Answers are grounded in this paper and cite their sources.
Study region Yangtze River Basin (YRB). Study focus This study examines how two intrinsic temporal properties of soil-moisture drought signals are associated with realized forecast skill: persistence-related singularity structure, characterized by the maximum singularity strength ( α max ), and multifractal complexity, quantified by the singularity-spectrum width (Δ α ). A Joint Standardized Soil Moisture Index (JSSI) was constructed across five accumulation scales, and multifractal detrended fluctuation analysis (MF-DFA) was used to characterize multiscale temporal structure . Support Vector Regression (SVR), Long Short-Term Memory (LSTM), and Temporal Convolutional Network (TCN) models were selected to forecast JSSI at 1-, 3-, and 6-month leads. New hydrological insights for the region the maximum singularity strengths ( α max ) and width of the singularity spectrum (Δ α ) show the most consistent associations with forecast skill and significant spatial organization across the basin. At the 1-month lead, α max provided the clearest common direction of skill variation, while responses to Δ α were more model dependent. Cross-model agreement was strongest at the 3-month lead, with correlations among fitted forecastability surfaces of 0.993–0.999. At the 6-month lead, α max remained a common organizing axis, while the association with Δ α became increasingly architecture dependent. Geopotential, long-term mean LAI, and soil type jointly explained 36.62% and 43.35% of the spatial variation in α max and Δ α , respectively. Overall, soil-moisture temporal structure provides an important context for drought forecastability.
No takes yet. Share an insight, caveat, or question.
Zhou et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: