Abstract
Systems exposed to hydroclimatic variability, such as the integrated electric system in Uruguay, increasingly require real-time multiscale information to optimize management. Monitoring of the precipitation field is key to inform the future hydroelectric energy availability. We present an operational implementation of an algorithm that merges satellite precipitation estimates with rain gauge data, based on a 3-step technique: (i) Regression of station data on the satellite estimate using a Generalized Linear Model; (ii) Interpolation of the regression residuals at station locations to the entire grid using Ordinary Kriging and (iii) Application of a rain/no rain mask. The operational implementation follows five steps: (i) Data download and daily accumulation; (ii) Data quality control; (iii) Merging technique; (iv) Hydrological modeling and (v) Electricity-system simulation. The hydrological modeling is carried with the GR4J rainfall-runoff model applied to 17 sub-catchments of the G. Terra basin with routing up to the reservoir. The implementation became operational at the Electricity Market Administration (ADME) on June 2020. The performance of the merged precipitation estimate was evaluated through comparison with an independent, dense and uniformly distributed rain gauge network using several relevant statistics. Further validation is presented comparing the simulated inflow to the estimate derived from a reservoir mass budget. Results confirm that the estimation that incorporates the satellite information in addition to the surface observations has a higher performance than the one that only uses rain gauge data, both in the rainfall statistical evaluation and hydrological simulation.
Highlights
The renewable contribution of the electric energy matrix in Uruguay has been increasing steadily during the last decades, with hydroelectric, wind and solar components that have different inherent variability and predictability
It uses a deterministic model to estimate a value of the variable by using actual ground measurements to calibrate a model for the satellite estimates and refines the estimate analyzing the residuals for spatial correlation; it combines the statistical fitting and deterministic modeling [33]
The merged precipitation field obtained is used in a hydrological modeling of the Rio Negro basin whose output is, in turn, coupled with an electric system modeling that guides planning and dispatch decisions for the following seven days
Summary
The renewable contribution of the electric energy matrix in Uruguay has been increasing steadily during the last decades, with hydroelectric, wind and solar components that have different inherent variability and predictability. This poses both a challenge and an opportunity to optimize planning at different embedded timescales and, dispatch. In most applications that use operational hydrological models, the spatial and temporal variability of precipitation constitutes one of the dominant factors and with greater associated uncertainty In this context, remote sensing products are ideal instruments to be used in real-time hydrological modeling, embedded in operational systems for flood warnings, drought monitoring and water resource management [4,5]. With high spatial and temporal resolution, are currently available in near-real-time (NRT): the Climate Prediction Center (CPC) MORPHing algorithm (CMORPH) [6], the CPC Quick MORPHing technique (QMORPH) [7], the Tropical
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
