EO-WM: A Physically Informed World Model for Probabilistic Earth Observation Forecasting

2026-06-25

AI Summary

This study addresses the challenge of Earth observation forecasting under meteorological forcing, where sparse observations and unobservable surface states hinder accurate modeling of weather-induced responses. To tackle this, the task is formulated as a partially observable, weather-driven world modeling problem. The authors propose a probabilistic prediction framework based on a video diffusion Transformer, featuring a novel dual-path conditioning mechanism that disentangles climatic baselines from weather anomalies and integrates cumulative physical stress signals. Two new benchmarks are introduced to evaluate model responsiveness to extreme and shifting weather conditions. Experimental results demonstrate that the proposed method reduces prediction error in NDVI decline magnitude by 5.63% and improves directional accuracy by 7.80%, while maintaining competitive performance on standard pixel-level metrics.

Abstract

Earth Observation (EO) forecasting aims to predict future Earth surface dynamics from satellite observations under changing meteorological conditions. In this paper, we view this task as a partially observed, weather-driven world modeling problem, in which weather acts as a conditioning signal, while forecasting remains uncertain due to sparse observations and unobserved land-surface states. However, existing methods do not fully capture this setting: deterministic models collapse uncertainty into a single future prediction, while diffusion-based methods typically treat weather variables as undifferentiated conditioning signals, and existing benchmarks focus mainly on reconstruction accuracy rather than whether forecasts respond correctly to changed weather forcing.We introduce EO-WM, a video diffusion transformer for multispectral EO forecasting. EO-WM incorporates a physically informed conditioning framework that represents meteorological forcing through a climatological baseline, weather anomalies, and cumulative physical stress signals. Specifically, it separates baseline and anomaly through distinct conditioning pathways, and accumulates anomalous forcing over time to capture sustained heat and drought stress. To evaluate weather-response behavior beyond standard metrics, we introduce two diagnostic benchmarks: an Extreme Summer Benchmark for severity-aware prediction of vegetation degradation under extreme weather, and a Seasonal Matched-Pair Benchmark for testing response fidelity under changed weather forcing. Experiments show that EO-WM reduces the error in predicted Normalized Difference Vegetation Index (NDVI) decline amplitude by a relative 5.63% and improves directional hit rate by a relative 7.80%, while remaining competitive on standard pixel-level metrics. The benchmarks and model will be made open-source at https://github.com/Luo-Z13/EO-WM.

Problem

• Earth Observation forecasting
• weather-driven world modeling
• uncertainty quantification
• meteorological forcing
• land-surface dynamics

Innovation

• physically informed conditioning
• video diffusion transformer
• weather anomaly accumulation
• extreme weather benchmarking
• probabilistic Earth observation forecasting