CAMELS-FI: Large scale catchment attributes and hydrometeorological time series in Finland

Comprehensive, large sample hydrological datasets, such as CAMELS (Catchment Attributes and MEteorology for Large-sample Studies), have provided the basis for advances in many aspects of hydrological research in recent years. They can be utilized for several purposes, such as training data driven hydrological models, comparisons between regions dominated by different types of hydrological processes and testing of general validity of hydrological theories. The value of these datasets is in combining multitude of data sources into one, easily accessible and usable, harmonized and high quality package. We present CAMELS-FI, an extensive hydro-meteorological dataset for over 160 catchments, which adheres to the blueprint established by the previous CAMELS-datasets. It combines hydrological and meteorological time series with static catchment attributes in a format that enables comparisons between catchments within the dataset but also between different CAMELS-datasets.

CAMELS-FI provides up to 30 years of daily data, containing variables similar to previous CAMELS-datasets, such as streamflow observations, rainfall, temperature, evapotranspiration and snow. In addition, static attributes describing among others the catchment’s soil type, land use and topography are provided. The selected catchments are either not impacted or only marginally impacted by actively managed reservoirs, and have observations from at least five years. We also intend to compare the differences between the hydrological and meteorological signatures in different catchments, as well as compare the regional variability of soil, land cover and topography in order to give deeper insights on the properties of Finnish catchments.

We are planning to use CAMELS-FI to train and test a deep learning neural network to make river flow predictions in unagauged catchments more accurate in Finland.