Using remotely sensed information to improve vegetation parameterization in a semi-distributed hydrological model (SMART) for upland catchments in Australia

Appropriate representation of the vegetation dynamics is crucial in hydrological modelling. To improve an existing limited vegetation parameterization in a semi-distributed hydrologic model, called the Soil Moisture and Runoff simulation Toolkit (SMART), this study proposed a simple method to incorporate daily leaf area index (LAI) dynamics into the model using mean monthly LAI climatology and mean rainfall. The LAI-rainfall sensitivity is governed by a parameter that is optimized by maximizing the Pearson correlation coefficient (R) between the estimated and satellite-derived LAI time series. As a result, the LAI-rainfall sensitivity is smallest for forest, shrub, and woodland regions across Australia, and increases for grasslands and croplands. The impact of the proposed method on catchment-scale simulations of soil moisture (SM), evapotranspiration (ET) and discharge (Q) in SMART was examined across six eco-hydrologically contrasted upland catchments in Australia. Results showed that the proposed method produces almost identical results compared to simulations by the satellite-derived LAI time series. In addition, the simulation results were considerably improved in nutrient/light limited catchments compared to the cases with the default vegetation parameterization. The results showed promise, with possibilities of extension to other hydrologic models that need similar specifications for inbuilt vegetation dynamics.