Use of Land Surface and Routing Models with Satellite Altimetry to Evaluate Risks to Hydropower during Drought

Hydropower provides a significant percentage of power to national grids in many regions of the world but drought, seasonal hydrologic fluctuations, and reservoir operations can jointly affect the ability of existing infrastructure to deliver power. This research uses several configurations of the Land Information System (Kumar et al 2008) to retrospectively model surface flows in the Upper Tigris watershed from 2004-2015. We compare streamflow outputs of HyMAP (Getirana et al 2012) routing model runs from GDAS-forced CLSMF2.5 and Noah3.3 land surface models at 0.125-deg and 0.25-deg spatial resolutions, and evaluate the degree to which satellite altimetry at the reservoir providing hydropower in this region can be utilized to select appropriate model configurations. From the perspective of an operational water user, we then assess the extent to which the combined models and altimetry data can improve our understanding of drought-related risks to hydropower operations in the area. We also evaluate several candidate drought and seasonality statistics appropriate to the limited time series of forcing data used in this study. Results show that while it is potentially reasonable to use these models to bracket a range of reservoir inflows in the early part of the water year, the uncertainty in cumulative inflows grows towards the end of the water year, with yearly differences in cumulative outflows ranging from 1-2 times the typical dynamic range of the reservoir. Therefore, although the models appear more reliable at the time of year when hydropower is most at risk due to low water levels, the model results diverge most during the periods of highest electricity demand. The analysis indicates that historically low water levels seen at the reservoir in January 2011 that temporarily stopped hydropower operations were consistent with natural water availability rather than water control changes upstream of the reservoir. The methods explored in this study constitute a step toward bringing integrated land surface modeling techniques at regional-to-global scales from the research to the operational realm.