Energy Balance and Hydrological Modelling of Zongo Glacier, Bolivia, Using ERA-40 Reanalysis Data

In the Andes several regions profit significantly from glacial melt water for drink water supply and electricity production. During the dry season, glacier melt is significant source of water in the semi-arid region of La Paz, Bolivia. The Andean glaciers are retreating and water resources after reaching a culmination, will decrease. This implicates serious environmental and socio-economical consequences. For an effective attenuation, it is crucial to furnish quantitative predictions of the glacier mass loss and its effects on the water resources in these regions. A distributed energy balance model has been developed to model mass balance and melt induced discharge of tropical glaciers. We want to predict the changes in glacier melt discharge in response to future climate change for the region of La Paz, Bolivia and later regionalize the model to a larger area. The model operates on daily steps, has a 20 m grid resolution, and is forced by daily data of air temperature, humidity, wind speed, global radiation and precipitation. As a test basin, we calibrate the model at Glaciar Zongo, Bolivia, 16°15'S , 68°°10'W which is monitored by the French Institute for Research for the Development (IRD) . Zongo Glacier is a 1,8 km2 large and the catchment is 63% glacierized. Mass balance, weather station and discharge data are available on daily basis from 1991 onward. The measurements have gaps and only two years (1994-95 and 1999-00) with continuous data are available. In order to allow for multi-year simulations we force the model by daily ERA-40 reanalysis data from the European Center for Weather Forecast (ECMWF). To downscale the data we compare the daily data 1991-2002 to the observations at the glacier. Results indicate a fair agreement for air temperature, but a rather poor correlation between the ERA-40 data and the observations for wind speed, global radiation and precipitation. The correlation is improved using monthly values. So far, test runs of the model using downscaled ERA-40 data for the Zongo Glacier show good agreement for the mass balances (r2=0.88) and relatively good estimate of the monthly melt discharge (r2=0.74). The transition between the wet and dry season is well captured by the model