Comparison of Phenology Estimated From Monthly Vegetation Indices and Solar-Induced Chlorophyll Fluorescence in China

Abstract

Phenology is an important biological indicator for monitoring terrestrial ecosystems and global change. Solar-induced chlorophyll fluorescence (SIF) emitted by chlorophyll has been proven to characterize vegetation photosynthesis and phenology. In this study, we used monthly normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), and SIF products to qualitatively compare the effectiveness at detecting the phenological characteristics (SOS (start-of-season), EOS (end-of-season), and LOS (length-of-season)) over China during 2007–2013. The phenological characteristics determined by gross primary productivity (GPP) were applied as the reference to validate the phenological characteristics derived from NDVI, EVI, and SIF. The results demonstrated that the phenological characteristics derived from SIF were more consistent with that of GPP than VIs (NDVI and EVI) when considering all latitude grades, different elevation grades, and different land cover types in China. In the middle- and high-latitude regions, SOS derived from the vegetation indices (SOSVIs) did not deviate from those from GPP (SOSGPP) and SIF (SOSSIF), while in low latitudes, SOSVIs were about 20 d later than SOSSIF and SOSGPP. The VIs (EOSVIs) had a severe lag behind those of SIF (EOSSIF) in estimating the EOS at all latitudes. The EOSSIF had a deviation of fewer than 5 d compared with EOS estimated by GPP (EOSGPP), whereas the deviation of EOSVIs from EOSGPP was about 10–31 d across low to high latitude regions. The biases of SIF and VIs were due to the inconsistency between vegetation photosynthesis and leaf greenness. Also, VIs overestimated the LOS at all latitudes, the difference of LOS between estimated by NDVI and estimated by GPP was as long as 39 d in the high-latitude region. Our study suggests that SIF is suitable for estimating the phenological characteristics of vegetation regardless of different latitudes, elevation grades, and land cover types in China, providing a basis for SIF to study the vegetation phenological characteristics in a regional scope.