RBI study shows satellite lights and Google data cut state growth estimation errors by up to 62%

The Reserve Bank of India has published a study showing that satellite night-time light imagery and Google search data can dramatically improve real-time estimates of state-level economic growth, reducing forecasting errors by up to 62 per cent.

Published in the central bank's October Bulletin, the paper titled Nowcasting State-Level Economic Activity in India with Night-Time Light and Google Trends outlines a quarterly framework designed to monitor economic momentum much faster than traditional statistical releases allow.

Official Gross State Value Added (GSVA) figures are usually released annually with a delay of 12 to 18 months. That time gap has long made it difficult for financial planners and government bodies to assess local economic health in real time. The proposed nowcasting framework provides timely signals to complement official statistics, reinforcing India's data-driven governance and fiscal management.

How night lights and search trends track state GDP

Researchers tested the framework across three major industrial powerhouses: Gujarat, Maharashtra and Karnataka. The model combined satellite measurements of night-time illumination with Google Trends search indices across 25 categories, including employment, education, finance, shopping, real estate and motor vehicles. It also incorporated historical GSVA trends.

When evaluated against a benchmark relying strictly on past growth figures, the combined approach delivered substantial gains in accuracy.

The research established a strong statistical link between year-on-year increases in night-time light intensity and overall state value added. Maharashtra displayed the highest correlation at 0.90, followed by Karnataka at 0.79 and Gujarat at 0.66. These relationships remained robust even after filtering out economic fluctuations caused by the Covid-19 pandemic.

Strengthening economic planning

The study noted that satellite night-light data provided predictive indications of future economic expansion, while search engine trends offered accurate real-time signals of consumer and business activity.

While the model has so far been evaluated on three leading industrial states, its success demonstrates the potential for modern data tools to enhance state-level economic monitoring across the country.