Climate data for parameterisation of lakes in Numerical Weather Prediction models

In Numerical Weather Prediction (NWP), to initialise lake variables for parameterisation of lakes, lake climatology is required. We obtained model lake climatology through offline runs of a lake model FLake (Freshwater Lake model) with different values of lake depth parameter. As a result, a global dataset with the resolution of 1° was developed. To project lake climatology onto a particular NWP model grid, data are extracted from the dataset depending on the lake depth value provided on the target grid. To prevent drifting of the bottom temperature in warm deep lakes, relaxation to the long-term mean air temperature was applied in FLake. Lake model climatology was validated against observations for different types of lakes. We suppose that detected errors for boreal lakes in spring are connected with inaccuracies in forcing data, but further study of errors is needed.