Benchmarking convection-permitting climate simulations for hydrological applications: A comparative study of WRF-SAAG and observation-based products

Abstract. Over the last years, significant progress has been made in the development of convection-permitting climate models (CPCMs), especially for improving precipitation modeling in regions with complex terrain. Recently, the South American Affinity Group (SAAG) developed a novel high-resolution dataset — hereafter referred to as the WRF-SAAG dataset — by dynamically downscaling the ERA5 reanalysis using the Weather Research and Forecasting (WRF) model over South America for the period 2000–2021. In this paper, we evaluate the quality of WRF-SAAG daily precipitation and temperature simulations using observations from meteorological stations over continental Chile for the period 2001–2018, and present comparisons against two gridded meteorological products – CR2MET and RF-MEP – which are based on in-situ meteorological station measurements and have been widely used for hydrometeorological applications in this region. We found that, although the precipitation products correctly replicated the percentage correct (PC) of observed events and non-events (PC ≥ 0.64), detection accuracy varied within each Chilean macrozone –defined by latitudinal bands – with worse performance in the Far North (between 17.5 – 26° S) and Patagonia (between 43.7 – 56° S) — median Critical Success Index (CSI) < 0.49 for events > 5 mm/d— compared to the central region (CSI ≥ 0.44 for events > 5 mm/d). The evaluation of daily precipitation and extreme temperatures against station observations using Tang’s Kling-Gupta efficiency (KGE_T) and its components reveals that all datasets performed better in reproducing precipitation in rainy regions (median KGE_T ≥ 0.65 in the Southern macrozone), while in arid areas such as the Near North during summer, the median KGE_T was negative. The CR2MET product consistently provided the best performance metrics for extreme precipitation and temperature, partly because it includes information from the stations used for evaluation. Finally, the application of the TUW hydrological model shows that WRF-SAAG simulations achieved runoff estimations comparable to the best observation-based products, with the best metrics obtained in the Southern macrozone, where the median objective function (OF) —defined as the average of KGE' and KGE' (1/q) — remains above 0.87 (0.67) during the calibration (evaluation) period. More broadly, the results presented here show that – despite some remaining challenges in arid climate regions – kilometer-scale climate models can deliver information of a quality comparable to that of observation-based products for hydrological applications in Chile.