Zhou, XJ; Wang, PX; Tansey, K; Zhang, SY; Li, HM; Wang, L (2020). Developing a fused vegetation temperature condition index for drought monitoring at field scales using Sentinel-2 and MODIS imagery. COMPUTERS AND ELECTRONICS IN AGRICULTURE, 168, 105144.

Time-series high spatial resolution drought monitoring is essential for effective agricultural management. For more than a decade, the multiyear vegetation temperature condition index (VTCI) based on the Advanced Very High Resolution Radiometer (AVHRR) and Moderate Resolution Imaging Spectroradiometer (MODIS) has been applied to regional drought monitoring. However, the spatial resolutions of AVHRR and MODIS (1 km) often represent mixtures of built-up areas and different vegetation or crop types. In this paper, a framework is proposed to obtain a ten-day interval multiyear VTCI at field scales, which is fused from Sentinel-2 data with a fine spatial resolution (20 m) and ten-day interval Terra MODIS data with a coarse spatial resolution using the Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM). This framework includes the disaggregation of the Terra MODIS land surface temperature (LST) using Sentinel-2 data and digital elevation model (DEM) data, the spatiotemporal fusion of normalized difference vegetation index (NDVI) and LST, and drought monitoring by the fused VTCI. The accuracy of this framework for regional drought monitoring was tested in the Guanzhong Plain of China. The results indicate that 1) the Sentinel-2 biophysical products combined with DEM data based on support vector regression can accurately disaggregate the Terra MODIS LST; 2) the ESTARFM has good capability to fuse the NDVI and LST derived from Sentinel-2 and Terra MODIS; 3) the multiyear spatiotemporally fused VTCI, a quantitative drought monitoring index at field scales, can be calculated by using the linear regression between the single-year fused VTCI based on the data in 2018 and the multiyear Terra MODIS VTCI based on an 18-year data record; and 4) the spatiotemporal fusion of the multiyear VTCI can significantly improve the drought monitoring accuracy in the winter wheat and woodland area of the Guanzhong Plain. From mid-March to early May, the multiyear spatiotemporally fused VTCIs have better correlation with the cumulative precipitation over the past 20 days (R-2 is 0.83 in the entire study area) than the multiyear Terra MODIS VTCIs (R-2 is 0.80 in the entire study area). The results of this study demonstrate the potential of using the spatiotemporal fusion algorithm to obtain the multiyear VTCI at field scales and provide an effective method for improving the accuracy of drought monitoring.