Time series analysis of rice cropping seasons and production using remote sensing images: A case study in Dong Thap province, Vietnam

Monitoring of rice crop variability in Dong Thap is limited by reliance on manual surveys, the lack of continuous data series, and difficulty distinguishing changes between farming models. To address this gap, the study aims to (i) identify changes in the structure of double rice crops and triple rice crops in the period 2017–2021, and (ii) assess the impact of such changes on rice yield. Sentinel-1A SAR data series were processed on Google Earth Engine through the following steps: speckle noise filtering, VH reflectance extraction over time, growth curve construction, and crop clustering using the K-means algorithm. The resulting layers were integrated with GIS for overlaying, analyzing spatial-temporal fluctuations, and combining with household survey data to estimate yield. The results show that the two rice-crop areas decreased by 7,038.5 hectares, and the three rice-crop areas decreased by 4,068.88 hectares from 2017 to 2021. Despite the decrease in area, the province's rice output still increased from 2.88 to 2.95 million tons, reflecting productivity improvements and changes in production structure. These findings provide important evidence to support managers in adjusting rice land planning, allocating water resources, reviewing the dike system, and developing climate change adaptation policies for production.