Harmonization of Landsat and Sentinel 2 for Crop Monitoring in Drought Prone Areas: Case Studies of Ninh Thuan (Vietnam) and Bekaa (Lebanon)

摘要: Proper satellite-based crop monitoring applications at the farm-level often require near-daily imagery medium to high spatial resolution. The combination of data from different ongoing satellite missions Sentinel 2 (ESA) and Landsat 7/8 (NASA) provides this unprecedented opportunity a global scale; however, is rarely implemented because these procedures are demanding computationally intensive. This study developed robust stream processing for harmonization 7, 8 in Google Earth Engine cloud platform, connecting benefit coherent structure, built-in functions computational power Cloud. harmonized surface reflectance images were generated two agricultural schemes Bekaa (Lebanon) Ninh Thuan (Vietnam) during 2018–2019. We evaluated performance several pre-processing steps needed including image co-registration, Bidirectional Reflectance Distribution Functions correction, topographic band adjustment. found that misregistration between varied 10 m 32 (Lebanon), posed great impact on quality final set if not treated. Analysis pair overlapped L8-S2 over region showed that, after harmonization, all band-to-band correlations greatly improved. Finally, we demonstrated an application dense mapping monitoring. An harmonic (Fourier) analysis was applied fit detected unimodal, bimodal trimodal shapes temporal NDVI patterns one year province. derived phase amplitude values cycles combined with max-NDVI as R-G-B false composite image. able highlight croplands bright colors (high amplitude), while non-crop areas shown grey/dark (low amplitude). sets (with 30 resolution) along scripts used provided public use.