Reconstruction of Daily 30 m Data from HJ CCD, GF-1 WFV, Landsat, and MODIS Data for Crop Monitoring

摘要: With the recent launch of new satellites and developments spatiotemporal data fusion methods, we are entering an era high resolution remote-sensing analysis. This study proposed a method to reconstruct daily 30 m for monitoring crop types phenology in two areas located Xinjiang Province, China. First, Spatial Temporal Data Fusion Approach (STDFA) was used time series from Huanjing satellite charge coupled device (HJ CCD), Gaofen no. 1 wide field-of-view camera (GF-1 WFV), Landsat, Moderate Resolution Imaging Spectroradiometer (MODIS) data. Then, reconstructed were applied extract using Hybrid Piecewise Logistic Model (HPLM). In addition, onset date greenness increase (OGI) decrease (OGD) also calculated simulated phenology. Finally, mapped information. The results show that had quality with proportion good observations (PGQ) higher than 0.95 HPLM approach can simulate Normalized Different Vegetation Index (NDVI) very well R2 ranging 0.635 0.952 Luntai 0.719 0.991 Bole, respectively. able single fields, which provided detailed pattern relative MODIS Moreover, be classified overall accuracy equal 0.91 is 0.028 0.046 those obtained by multi-temporal Landsat NDVI