Democratization of Data: Spatial-Based Poverty Predictive Modeling in South Solok Regency Using Google Earth Engine

Abstract The availability of accurate and granular poverty data down to the sub-district level is often a challenge in regional development planning. Conventional data collection methods tend to be slow, high-cost, and carry the risk of subjectivity. This research introduces a prototype of a poverty monitoring and prediction system in South Solok Regency for the 2020-2023 period using the Google Earth Engine (GEE) platform. By applying the Ordinary Least Squares (OLS) Regression method to the synthesis of disaggregated secondary data, the model successfully achieved an accuracy level (R²) of 0.984. The results of this research not only provide a spatial visualization of poverty trends but also give a projection of the 2024 poverty figures as a basis for evidence-based policy and mitigation of pressure on protected forest areas. Keywords: Google Earth Engine, Poverty, South Solok, OLS Linear Regression, Spatial Intelligence.