Historical and future water and energy balances of Lake Tsomoriri, a high altitude lake in the Indian Himalaya, were assessed using a combination of remote sensing and modeling to better understand the lake's historical hydrology and the potential impact due to climate change. The model, which included watershed hydrology to determine inflow and a hydrodynamic water and energy balance model (CE-QUAL-W2) was constructed and calibrated using data from remote sensing sources and temperature data from a nearby meteorological station. Tropical Rainfall Monitoring Mission (TRMM) precipitation data and Surface Radiation Budget (SRB) shortwave radiation and cloud cover data were used for model inputs. Landsat Multispectral Scanner, Thematic Mapper and Enhanced Thematic Mapper Plus, and Terra Moderate Resolution Imaging Spectroradiometer (MODIS) imagery were used to assess historical changes in lake size and freeze/thaw dates for use as calibration data in the model. Historical imagery showed that lake area was dynamic, and fluctuated during the study period (1973-2009). Despite increases in winter temperature, the length of the frozen season did not change over 2001-2010. The modeled water levels increased with increased precipitation, and the water balance was highly sensitive to wind speed. Model runs with default input parameters resulted in rapid, sustained increases in lake levels, which were not observed in the historical imagery. Only model runs that included loss to groundwater or high wind speeds maintained stable water levels in the lake. Increased air temperature caused increased evaporation and reduced modeled frozen season length. Increases in watershed inputs due to increased glacial melt in the Tsomoriri watershed were minimal compared to overall inflow, due to low glacial volume in the watershed. Model results could not be replicated using the longer term and lower spatial resolution Climate Research Unit (CRU) meteorological data as model input, highlighting uncertainty in the main model inputs and the importance of in-situ data. A potential future scenario was modeled using a portion of the CRU data set that represented a stable lake scenario and a temperature increase based on an ensemble of eight general circulation models (GCM) compiled by the Intergovernmental Panel on Climate Change (IPCC). The ensemble average temperature increase of 0.055°C per year decreased water levels by 14 meters by 2085 and decreased frozen season lengths at the simulated lake from 100 to 62 days on average. The model was used to identify key variables that control the water balance. Inflow, air temperature, precipitation, and wind speed were identified as crucial unknown input variables that change model output significantly. The modeling suggested a strategy for field measurements that can most efficiently constrain the remaining uncertainty
