground-based observations. In recent years, deep learning-based models have been increasingly replaced traditional statistical methods to generate high-resolution, bias free projections of climate variables. For example, Max-Average U-Net (MAUNet) architecture has been demonstrated for its ability to downscale precipitation estimates. The versatility and adaptability of these neural models make them highly effective across a range of applications, though this often come at the cost of high computational and memory requirements. The aim of this research is to develop light-weight neural network architectures for both bias correction and downscaling of precipitation, for which the teacher-student based learning paradigm is explored. This research demonstrates the adaptability of MAUNet to the task of bias correction, and further introduces a compact, lightweight neural network architecture termed MAUNet-Light.The proposed MAUNet-Light model is developed by transferring knowledge from the trained MAUNet, and it is designed to perform both downscaling and bias correction with reduced computational requirements without any significant loss in accuracy compared to state-of-the-art.

Figures (14)

• Figure 1: Schematic diagram of the MAUNet architecture.
• Figure 2: Schematic diagram of the MAUNet-Light architecture.
• Figure 3: Overview of the MAUNet-Light training process.
• Figure 4: Gridwise RMSE plots for MAUNet and MAUNet_Light_KR in downscaling.
• Figure 5: Gridwise correlation coefficient values illustrating the spatial variation in correlation between model predictions and observations.