Spatio-temporal Drought Dynamics in theParnaíba River Basin (2000–2024): A Multi-scale Analysis Integrating MODIS-derived Vegetation Anomalies and Climatic Trends

Drought represents one of the most severe hydroclimatic hazards affectingthe Brazilian Northeast, yet the spatio-temporal dynamics of vegetation-baseddrought and its relationships with multi-variable climatic trends remain poorlycharacterised at sub-basin scale in the Parna´ıba River Basin (PRB). This studyaddresses this gap by providing a comprehensive multi-scale assessment ofdrought dynamics across the PRB over the 2000–2024 period. The Standard-ized Vegetation Index (SVI) was computed pixel-wise from the MODIS TerraMOD13A1 500 m time series processed in Google Earth Engine, and integratedwith annual hydroclimatic variables (precipitation, air temperature, referenceevapotranspiration, actual evapotranspiration, and soil moisture) derived fromCHIRPS and ERA5-Land reanalysis. Long-term trends were assessed using theMann–Kendall test and Sen’s slope estimator at both basin and pixel scales.Basin-scale and sub-basin-scale climate–vegetation relationships were examinedthrough Spearman rank correlations with temporal lags of 0, 1, and 2 years. Sub-basin drought behaviour was classified using k-means clustering applied to SVItemporal profiles across 12 sub-basins. The SVI accurately captured the 2012–2017 megadrought as the most severe event of the record, with basin-mean SVIreaching −1.02 in 2012 and −1.08 in 2015. A statistically significant basin-widewarming trend of +0.034 °C yr−1 (p = 0.006) was detected, accompaniedby significant negative trends in actual evapotranspiration (−3.33 mm yr−1; p = 0.024) and soil moisture (p = 0.034), despite the absence of a significantprecipitation trend (p = 0.761). Precipitation emerged as the dominant driverof vegetation dynamics (r = 0.65 at lag-0; r = 0.57 at lag-2 years), evidencingmulti-year vegetation memory effects. Clustering identified three ecohydrologi-cal sub-basin groups reflecting distinct drought response regimes associated withthe Caatinga semi-arid interior, the Cerrado–MATOPIBA agricultural frontier,and the coastal lower basin. These results demonstrate that the PRB operatesunder a water-limited regime in which regional warming intensifies atmosphericevaporative demand, driving progressive land-surface drying particularly in thesouthern and central sectors. The spatially explicit framework developed hereprovides a scientific basis for adaptive water resource management and sustain-able land-use planning in large basins facing the synergistic pressures of climatechange and agricultural expansion.