Are dispersed rice planting dates linked to higher provincial yields? Evidence from Northeast Thailand

Abstract

Abstract Planting date is an important determinant of wet-season rice performance in monsoon Asia. However, it is influenced not only by rainfall but also by irrigation access, drainage conditions, and field operability. This study quantified how the spatial distribution and dispersion of planting dates were associated with provincial wet-season rice yields in Northeast Thailand. Planting dates at 1 km resolution for 2002–2023 were estimated from time series of the Enhanced Vegetation Index (EVI) derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) using the PhenoRice algorithm, and province-year indicators of mean planting date and within-province spatial standard deviation were derived. These indicators were analyzed together with June–August precipitation, June–October air temperature, and the Standardized Precipitation–Evapotranspiration Index (SPEI) for September–October. Satellite-derived monthly planted area for May–August showed moderate agreement with reported agricultural statistics, with a correlation coefficient of 0.65. Spatially, mean planting dates were earlier in the southern part of the region and later in the wetter northeast, suggesting that drainage and waterlogging constraints, rather than early-season rainfall availability alone, influenced establishment timing. Two-way fixed-effects regressions showed that planting-date dispersion was positively associated with yield, and that this association was stronger under wetter late-season conditions. These results suggest that spatial heterogeneity in planting date is a yield-relevant characteristic of wet-season rice systems beyond rainfall totals alone.