Modeling terrestrial water balance using regionalization of hydrological model parameters at Kalli watershed, Estonia.

Terrestrial water storage and evapotranspiration (ET) have strong dependence on vegetation type at the catchment scale, and they also play an important role in the forest management and carbon exchange. Regionalization methods dependent on rainfall-runoff models comprise techniques for transferring calibrated parameters in instrumented basins (donor basins) to non-instrumented basins (target basins). In this study, regionalization was used to transfer of parameters from the GR4J-Cemaneige model calibrated in Reola hydrographic basin to predict daily flows in Kalli basin, both watersheds located in the south-east of Estonia. The basins are covered with hemi-boreal forests and we used the hypothesis that both basins have physiographic and climate similarity, besides being close to each other. ET historical data was collected from the Station for Measuring Ecosystem-Atmosphere Relations (SMEAR Estonia), located in Järvselja forest in Kalli basin. Precipitation data was collected from Tartu-Tõravere and SMEAR stations, and the river flow from Reola hydrometric station. The time series period adopted for model calibration was 2015-2018 and the 2019 year was used for validation. Point precipitation measurements are generally used to calculate average precipitations at the watershed scale. However, several studies point out that the use of rain gauges can generate errors in undermeasurement of precipitation, both for snow (solid precipitation) and rain (liquid precipitation), mainly due to the effect of wind. Water Balance Method (WBM) was used to correct the precipitation from Reola Watershead and then GR4J-Cemaneige model was calibrated at Reola Basin, with Nash-Sutcliffe Efficiency index of 0.77. The precipitation series from SMEAR at Kalli watershed was corrected to consider the solid precipitation, using data from Tartu-Tõravere station. For modeling the Kalli watershed streamflow an iterative process was created and it consists of: i) run the hydrological model and obtain the streamflow, ii) apply the WBM correction on the precipitation data, iii) repeat step (i) until the correction factor stabilizes to a fixed value. This methodology allowed to obtain the terrestrial water storage daily variation from the beginning of 2015 to the end of 2019 with high reliability, since the watershed is treated as a system. The results indicate that the terrestrial water storage is positive in winter and negative in the summer period, reaching absolute maximum values of 200 mm.  This research allows a better understanding of the Järvselja hemi-boreal forest water dynamics and its variation through the seasons, especially in the context of climate change. Next steps are to analyze the relations between carbon, energy and water balance at stand and catchment scales.