Probabilistic Deep Learning for Drought Forecasting: Role of Internal Climate Variability

2026-08-03

AI Summary

Traditional drought prediction often treats internal climate variability as unstructured noise, leading to underestimation of extreme drought risk. This study addresses this limitation by explicitly incorporating internal variability as a predictable component within a deep learning framework. We propose a physically consistent, risk-averse approach to estimate lower-bound drought projections by leveraging a large ensemble of climate model simulations to drive a probabilistic deep learning model. The method integrates reanalysis data with rigorous uncertainty quantification to characterize the most severe drought trajectories under adverse internal variability scenarios. The resulting ensemble-driven lower bound demonstrates superior calibration across most regions and seasons, particularly during anomalously dry periods, outperforming approaches relying solely on reanalysis data. By more accurately capturing tail-end drought risks, our framework provides a conservative yet informative reference for climate adaptation planning.

Abstract

Predicting drought risk is essential for anticipating impacts on water resources, agriculture, ecosystems, and climate adaptation planning. Yet drought forecasts remain uncertain because variability can substantially alter regional precipitation and evaporative demand. Treating this variability as unstructured noise ignores the fact that internal variability has spatial, seasonal, and temporal structure and thus contains information that can be used to improve drought forecasting. We propose a deep-learning-based forecasting framework for European drought prediction and extend it with an uncertainty-aware drought bound that explicitly incorporates internal forecast variability from a large climate model ensemble. This bound represents a physically plausible lower-tail trajectory of future drought conditions and marks how severe drought could plausibly become under an unfavourable realisation of internal variability, giving adaptation planning a conservative, risk-averse reference. We compare the proposed bound with a lower bound derived from reanalysis data only and show that our proposed ensemble-informed bound is better calibrated across most regions and seasons. This is specifically true during anomalously dry conditions, when historical reanalysis alone underestimates lower-tail drought risk. Our results show that internal variability should be treated as a forecast quantity in its own right. More broadly, large ensembles provide a practical way to transfer physically plausible climate variability into machine-learning drought forecasts, yielding risk-aware bounds that are more informative for drought assessment under shifting climate conditions.

Problem

• drought forecasting
• internal climate variability
• uncertainty quantification
• climate adaptation
• extreme drought risk

Innovation

• probabilistic deep learning
• internal climate variability
• drought forecasting
• large ensemble
• uncertainty quantification