Satellite Data for Nowcasting: Estimating Cambodia’s GDP in Real Time Using Satellite Data in a Machine Learning Framework

Selected Issues Papers, January 8, 2026

Source details

Canonical URL: Satellite Data for Nowcasting: Estimating Cambodia’s GDP in Real Time Using Satellite Data in a Machine Learning Framework

Combining satellite-derived indicators (e.g., nighttime lights, NO₂ emissions, vegetation indices) with traditional high-frequency indicators in a machine learning framework significantly improves the accuracy of GDP nowcasts.
Satellite data enables closer examination of subnational patterns, providing granular, near-real-time insights into economic activity.
Non-traditional data sources can complement conventional methods and strengthen macroeconomic surveillance in data-scarce environments.

Use of satellite-derived indicators: nighttime lights, NO₂ emissions, vegetation indices.
Integration with traditional high-frequency indicators.
Application of machine learning methods (keywords include: random forest, machine learning, nowcast, nowcasting).

Non-traditional, high-frequency satellite data should be considered to supplement official statistics where capacity limitations exist in the production and timely release of key indicators.
Incorporating satellite data into nowcasting frameworks can improve the evidence base for data-driven policy decisions and macroeconomic surveillance.
Subnational analysis enabled by satellite indicators can inform more granular policy targeting.

Subjects: Agricultural sector, Economic forecasting, Economic sectors, Health.
Keywords: Agricultural sector, big data, Global, IMF country, IMF staff, issues paper, machine learning, non-traditional data, nowcast, nowcasting, policy decision, random forest, satellite, satellite data.