Probabilistic Tracking of Annual Cropland Changes over Large, Complex Agricultural Landscapes Using Google Earth Engine

Cropland expansion is expected to increase across sub-Saharan African (SSA) countries in the next thirty years meet growing food needs continent. These land transformations will have cascading social and ecological impacts that can be monitored using novel Earth observation techniques produce datasets complementary national cropland surveys. In this study, we present a flexible Bayesian data synthesis workflow on Google Engine (GEE) used fuse optical synthetic aperture radar demonstrate its ability track agricultural change at scales. We adapted previously developed Updating of Land Cover (Unsupervised) algorithm (BULC-U) by integrating shapelet slope thresholding identify locations dates implemented tiling scheme allow processing large volumes imagery. apply approach map annual from 2000 2015 for Zambia (750,000 km2), country experiencing rapid growth land. applied our mapping time series unsupervised classifications Landsat 5, 7, 8, Sentinel-1, ALOS PALSAR within 1476 tiles covering Zambia. The changes maps reveal active between Zambia, especially Southern, Central, Eastern provinces. Our accuracy assessment estimates identified 27.5% 69.6% total (commission errors 6.1% 37.6%), depending threshold. results usefulness fusion shapelet, slope-based synthesize monitoring situations where training are scarce. addition, provide one first spatially continuous, annually incremented accounts region. flexible, cloud-based GEE enables multi-sensor, national-scale low cost users.