MP-MoE: Matrix Profile-Guided Mixture of Experts for Precipitation Forecasting

TL;DR

The paper introduces MP-MoE, a novel framework that combines intensity and structural-aware losses to improve precipitation forecasts by better capturing storm morphology and reducing phase shift penalties.

Contribution

It proposes a Matrix Profile-guided Mixture of Experts framework that enhances rainfall prediction accuracy by integrating subsequence-level similarity into the loss function.

Findings

Outperforms baseline models in heavy rainfall detection.

Reduces DTW values, indicating better storm shape preservation.

Improves CSI-M scores for multiple rainfall horizons.

Abstract

Precipitation forecasting remains a persistent challenge in tropical regions like Vietnam, where complex topography and convective instability often limit the accuracy of Numerical Weather Prediction (NWP) models. While data-driven post-processing is widely used to mitigate these biases, most existing frameworks rely on point-wise objective functions, which suffer from the ``double penalty'' effect under minor temporal misalignments. In this work, we propose the Matrix Profile-guided Mixture of Experts (MP-MoE), a framework that integrates conventional intensity loss with a structural-aware Matrix Profile objective. By leveraging subsequence-level similarity rather than point-wise errors, the proposed loss facilitates more reliable expert selection and mitigates excessive penalization caused by phase shifts. We evaluate MP-MoE on rainfall datasets from two major river basins in Vietnam across multiple horizons, including 1-hour intensity and accumulated rainfall over 12, 24, and 48 hours. Experimental results demonstrate that MP-MoE outperforms raw NWP and baseline learning methods in terms of Mean Critical Success Index (CSI-M) for heavy rainfall events, while significantly reducing Dynamic Time Warping (DTW) values. These findings highlight the framework's efficacy in capturing peak rainfall intensities and preserving the morphological integrity of storm events.