This project responds directly to the solicitation for LCLUC studies in Southeast Asia by examining how the region is responding to simultaneous loss of agricultural labor and intensification of rice production. Major project objectives include: 1) Build a comprehensive multi-resolution, satellite image-derived database to characterize variability across and long-term changes within regional rice production systems; 2) Model current and past rice production under changing socio-economic and environmental conditions; 3) Use national population and agricultural censuses and other spatially-explicit secondary datasets compiled for sub- district units to quantify how changing conditions are correlated with changes in rice production systems through time; and 4) Conduct field interviews at selected sites to develop a place-based understanding of how rice farming is being revolutionized by changing demographics, economic opportunities, and technological innovations. We will explore these objectives for the major rice producing areas of four Mainland Southeast Asia (MSEA) countries (a total of six rice producing regions) between 1995 and 2018. The four countries and six regions include: 1) Vietnam (Red River and Mekong River Deltas), 2) Thailand (Northeast and Central Regions), 3) Laos (Savannakhet Province), and 4) Cambodia (Battambang Province). We will quantify changes in rice production systems between 1995 and 2018. As a means of quantifying long-term landscape dynamics in the persistently clouded study area, we will use an assemblage of complementary, cloud-resilient remote sensing analytical methods. First, we will classify Sentinel 2 SAR time series data (2014-2018) through an unsupervised rule-based clustering algorithm to differentiate stable standing water from flooded rice paddies to map locations and timing of rice production. Second, we will apply the Noise Insensitive Trajectory Algorithm (NITA) on Landsat (1995- 2018) and Sentinel 2 (2015-2018) time series data to quantify and map sub-annual changes in timing and pattern of rice production across our six study regions. NITA models land cover dynamics for every satellite image pixel across all available image dates and is the first all-available-images time series algorithm specifically designed to process data suffering from signal degeneration due to atmospheric effects. We will then input satellite-derived measures of area under rice production to the CSM-CERES-Rice model to estimate plot-level as well as regional annual rice yields. We will examine quantitative relationships between rice production, physiographic variables, and socioeconomic data gathered from agricultural and national censuses. A regression forest relating socioeconomic and physiographic variables to change in rice production systems will be used to identify the most significant predictors of change in each study region and across the study area. Finally, to understand how changes in labor dynamics and increasing demands for off-farm employment alter processes associated with rice production in land preparation, planting, weeding, harvesting, and the number of crops grown per year, we will conduct semi-informal interviews with key informants and survey 100 households in each rice growing region (total of 600 households). The project’s significance to NASA lies in its improved, multi-sensor approach for mapping changes in rice production systems-a change in land use rather than land cover, its use of novel cloud-resilient LCLUC monitoring approaches, and its integration of regional and local-scale perspectives of conditions that underlie observed changes to rice production systems (e.g., urbanization or industrialization). The knowledge generated by the proposed research will improve understanding of the social and ecological transformations affecting MSEA rice production and broadly advance globally-relevant theory on agriculture adaptation and change.