The CAMELS-CL dataset: catchment attributes and meteorology for large sample studies – Chile dataset

Abstract

Abstract. We introduce the first catchment data set for large sample studies in Chile (South America). The data set includes 516 catchments and provides catchment boundaries, daily streamflow records and basin-averaged time series of the following hydrometeorological variables: 1) daily precipitation retrieved from four gridded sources; 2) daily maximum, minimum and mean temperature; 3) daily potential evapotranspiration (PET); 4) 8-day accumulated PET; and 5) daily snow water equivalent. In addition to the hydro-meteorological time series, we use diverse data sets to extract key landscape attributes characterizing climatic, hydrological, topographic, geological and land cover features. We also describe the degree of anthropic intervention within the catchments by relying on publicly available water rights data for the country. The information is synthetized in 64 catchment attributes describing the landscape and water use characteristics of each catchment. To facilitate the use of the dataset presented here and promote common standards in large-sample studies, we computed most catchment attributes introduced by Addor et al., (2017) in their Catchment Attributes and MEteorology for Large-sample Studies dataset (CAMELS dataset) created for the United States, and proposed several others. Following this nomenclature, we named our dataset CAMELS-CL, which stands for CAMELS dataset in Chile. Based on the constructed dataset, we analysed the main spatial patterns of catchment attributes and the relationships between them. In general, the topographic attributes were explained by the Andes Cordillera; climatic attributes revealed the basic features of Chilean climate; and hydrological signatures revealed the leading patterns of catchment hydrologic responses, resulting from complex, non-linear process interactions across a range of spatiotemporal scales, enhanced by heterogeneities in topography, soils, vegetation, geology and other landscape properties. Further, we analysed human influence in catchment behaviour by relating hydrological signatures with a novel human intervention attribute. Our findings reveal that larger human intervention results in decreased annual flows, runoff ratios, decreased elasticity of runoff with respect to precipitation, and decreased flashiness of runoff, especially in drier catchments. CAMELS-CL provides unprecedented information in South America, a continent largely underrepresented in large-sample studies. The proximity of the Andes means that this dataset includes high-elevation catchments, which are generally poorly represented world-wide due to data-scarcity. The CAMELS-CL dataset can be used to address a myriad of applications, including catchment classification and regionalization studies, the modelling of water availability under different management scenarios, the characterisation of drought history and projections, and the exploration of climate change impacts on hydrological processes. This effort is part of an international initiative to create a multi-national large sample data sets freely available for the community.