Earlier Ecological Drought Detection by Involving the Interaction of Phenology and Eco‐Physiological Function

Abstract

Abstract Meteorological drought indices like the Standardized Precipitation Evaporation Index (SPEI) are frequently used to diagnose “ecological drought,” despite the fact that they were not explicitly designed for this purpose. More recently developed indices like the Evaporative Stress Index (ESI), which is based on the degree of coupling between actual to potential evapotranspiration, may better capture dynamic plant response to moisture limitations. However, the skill of these indices at describing plant water stress is rarely evaluated at sub‐seasonal timescales over which drought evolves. Moreover, it remains unclear how variability in phenological timing impacts and complicates early drought detection. Here, we compared the ability of ESI and SPEI to reflect the dynamics of ecological drought in forests and grasslands, based on anomalies of Gross Primary Productivity (GPP), surface conductance (Gs, a proxy for stomatal conductance), soil moisture, and vapor pressure deficit. ESI performed better than SPEI in capturing the dynamics of GPP and Gs, but still missed early ecological drought signals due to biases linked to earlier onset of spring leaf development. Thus, we developed a modified variant of the ESI (ESILAI) that accounts for the complicating effects of phenological shifts in leaf area index (LAI). The ESILAI detected drought onset up to 7–10 weeks earlier than SPEI and ESI. Additionally, drought onset dates determined from ESILAI are close to (±2 weeks) the dates determined from LAI‐corrected anomalies of Gs, and GPP, as well as the onset dates of soil water deficit and atmospheric aridity.

Keywords

• agricultural drought
• meteorological drought
• drought early warning
• drought forecasting