A Priori Parameter Estimation for Distributed Models and for Predictions in Ungauged Basins (PUB) II Posters
Presiding: Q Duan, Lawrence Livermore National Laboratory; T Wagener, Pennsylvania State University
H51B-01 0830h
Split-Sample Tests for Vegetation-Type Dependence of Parameters of the Land Dynamics (LaD) Model
Poorly constrained vegetation parameters (a stomatal resistance parameter and a plant-root parameter) of the Land Dynamics (LaD) model for global land water and energy balance are estimated by model calibration. Very fast simulated annealing is used to minimize selected measures of the deviation between modeled and measured streamflow over a set of streamgauges (the calibration gauges). Estimated parameters are then transferred for use in other gauged basins that were excluded from the optimization (evaluation gauges). This calibration-evaluation process is performed in two ways: with parameters assumed to be globally constant, and with parameters assumed to depend on vegetation type. Performance of this parameter-transfer procedure is evaluated by computing, for the evaluation gauges, the value of the error measure that had been used in the calibration. Does stratification of the optimization by vegetation type yield predictive power? Comparison of the calibration and evaluation error measures for the global and vegetation-stratified approaches may help to answer that question.
H51B-02 0830h
The Importance of Drainage Network Definition in the Determination of the Basin Response.
The identification of channel initiation points is central to geomorphology and hydrology. The Geomorphologic Instantaneous Unit Hydrograph (GIUH) concept is extensively used to describe surface water transport phenomena in small catchments. Recent works pointed out the relative importance, in specific environments, of three different sources of dispersion in characterizing the hydrologic response function through the GIUH application: geomorphologic, kinematic and hydrodynamic. The present work shows that, accounting for a correct representation of the geomorphologic dispersion and of the kinematic dispersion, only due to the differences between hillslope and channel routing process, is sufficient to determining the main features of the hydrologic response in small steep catchments. An objective procedure for the calibration of the geomorphologic filter used to distinguish between hillslope and channel paths is applied to identify the drainage network for several catchments. Different case studies are presented.
H51B-03 0830h
Using Dual-Region Calibration to Improve Recharge and Hydraulic Conductivity Estimates for Hydrologic Modeling
High-resolution regional groundwater flow models commonly have large uncertainties as a result of poorly calibrated parameters. One source of these uncertainties is the inherently non-unique nature of recharge and hydraulic conductivity parameters estimated using either stream fluxes or hydraulic head as the primary data set. Elevated recharge can have the equivalent effect of a decrease in conductivity on the simulated hydraulic heads. Likewise, a predominately flux-calibrated model can have very low sensitivity to conductivity parameters. Thus simply calibrating a transient model to either head or stream flux measurements is inadequate. This study presents a novel solution to these problems by demonstrating that the linked calibration of two separate regional groundwater flow models can produce a parameter set superior to those obtained from either individual model calibration alone. The two regions are geological and climatologically similar, and each region was calibrated with predominantly one data type. The two modeled regions for this study are the Grand Traverse Bay watershed (GTBW) and the Muskegon River watershed (MRW) in northern lower-Michigan. Within the GTBW, a network of 16 pressure transducers was used to monitor hydraulic head continuously for almost two years. In the MRW, approximately 150 base flow stream discharges were collected over the same time period, yielding a dense set of flux calibration targets. A simultaneous, parallel, parameter optimization method is presented, and the results are compared to separate parameter optimizations of each individual model. The overall effectiveness of the technique is evaluated by comparing optimized parameters to available estimates of recharge and conductivity from other studies, demonstrating that the dual-model calibration method can improve our ability to jointly estimate recharge and hydraulic conductivity for regional aquifers.
H51B-04 0830h
Area-slope Threshold Value Assessment for Drainage Network Recognition in the Framework of Rainfall Runoff Modelling
The estimate of model's parameters is a central feature for the application of rainfall runoff models. The applicability of a GIUH-based network response model is conditioned by the identification of the proper drainage network. Among the various filtering criteria that can be found in the literature for channel recognition from digital elevation models, the one using contributing area slope shows interesting features. Nevertheless, the area-slope criterion has been poorly applied, mainly because of the difficulties in objectively defining the appropriate threshold values. In this work, a structured approach to assess the area-slope threshold value is proposed within the hydrologic applications perspective. The resulting channel network is then used as input to a semi-distributed, event-based rainfall-runoff model able to describe severe rainfall events in small, steep basins. This model accounts for network and hillslope contributions to the total dispersion in the routing process, a key factor in determining the main features of the hydrologic response. In a geomorphologically homogeneous region, the set of model parameters shows interesting invariance properties with respect to storm and basin characteristics.
H51B-05 0830h
A simulation approach for investigating the hydrologic signature on the probability distribution of floods
In the framework of flood prediction, theoretically derived distributions represent a promising tool that still posses the capability for greater development. Nevertheless, a wider investigation on the main control of hydrological processes acting at the basin scale is needed, but at the same time extremely complex to be assessed using experimental approaches. The only way is probably the hydrological simulations via distributed models. With this aim, a continuous simulation scheme is developed to reproduce the basin behavior through the time. The scheme uses a distributed hydrological model (DREAM) in cascade with a rainfall generator (IRP) reproducing in a consistent way the flood peak CDFs (Cumulative Distribution Functions). The numerical simulation allows the reconstruction of large number of extreme events and the derivation of interesting evidence on the main control for flood generation mechanisms highlighting the relevance of soil texture and morphology different climatic environments. The proposed methodology has been applied to the Agri and the Bradano, in the region of Basilicata, Italy.