The 2010 Russian drought impact on satellite measurements of solar-induced chlorophyll fluorescence: Insights from modeling and comparisons with parameters derived from satellite reflectances

Abstract

We examine satellite-based measurements of solar-induced chlorophyll fluorescence (SIF) over the region impacted by the Russian drought and heat wave of 2010. Like the popular Normalized Difference Vegetation Index (NDVI) that has been used for decades to measure photosynthetic capacity, SIF measurements are sensitive to the fraction of absorbed photosynthetically-active radiation (fPAR). However, in addition, SIF is sensitive to PAR as well as the fluorescence yield that is related to the photosynthetic yield. Both SIF and NDVI from satellite data show drought-related declines early in the growing season in 2010 as compared to other years between 2007 and 2013 for areas dominated by crops and grasslands. This suggests an early manifestation of the dry conditions on fPAR. Using MODIS fPAR retrievals, we computed SIF/APAR which is related to light use efficiencies (LUEs) for fluorescence and photosynthesis. We found drought-related losses in fluorescence efficiency for all areas including those dominated by mixed forests. Unlike croplands and grasslands, areas of mixed forest did not show significant drought-related declines in fPAR. We also simulated SIF and Gross Primary Productivity (GPP) using a global land surface model driven by observation-based meteorological fields. The model provides a reasonable simulation of the drought and heat impacts on SIF in terms of the spatial extents of anomalies, but some differences were found in timing of the peak drought response between modeled and observed SIF. Model data also suggested that drought-related declines in LUE for fluorescence and photosynthesis in areas of mixed forest produce losses in SIF and GPP. SIF and GPP losses due to drought in croplands and grasslands result from both LUE and fPAR reductions. The combination of SIF and NDVI or fPAR data is shown to be an important source of information for evaluating model performance.