Hydrological Evaluation of the LPX Dynamic Global Vegetation Model for Small River Catchments in the UK

Assessments of water resources using macro-scale models tend to be conducted at the continental or large-basin scale. However, security of freshwater supplies is a local issue, and thus necessitates study at such a scale. This research aims to evaluate the suitability of the Land Processes and eXchanges Dynamic Global Vegetation Model (LPX-DGVM, a development of the LPJ model) for simulating runoff for small catchments in the UK. Simulated runoff is compared against National River Flow Archive streamflow and precipitation observations from 12 catchments of varying size and climate regime. Annual and seasonal variability are analysed to evaluate model performance. Results show that LPX replicates observed inter- and intra-annual runoff variability successfully in terms of both flow timings and magnitudes. Interannual variability in flow timings are reproduced particularly well, whereas runoff magnitudes are slightly overestimated for the majority of catchments. These overestimations are shown to be largely accounted for by water extractions in populated catchments. Absence of river routing and storage from the model, in addition to precipitation errors, have also been identified as contributing to the runoff discrepancies. Seasonal variability in runoff is also reproduced successfully by LPX. Unlike in previous global evaluation studies of the model, flow timings correlate well with observations and there is no tendency towards early estimation of extremes. Runoff magnitudes are, however, generally slightly overestimated for winter and underestimated for summer months. LPX simulates the frequency of low flow months well, particularly for summer months. Monthly high flow frequencies tend to be underestimated, although are well captured for winter months. Overall, the results show that LPX-DGVM can successfully simulate runoff processes at the scale of small catchments in the UK. This study offers promising insights for the prospects of using dynamic global vegetation models for simulating future water resources at the local scale.