Answers are grounded in this paper and cite their sources.
In this study, we present an updated version of PISCOp (Peruvian Interpolated data of SENAMHI’s Climatological and Hydrological Observations), which consists of a gridded precipitation dataset covering the entire geographical area of Peru. The analysis period spans from 1981 to 2025, with a daily temporal resolution and a spatial resolution of 0.1° (≈10 km). The merging algorithm is based on geostatistical and deterministic interpolation methods integrating three different rainfall sources: (i) the observed rain network quality-controlled, infilled and homogenized dataset, (ii) radar-gauge merged precipitation climatologies and (iii) the Climate Hazards Group Infrared Precipitation (CHIRP V3.0) estimates. This latest version improves on the previous one by increasing the number of observed stations, expanding coverage of transboundary basins, refining quality-control processes, adding a machine learning-based imputation method, and including a process to homogenize the completed series. Based on the results of spatial autocorrelation, teleconnection identification, hydrological modeling, and independent validation, we suggest that PISCOp V3.0 estimates exhibit moderate to high performance and outperform other regional gridded rainfall datasets across the Amazon and Andes regions (only Andes ecoregion) at a monthly (daily) scale with an Refined Index of Agreement in the upper 0.72 (0.60). Furthermore, selecting bias correction and/or homogenization approaches for the completed series, based on the target station, reduced the formation of inhomogeneities in the final gridded product. Finally, the PISCOp V3.0 gridded dataset aims to support hydrological studies, water management, drought analysis and the monitoring of long-term hydrological events, in basins with available rain gauge stations.
No takes yet. Share an insight, caveat, or question.
Lope et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: