REMOiso forcing for the iPILPS Phase 1 experiments and the performance of REMOiso in three domains

Land surface schemes (and their evaluation) require high-quality forcing at subdiurnal resolution. The paucity of such data for land surface scheme experiments is exacerbated in iPILPS (Isotopes in the Project for Intercomparison of Land Surface Parameterisation Schemes), where high-quality, high-resolution isotope forcing is also required. The REgional MOdel REMOiso has been used to provide meteorological and isotopic forcing for several years for three sites for the iPILPS Phase 1 experiment: Manaus (Brazil), Neuherberg (near Munich, Germany) and Tumbarumba (Australia). Thus, this paper is about how the forcing data for the iPILPS Phase 1 experiment were generated, and how the model-derived forcing compares with observational data. The comparison between monthly aggregations of the REMOiso simulations and observational data such as GNIP (Global Network of Isotopes in Precipitation) shows that REMOiso produces plausible results for these three regions and sites. An extended example of the strength and weaknesses of REMOiso is shown, by presenting the first application of REMOiso to the Australian domain. These simulations are compared with temperature, precipitation and isotopic data, collected by GNIP and the Australian Bureau of Meteorology, for seven stations in Australia. Observational sampling campaigns will provide meteorological and isotopic data to compare how REMOiso performs at subdiurnal resolution. The current forcing from REMOiso, however, provides a set of high-resolution data, physically consistent with each other, that can be additionally used by modellers wishing to test the performance of their own isotope-enabled land surface schemes.