Modelling hydrological impact of remotely sensed vegetation change

Abstract

: Vegetation cover over a catchment could be altered by a changing climate, bushfire or human interventions. This will result in changes of catchment hydrological response that may affect catchment water resources management. For developing adaptative water resources management strategies under a changing climate, it is essential to take into account hydrological responses to the dynamics of vegetation. In this study, we adapt an existing hydrological model (GR4J) by incorporating remotely sensed vegetation cover (represented by leaf area index) into the model, in an attempt to better reflect the relationship between catchment evapotranspiration and vegetation cover. The model is designed to be parsimonious and plausible for quantifying hydrological impacts of vegetation change. The model has been tested in 122 catchments across the Murray Darling Basin (MDB), with remotely sensed leaf area index (LAI) from GIMMS3g and climate inputs from the SILO gridded dataset. Results show that the model performs reasonably well in most catchments (with NSE>0.5 for 95% of the catchments). The model performance is comparable to the original GR4J for most tested catchments but is notably better for 20% of the studied catchments (Figure 1a). The results indicate that remotely sensed LAI can help improve hydrological modelling, particularly by better reflecting the impact of vegetation dynamics on evapotranspiration. However, uncertainty exists in the remotely sensed LAI, which in some cases could affect model performance negatively.