2026/04/30 by David Herrera, Alexandre Belleflamme, Klaus Görgen +2
Energy · Earth and Planetary Sciences · Environmental Science · #TiO2 Photocatalysis and Solar Cells #Marine and coastal ecosystems #Fecal contamination and water quality
paper · doi:10.1016/j.rse.2026.115456
Passive optical satellite products have long been used to trace drought effects. Effective mitigation and understanding land–atmosphere interactions require variables reflecting plant physiological status under water scarcity. One such variable is Solar-Induced Chlorophyll Fluorescence (SIF), emitted directly from the photosynthetic apparatus. It provides direct information on vegetation functioning and can reveal stress within days. Spaceborne SIF observations have been available for over a decade and have been widely applied for vegetation stress detection. However, robust daily drought monitoring remains challenging because top-of-canopy SIF is strongly modulated by canopy structure and illumination, and retrieval noise can obscure short-term drought responses. This motivates normalization approaches that better isolate the physiological component of the fluorescence signal for near-real-time drought monitoring. To address this limitation, we estimated leaf-level fluorescence quantum efficiency (ΦF) by integrating TROPOMI SIF with photosynthetically active radiation (PAR) from the Breathing Earth Simulator (BESS), generating a daily 0.05° dataset for Germany (2018–2023). We evaluated ΦF as an early drought indicator in agricultural and forest ecosystems by comparing it with a subsurface water storage anomaly (SSWS) product from the coupled ParFlow/CLM model. Drought periods were identified as prolonged negative SSWS anomalies. Daily ΦF was aggregated, smoothed with a two-day rolling average, and analyzed via lagged cross-correlation with surrogate-based significance testing. ΦF consistently tracked negative SSWS anomalies with a two-day lag. This pattern was consistent across across the two analyzed land-cover classes, indicating that ΦF detects emerging reductions in subsurface water storage with a short delay. MODIS land surface temperature (LST) exhibited a complementary inverse response, peaking at 1–2 days underscoring that the correlation found was in fact linked to water availability. In contrast, TOC SIF and common vegetation indices (NIRv, NDVI) showed weak or inconsistent correlations. These results demonstrate that ΦF enables near-real-time detection of vegetation water stress and outperforms traditional optical indices for this purpose. The study highlights the need for downscaling and normalization to transform canopy SIF observations into an effective signal for early drought detection.