Using global products to improve under-standing of eco-hydrological processes and patterns in tropical catchments

Abstract

Understanding the ecohydrological processes at large-scale in tropical regions across Latin America has been a challenge due to climate variability and the lack of long temporal data series. As an alternative, remote sensing data and global products have been widely used in ungauged basins to get insights into key hydro-climatic patterns. Here, we study eco-hydrological processes in 27 tropical catchments across Colombia, Costa Rica and Mexico through the vegetation water use efficiency described by the Horton Index (HI) and the long-term water-energy balance described by the Budyko curve. Streamflow records from 2000 to 2014 were correlated with vegetation and climate indexes derived from MODIS, TRMM and CHIRPS at long-term and annual scales in order to find which factors presented more influence on the catchment’s response. Most of the Colombian and Costa Rican catchments were classified as energy-limited according to the dryness index (AI>1) and low water efficiency (HI ~ 0.48), meanwhile, Mexican catchments tended to be water-limited (AI>1) and more water efficient (HI ~ 0.52). Results suggested that climate is the strongest driver of water partitioning, where mean annual precipitation presented a strong linear relationship (ρ=0.93) with mean annual and intra-annual streamflow. Furthermore, the dryness index was a good predictor (r²=0.83 and 0.86, respectively) of the annual and long-term streamflow through a non-linear relationship similar to Budyko theory. Vegetation water use efficiency plays a secondary role in water partitioning, where HI was linearly negatively correlated with streamflow and baseflow (ρ=-0.78 and 0.82)