Which PET formulas for my rainfall-runoff modelling?

The combined effect of evaporation and transpiration plays a key role in the water balance for hydrological modelling at catchment scale.
However, the direct measurement of these processes is very challenging, and a formula is typically used to estimate the potential evapo-transpiration (PET), i.e. the maximum rate of water leaving the catchment to the atmosphere in ideal conditions.
Many rainfall-runoff models take the PET as input, and improving the quality of the PET data can directly enhance the performance of the model.
Moreover, PET is a crucial factor to characterize the hydrological behavior and to find similarities between basins.

Many PET formulas have been proposed and several of them are commonly used with excellent results, but the choice is up to the single researcher, that each time must decide based on the input data available and their personal preferences.

Here we analyze the differences between formulas in the estimated values, obtained for a large set of Caravan catchments (Kratzert et al., 2023), in order to give insights on which PET formulas could be more suited for the use in rainfall-runoff modelling.

We selected a group of PET formulas among the most used in literature, with diverse types of methods and required inputs.
The data to feed the formulas were taken from global datasets, derived from reanalysis products: the availability in such sets of temperature, radiation, pressure, humidity and wind allows us to include in the study the FAO Penman-Monteith formula, that we used as benchmark. Additionally, we selected some temperature and/or radiation-based formulas, which represent important tools for data-scarce applications and large-scale hydrology.

For every catchment we calculated the daily time series of PET for each formula, then we analyzed and compared the aggregated yearly and seasonal mean values.
We illustrated the main differences and distribution variations between catchments at local and global scale, highlighting climatological patterns and how the choice of the PET formula affects the catchment aridity classification in the Budyko curve.

References:

Kratzert, F., Nearing, G., Addor, N., Erickson, T., Gauch, M., Gilon, O., Gudmundsson, L., Hassidim, A., Klotz, D., Nevo, S., Shalev, G., and Matias, Y.: Caravan – A global community dataset for large-sample hydrology. Sci Data 10, 61 (2023). https://doi.org/10.1038/s41597-023-01975-w