Tutorial for merging satellite-based precipitation datasets with ground observations using RFmerge (>=0.3-0)

Le résumé fourni par la source

About This vignette updates an earlier tutorial developed for RFmerge version 0.1-6 and previous releases, which relied on the now-superseded `raster` package. The current workflow uses the terra package and demonstrates how RFmerge can be used to create an improved gridded precipitation product by combining satellite-based precipitation estimates with in situ observations and physically meaningful covariates, including elevation and distances to rain gauge stations. We use the Valparaiso Region in central Chile as a case study. The example shows how to generate a daily merged precipitation product at 0.05° spatial resolution for January-August 1983. The workflow requires the following inputs: i) daily rainfall time series from rain gauges, ii) metadata describing the spatial coordinates and identifiers of the rain gauges, iii) the Climate Hazards Group InfraRed Precipitation with Station data version 2.0 (CHIRPSv2), iv) the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks - Climate Data Record (PERSIANN-CDR), and v) the Shuttle Radar Topography Mission version 4 (SRTM-v4) digital elevation model (DEM). In addition, distances from rain gauges to grid cells can be used as covariates. When requested, these distances are computed internally by `RFmerge` from each rain gauge station to every grid cell within the study area. Citation If you find this tutorial useful, please cite it as Zambrano-Bigiarini et al. (2026): Zambrano-Bigiarini, M.; Baez-Villanueva, O.M.; Giraldo-Osorio, J.D. (2026). Merging satellite-based precipitation datasets with ground observations using RFmerge (>=0.3-0). doi:10.5281/zenodo.20061103. The theoretical basis of the merging algorithm is described in the following *Remote Sensing of Environment* article: Baez-Villanueva, O. M.; Zambrano-Bigiarini, M.; Beck, H.; McNamara, I.; Ribbe, L.; Nauditt, A.; Birkel, C.; Verbist, K.; Giraldo-Osorio, J.D.; Thinh, N.X. (2020). [RF-MEP: a novel Random Forest method for merging gridded precipitation products and ground-based measurements](https://doi.org/10.1016/j.rse.2019.111606), Remote Sensing of Environment, 239, 111610. doi:10.1016/j.rse.2019.111606. Please also cite the `RFmerge` R package: Zambrano-Bigiarini, M.; Baez-Villanueva, O.M., Giraldo-Osorio, J. (2026). RFmerge: Merging of Satellite Datasets with Ground Observations using Random Forests. R package version 0.3-3. URL: https://hzambran.github.io/RFmerge/. doi:10.32614/CRAN.package.RFmerge. Required `RFmerge` version This tutorial was developed for `RFmerge >= 0.3-0`, the first release series in which `RFmerge` is based on the terra package. All previous `RFmerge` versions up to 0.1-6 were based on the superseded raster package. `RFmerge` 0.1-6 was removed from CRAN on 2023-02-25 because issues related to retired spatial dependencies, such as `rgdal`, could not be resolved in time.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.