eprintid: 50335 rev_number: 8 eprint_status: archive userid: 1482 importid: 105 dir: disk0/00/05/03/35 datestamp: 2025-02-12 08:53:30 lastmod: 2025-02-12 08:58:16 status_changed: 2025-02-12 08:53:30 type: article metadata_visibility: show creators_name: Doury, Antoine creators_name: Gadat, Sébastien creators_name: Somot, Samuel creators_idrefppn: 256748632 creators_idrefppn: 080889433 creators_idrefppn: 103969632 creators_halaffid: 1002422; 490594 creators_halaffid: 1002422;56663 creators_halaffid: 443875,490594 title: On the suitability of a Convolutional Neural Network based RCM-Emulator for fine spatio-temporal precipitation ispublished: pub subjects: subjects_ECO abstract: High resolution regional climate models (RCM) are necessary to capture local precipitation but are too expensive to fully explore the uncertainties associated with future projections. To resolve the large cost of RCMs, Doury et al. (2023) proposed a neural network based RCM-emulator for the near-surface temperature, at a daily and 12 km-resolution. It uses existing RCM simulations to learn the relationship between low-resolution predictors and high resolution surface variables. When trained the emulator can be applied to any low resolution simulation to produce ensembles of high resolution emulated simulations. This study assesses the suitability of applying the RCM-emulator for precipitation thanks to a novel asymmetric loss function to reproduce the entire precipitation distribution over any grid point. Under a perfect conditions framework, the resulting emulator shows striking ability to reproduce the RCM original series with an excellent spatio-temporal cor- relation. In particular, a very good behaviour is obtained for the two tails of the distribution, measured by the number of dry days and the 99th quantile. Moreover, it creates consistent precipitation objects even if the highest frequency details are missed. The emulator quality holds for all simulations of the same RCM, with any driving GCM, ensuring transferability of the tool to GCMs never downscaled by the RCM. A first showcase of downscaling GCM simulations showed that the RCM-emulator brings significant added-value with respect to the GCM as it produces the correct high resolution spatial nd heavy precipitation intensity. Nevertheless, further work is needed to establish a relevant evaluation framework GCM applications. date: 2024-07-26 date_type: published publisher: Springer id_number: 10.1007/s00382-024-07350-8 official_url: http://tse-fr.eu/pub/130256 faculty: tse divisions: tse keywords: Emulator keywords: Hybrid downscaling keywords: Regional climate modeling keywords: Statistical downscaling keywords: Precipitations keywords: Deep neural network keywords: Machine learning keywords: EURO-CORDEX keywords: CORDEX language: en has_fulltext: FALSE doi: 10.1007/s00382-024-07350-8 view_date_year: 2024 full_text_status: none publication: Climate Dynamics volume: vol.62 place_of_pub: Berlin pagerange: 8587-8613 refereed: TRUE issn: 0930-7575 oai_identifier: oai:tse-fr.eu:130256 harvester_local_overwrite: volume harvester_local_overwrite: date harvester_local_overwrite: issn harvester_local_overwrite: pending harvester_local_overwrite: creators_idrefppn harvester_local_overwrite: creators_halaffid harvester_local_overwrite: abstract harvester_local_overwrite: publisher harvester_local_overwrite: place_of_pub oai_lastmod: 2025-02-10T07:58:46Z oai_set: tse site: ut1 citation: Doury, Antoine , Gadat, Sébastien and Somot, Samuel (2024) On the suitability of a Convolutional Neural Network based RCM-Emulator for fine spatio-temporal precipitation. Climate Dynamics, vol.62. pp. 8587-8613.