A Random Forest Algorithm for Landsat Image Chromatic Aberration Restoration Based on GEE Cloud Platform—A Case Study of Yucatán Peninsula, Mexico

With the growth of cloud computing, use Google Earth Engine (GEE) platform to conduct research on water inversion, natural disaster monitoring, and land change using long time series Landsat images has also gradually become mainstream. are currently one most important image data sources for remote sensing inversion. As a result changes in weather conditions single-view images, varying radiances acquired; hence, monthly or annual scale mosaic multi-view results strip color variation. In this study, NDWI MNDWI within 50 km coastline Yucatán Peninsula from 1993 2021 used as object study GEE platform, areas with chromatic aberrations reconstructed TOA (top atmosphere reflectance) SR (surface data. The DN (digital number) values probability distributions reference be restored classified counted independently random forest algorithm, classification mapped area histogram-matching manner. MODIS Sentinel-2 products comparison validation. demonstrate that do not exhibit obvious band aberration, stacking is smoother; provide improved bodies, correlation between high 0.5358 0.5269, respectively. addition, none existing can effectively eliminate aberration bands.