Soil Moisture as a Persistent Source of Subseasonal Flash-Drought Predictability

Flash droughts (FDs) intensify within days to weeks, threatening agriculture, ecosystems, and water resources. Their subseasonal predictability depends on atmospheric forcing, land-surface memory, and groundwater dynamics. Here, we assess global FD predictability using operational ECMWF Subseasonal-to-Seasonal (S2S) forecasts by integrating soil moisture (SM), total precipitation (TP), latent heat flux (LHF), and vapor pressure deficit (VPD) into a multivariate Flash Drought Intensity Index (FDII), with additional analyses of groundwater-related memory. SM exhibits the strongest and most persistent forecast-reference anomaly correlation coefficient (ACC), retaining over 93% of its Pentad-4 spatially pooled ACC through Pentad-9, with the hierarchy SM > LHF > VPD > TP. SM attribution increases from approximately 34% in the tropics to 69% at high latitudes, while LHF remains important in the tropics. Attribution weighting increases the global mean grid-cell correlation across forecast leads from 0.21 for equal weighting to 0.35, a 67% relative improvement, although both multivariate formulations remain below the SM-only correlation of 0.43. Importantly, FDII SM exceeds persistence in only approximately 4% of grid cells, limiting evidence of widespread improvement over persistence. Groundwater-inclusive forecasts show localized ACC improvements exceeding 0.1, although groundwater's independent contribution remains unresolved. Across evaluated croplands, attribution-weighted forecasts achieve a mean forecast-minus-persistence difference in the Symmetric Extremal Dependence Index (Delta SEDI) of 0.023, with positive Delta SEDI across 53.5% of evaluated cropland. These findings identify SM as the strongest baseline for subseasonal FD predictability while highlighting the limited spatial extent of improvement over persistence and the need for further evaluation before operational warning applications.

Authors

Institutions

Publication Details

Published
2026-11-12
DOI
https://doi.org/10.22541/essoar.15012099/v1
Primary Topic
Hydrology and Drought Analysis
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Soil Moisture as a Persistent Source of Subseasonal Flash-Drought Predictability

Yusang Oh, Venkatesh Budamala, Jeffrey B. Basara, Hyunglok Kim et al.
Hydrology and Drought Analysis
preprint

Soil Moisture as a Persistent Source of Subseasonal Flash-Drought Predictability

Yusang Oh, Venkatesh Budamala, Jeffrey B. Basara, Hyunglok Kim, Hoang Hai Nguyen, Euiyoung Choi
preprint en

Abstract

Flash droughts (FDs) intensify within days to weeks, threatening agriculture, ecosystems, and water resources. Their subseasonal predictability depends on atmospheric forcing, land-surface memory, and groundwater dynamics. Here, we assess global FD predictability using operational ECMWF Subseasonal-to-Seasonal (S2S) forecasts by integrating soil moisture (SM), total precipitation (TP), latent heat flux (LHF), and vapor pressure deficit (VPD) into a multivariate Flash Drought Intensity Index (FDII), with additional analyses of groundwater-related memory. SM exhibits the strongest and most persistent forecast-reference anomaly correlation coefficient (ACC), retaining over 93% of its Pentad-4 spatially pooled ACC through Pentad-9, with the hierarchy SM > LHF > VPD > TP. SM attribution increases from approximately 34% in the tropics to 69% at high latitudes, while LHF remains important in the tropics. Attribution weighting increases the global mean grid-cell correlation across forecast leads from 0.21 for equal weighting to 0.35, a 67% relative improvement, although both multivariate formulations remain below the SM-only correlation of 0.43. Importantly, FDII SM exceeds persistence in only approximately 4% of grid cells, limiting evidence of widespread improvement over persistence. Groundwater-inclusive forecasts show localized ACC improvements exceeding 0.1, although groundwater's independent contribution remains unresolved. Across evaluated croplands, attribution-weighted forecasts achieve a mean forecast-minus-persistence difference in the Symmetric Extremal Dependence Index (Delta SEDI) of 0.023, with positive Delta SEDI across 53.5% of evaluated cropland. These findings identify SM as the strongest baseline for subseasonal FD predictability while highlighting the limited spatial extent of improvement over persistence and the need for further evaluation before operational warning applications.

Agricultural Research Service (US), University of Massachusetts Lowell (US), Gwangju Institute of Science and Technology (KR)
Climate action, Zero hunger
Hydrology and Drought Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon