The Added Value of Different Data Types for Calibrating and Testing a Hydrologic Model in a Small Catchment

Abstract

This study investigated the added value of different data for calibrating a runoff model for small basins. The analysis was performed in the 66 ha Hydrological Open Air Laboratory, in Austria. An Hydrologiska Byråns Vattenbalansavdelning (HBV) type, spatially lumped hydrologic model was parameterized following two approaches. First, the model was calibrated using only runoff data. Second, a step-by-step approach was followed, where the modules of the model (snow, soil moisture, and runoff generation) were calibrated using measurements of runoff and model state variables and output fluxes. These measurements comprised laser-based measurements of precipitation, satellite and camera observations of snow, ultrasonic measurements of snow depth, eddy covariance measurements of evapotranspiration, time domain transmissometry-based soil moisture measurements, time-lapse photography of overland flow, and groundwater level measurements by piezometers. The two model parameterizations were evaluated on annual, seasonal, and daily time scales, in terms of how well they simulated snow, soil moisture, evapotranspiration, overland flow, storage change in the saturated zone, and runoff. Using the proposed step-by-step approach, the relative runoff volume errors in the calibration and validation periods were 0.00 and −0.01, the monthly Pearson correlation coefficients were 0.92 and 0.82, and the daily logarithmic Nash Sutcliffe efficiencies were 0.59 and 0.18, respectively. By using different sources of data besides runoff, the overall process consistency improved, compared to the case when only runoff was used for calibration. Soil moisture and evapotranspiration observations had the largest influence on simulated runoff, while the parameterization of the snow and runoff generation modules had a smaller influence.

Multi-Step Calibration Approach for SWAT Model Using Soil Moisture and Crop Yields in a Small Agricultural Catchment

“…Since the curve number (CN2) was the most sensitive parameter for streamflow, these results suggest that adjusting CN2 value is sufficient to obtain a good flow calibration for the HOAL catchment. These results are similar to Széles et al [36] who applied the Hydrologiska Byråns Vattenbalansavdelning (HBV) model [81] in the HOAL catchment. They noted that the model was able to produce very good streamflow results after runoff calibration, whereby their model obtained a logarithmic NSE value of 0.81.…”

“…Highquality soil data used during model set-up in our study can, therefore, be attributed to the good soil moisture simulation and, therefore, the negligible impact of soil moisture calibration on streamflow. Széles et al [36] also did not report any significant improvement in streamflow prediction for the HOAL catchment after incorporating soil moisture data in their research; however, they reported an improved soil moisture prediction by the model. Rajib et al [34] used in-situ soil moisture estimates for top 60 cm of soil profile in their multi-objective approach and reported an improvement of R 2 from 0.69 to 0.72 in streamflow calibration after soil moisture calibration.…”

“…The basin has an area of 66 hectares. The elevation ranges from 257 m to 323 m above sea level with a mean slope of 8% [36]. Agricultural land is the primary land use in the area occupying 87% of the total watershed [37].…”

Sensitivity of the agro-hydrological model CRITERIA-1D to the Leaf Area Index parameter

“…These models, however, are very sensitive to many input and parameters [6] and their applicability at large scales can be limited by the availability of input data and parameters and their accuracy [7]. For these reasons, several studies aimed to improve model performances by integrating new observations [8], calibrating the model in comparison to independent data and based on different objective functions [9] or by data assimilation techniques [10].…”

“…Nowadays many satellite products are in fact free available in almost real time (within a few hours from the acquisition time), and they can be quickly integrated for improving model prediction [12]. In this context, it becomes very important to assess which input and parameters might be more relevant for improving the performance of the model and to prioritize data collection and model integration [8], [13], [14].…”