Abstract
Spatial variability plays an important role in nonlinear hydrologic processes. Due to the limitation of computational efficiency and data resolution, subgrid variability is usually assumed to be uniform for most grid-based rainfall-runoff models, which leads to the scale-dependence of model performances. In this paper, the scale effect on the Grid-Xinanjiang model was examined. The bias of the estimation of precipitation, runoff, evapotranspiration and soil moisture at the different grid scales, along with the scale-dependence of the effective parameters, highlights the importance of well representing the subgrid variability. This paper presents a subgrid parameterization method to incorporate the subgrid variability of the soil storage capacity, which is a key variable that controls runoff generation and partitioning in the Grid-Xinanjiang model. In light of the similar spatial pattern and physical basis, the soil storage capacity is correlated with the topographic index, whose spatial distribution can more readily be measured. A beta distribution is introduced to represent the spatial distribution of the soil storage capacity within the grid. The results derived from the Yanduhe Basin show that the proposed subgrid parameterization method can effectively correct the watershed soil storage capacity curve. Compared to the original Grid-Xinanjiang model, the model performances are quite consistent at the different grid scales when the subgrid variability is incorporated. This subgrid parameterization method reduces the recalibration necessity when the Digital Elevation Model (DEM) resolution is changed. Moreover, it improves the potential for the application of the distributed model in the ungauged basin.
Highlights
For the last few decades, the development of numerous distributed rainfall-runoff models enables the spatial variations to be represented by a network of grid elements
The rest of paper is organized as follows: we provide a description of the study area; a brief description of the Grid-Xinanjiang model and the subgrid parameterization method are presented in the model description section; the following section includes a set of numerical experiments conducted in the Yanduhe Basin to compare the model performance with and without incorporating the subgrid variability; and the final section provides conclusions and perspectives
The watershed is assumed to drain from a full saturation status without rainfall and evapotranspiration, until the watershed soil moisture content reaches the condition that we choose
Summary
For the last few decades, the development of numerous distributed rainfall-runoff models enables the spatial variations to be represented by a network of grid elements. Advances in geographic information systems (GIS), remote sensing (RS) and computational technology have offered the potential to build complex distributed hydrologic models and improve the accuracy of hydrologic prediction in time and space [1,2,3]. Limited by the resolution of available data and the computational efficiency, most grid-based distributed models do not take into account the subgrid variability of model input, parameters and model state [7,8,9]. With the assumption of the uniform subgrid, the high frequency information of hydrologic variables and parameters will be lost as the large sampling dimensions act as a filter [10,11]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
