A comprehensive calibration and validation of SWAT-T using local datasets, evapotranspiration and streamflow in a tropical montane cloud forest area with permeable substrate in central Veracruz, Mexico

Document Type


Publication Date



Department of Civil, Environmental, and Geospatial Engineering


Tropical montane cloud forests (TMCF) are threatened ecosystems despite their capacity to maintain high dry-season baseflow. A number of conservation policies, including payments for hydrological services, have been implemented to protect these forests. However, since most of the modeling tools used to assess the impacts of these policies were developed for temperate zones, more work is needed to understand and improve the applicability of popular models in tropical contexts. This study uses local evapotranspiration and streamflow datasets to calibrate and validate an improved version of the Soil and Water Assessment Tool model for the Tropics (SWAT-T). Vegetation growth and canopy water storage capacity were calibrated using field data. Three methods provided by SWAT-T to calculate potential evapotranspiration (PET) were compared: Penman-Monteith (SWAT-T-PM), Hargreaves (SWAT-T-HA), and Priestly-Taylor (SWAT-T-PT). Sensitivity analysis and calibration of daily streamflow were conducted at the catchment scale (34 km2). Furthermore, the calibrated models were validated at three sites with evapotranspiration data, and at four distinct micro-catchments (0.137–0.446 km2) with gauged streamflow data. Overall, SWAT-T satisfactorily simulated streamflow during the calibration period producing acceptable goodness of fit indices. However, the model incorrectly predicted the dominance of lateral flow instead of the deep groundwater flow observed from isotope-based studies. SWAT-T-HA performed better than SWAT-T-PM and SWAT-T-PT, but all models underestimated the influence of rainfall interception losses since evaporation is limited by daily PET in forests. Finally, SWAT-T largely over- and underestimated mean annual daily low flow in pastures and forests, respectively. Taken together, these results indicate that improvements in the parametrization of rainfall interception and deep subsurface flow dynamics in SWAT-T are required to improve applicability of this modeling tool in tropical montane areas underlain by permeable substrates.

Publication Title

Journal of Hydrology