Performance of quantitative precipitation estimates in the complex terrain of the Peruvian Andes. IMERG V06 and the development of user-driven downstream applications.

Quantitative precipitation estimates obtained from satellite data are of critical importance to research and applications. Not only is precipitation a key component of important water and energy cycles, but the immediate societal benefits offered by reliable products are undeniable.

The complex terrain of the Peruvian Andes creates significant challenges to precipitation retrievals from space and to the establishment of dense ground monitoring networks. Nonetheless, for the communities and authorities of the region (home to nearly one third of the Peruvian population), this information is vital.

The performance of the quasi-global Integrated Multi-Satellite Retrievals for Global Precipitation Measurement (IMERG) V06 was assessed in Peru as part of Project e-Andes. Data covering the period between 2011 and 2020 were compared against gauge data from 35 stations maintained by SENAMHI. The gauges are located in the Andes and arid Pacific coast. Mean Pearson correlation values ranged from 0.34 (Early, Daily), to 0.80 (Final, Monthly), showing a clear improvement with temporal aggregation and from Early to Final runs. The trend was also observed across other metrics including Bias, RMSE, and MAE.

 Challenges to the validation of IMERG Final in sparsely gauged regions is also discussed. The study was an important component of capacity-building efforts and the development of user networks. The information provides important guidance for the development of monitoring services that incorporate both IMERG and gauge networks to create estimates with reduced bias.