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.
Accurate estimation of the soil water balance of the soil-plant-atmosphere system is key to determining the availability of water resources and their optimal management. Evapotranspiration and leaching are the main sinks of water from the system affecting soil water status and hence crop yield. The accuracy of soil water content and evapotranspiration simulations affects crop yield simulations as well. DSSAT is a suite of field‐scale, process‐based crop models to simulate crop growth and development. A “tipping bucket” water balance approach is currently used in DSSAT for soil hydrologic and water redistribution processes. By comparison, HYDRUS-1D is a hydrological model to simulate water flow in soils using numerical solutions of the Richards equation, but its approach to crop-related process modeling is rather limited. Both DSSAT and HYDRUS-1D have been widely used and tested in their separate areas of use. The objectives of our study were: (1) to couple HYDRUS-1D with DSSAT to simulate soil water dynamics, crop growth and yield, (2) to evaluate the coupled model using field experimental datasets distributed with DSSAT for different environments, and (3) to compare HYDRUS-1D simulations with those of the tipping bucket approach using the same datasets. Modularity in the software design of both DSSAT and HYDRUS-1D made it easy to couple the two models. The pairing provided the DSSAT interface an ability to use both the tipping bucket and HYDRUS-1D simulation approaches. The two approaches were evaluated in terms of their ability to estimate the soil water balance, especially soil water contents and evapotranspiration rates. Values of the d index for volumetric water contents were 0.9 and 0.8 for the original and coupled models, respectively. Comparisons of simulations for the pod mass for four soybean and four peanut treatments showed relatively high d index values for both models (0.94–0.99).