The event runoff coefficient (Rc) and the recession coefficient (tc) are of theoretical importance for understanding catchment response and of practical importance in hydrological design. We analyse 57 event periods in the period 2013 to 2015 in the 66 ha Austrian Hydrological Open Air Laboratory (HOAL), where the seven subcatchments are stratified by runoff generation types into wetlands, tile drainage and natural drainage. Three machine learning algorithms (Random forest (RF), Gradient Boost Decision Tree (GBDT) and Support vector machine (SVM)) are used to estimate Rc and tc from 22 event based explanatory variables representing precipitation, soil moisture, groundwater level and season. The model performance of the SVM algorithm in estimating Rc and tc is generally higher than that of the other two methods, measured by the coefficient of determination R2, and the performance for Rc is higher than that for tc. The relative importance of the explanatory variables for the predictions, assessed by a heatmap, suggests that Rc of the tile drainage systems is more strongly controlled by the weather conditions than by the catchment state, while the opposite is true for natural drainage systems. Overall, model performance strongly depends on the runoff generation type.
Measuring evaporation and transpiration at the field scale is complicated due to the heterogeneity of the environment, with point measurements requiring upscaling and field measurements such as eddy covariance measuring only the evapotranspiration. During the summer of 2014 an eddy covariance device was used to measure the evapotranspiration of a growing maize field at the HOAL catchment. The stable isotope technique and a Lagrangian near field theory (LNF) were then utilized to partition the evapotranspiration into evaporation and transpiration, using the concentration and isotopic ratio of water vapour within the canopy. The stable isotope estimates of the daily averages of the fraction of evapotranspiration (Ft) ranged from 43.0–88.5%, with an average value of 67.5%, while with the LNF method, Ft was found to range from 52.3–91.5% with an average value of 73.5%. Two different parameterizations for the turbulent statistics were used, with both giving similar R2 values, 0.65 and 0.63 for the Raupach and Leuning parameterizations, with the Raupach version performing slightly better. The stable isotope method demonstrated itself to be a more robust method, returning larger amounts of useable data, however this is limited by the requirement of much more additional data.