Satellite-based NDVI drought monitoring and its linkage with crop yield variability under climate uncertainty in India

Abstract:

Drought is among the most economically and agriculturally devastating manifestations of climate variability in India, affecting an estimated 68% of the country's geographical area and threatening the food security of hundreds of millions of smallholder farming households. The Normalised Difference Vegetation Index (NDVI), derived from satellite remote sensing platforms including MODIS, Landsat, and Sentinel-2, has emerged as the most widely applied proxy for drought-risk assessment and vegetation stress monitoring at regional and national scales. This systematic review synthesises peer-reviewed literature published between 2017 and 2026 to critically examine the methods, findings, and policy implications of satellite-based NDVI drought monitoring and its documented linkages with crop yield variability across India's diverse agro-climatic zones.

Drawing on 68 primary studies identified through systematic database searches, the review evaluates the performance of NDVI and derivative indices — including the Vegetation Condition Index (VCI), Standardised Vegetation Index (SVI), and Temperature Vegetation Dryness Index (TVDI) — in characterising drought onset, duration, and spatial extent. The review further examines the statistical and machine learning frameworks through which NDVI-derived signals have been linked to district-level and state-level crop yield records for rice, wheat, maize, sorghum, and groundnut. Evidence consistently demonstrates that NDVI-based drought indices explain between 40% and 80% of interannual crop yield variability under rainfed conditions, with predictive accuracy substantially enhanced through integration with soil moisture data, climate indices (SPI, PDSI), and Google Earth Engine-enabled cloud computing workflows. Critical evidence gaps — including insufficient sub-district spatial resolution, inadequate seasonal phenological disaggregation, and the underrepresentation of smallholder heterogeneity in existing models — are identified alongside a structured research agenda for next-generation drought monitoring systems. The review is intended as a comprehensive reference for postgraduate researchers and practitioners in remote sensing, agricultural geosciences, climate science, and food security planning.