Evaluación de la influencia de la cobertura boscosa en la regulación hídrica de la cuenca abastecedora del sistema Chingaza

Abstract

This study evaluated the spatiotemporal influence of forest cover on the hydrological regulation of the Chingaza Water Supply Basin during the 2016–2025 period. The research focused on the spatial characterization linking forest dynamics to streamflow variations within this strategic ecosystem. Using remote sensing in Google Earth Engine and Landsat 8 imagery, the Random Forest and Support Vector Machine (SVM) algorithms were compared, with Random Forest being selected based on its superior classification performance, achieving overall accuracies of 95.8% in 2016 and 94.3% in 2025. Multitemporal algebraic analysis revealed that, across an area of 246.45 km², vegetation regeneration exceeded deforestation, resulting in a net gain of 17.76 km² of High Andean forest. Subsequently, the InVEST Seasonal Water Yield model was implemented and parsimoniously calibrated using hydroclimatic time series from 1990 to 2025, achieving good performance with a Nash–Sutcliffe Efficiency (NSE) of 0.77 during validation. Scenario simulations demonstrated that forest conservation increased the Baseflow Index by up to 7.7% during the dry season while reducing quickflow by 17.5%. The main conclusion was that the recovery of High Andean forest enhances seasonal hydrological regulation by attenuating streamflow fluctuations, although it slightly reduces the total annual water yield due to increased evapotranspiration. These findings suggest that land management should preserve a balanced mosaic of High Andean forest and páramo ecosystems.