McCast: Memory-Guided Latent Drift Correction for Long-Horizon Precipitation Nowcasting

2026-05-13

AI Summary

This work addresses the challenge of trajectory drift in long-horizon precipitation nowcasting, which arises from error accumulation in autoregressive models and the lack of mechanisms to model and correct the global temporal evolution of meteorological systems. To this end, the authors propose McCast, a novel approach that reimagines memory mechanisms from passive conditioning to active drift correction. Central to McCast is the Drift-Correction Memory Bank (DCBank), which explicitly refines prediction trajectories in latent space through a correction latent extractor and a memory-based retrieval module. Evaluated on the SEVIR and MeteoNet benchmarks, McCast achieves state-of-the-art performance, demonstrating significantly improved temporal consistency and reliability in long-horizon forecasting scenarios.

Abstract

Existing precipitation nowcasting methods typically adopt an autoregressive formulation, where future states are predicted from previous outputs. However, such an approach accumulates errors over long rollouts, causing forecasts to drift away from physically plausible evolution trajectories. Although various studies have attempted to alleviate this problem by improving step-wise prediction accuracy, they largely neglect the global temporal evolution of meteorological systems and lack mechanisms to actively correct drift during rollouts. To address this issue, we propose McCast, a memory-guided latent drift correction method for precipitation nowcasting. Rather than treating memory as an unordered dictionary of latent states for passive conditioning, McCast leverages temporally organized memory to actively correct autoregressive latent evolution. Specifically, McCast introduces a Drift-Corrective Memory Bank (DCBank) that explicitly estimates the temporally consistent drift corrections to calibrate the divergent trajectory. DCBank performs drift correction in two stages: a Corrective Latent Extractor first predicts an initial correction from the current prediction and a reference latent state, and a Correction-Aware Memory Retrieval module then refines the initial correction using temporally organized historical memory. By explicitly correcting latent evolution, instead of improving step-wise prediction accuracy only, McCast produces more temporally coherent and reliable long-horizon forecasts. Experiments on two widely used benchmarks, SEVIR and MeteoNet, show that McCast achieves state-of-the-art performance, particularly in challenging long-horizon forecasting scenarios.

Problem

• precipitation nowcasting
• autoregressive error accumulation
• latent drift
• long-horizon forecasting
• temporal coherence

Innovation

• memory-guided
• latent drift correction
• autoregressive nowcasting
• temporal coherence
• precipitation forecasting