The Influence Of Different Spatial Resolution And Algorithm Selection On The SWIM Hydrological Model

In the context of climate change, the extreme climate events, like floods and droughts, occur frequently. The runoff plays a decisive role in the extreme flood events as the carrier and foundation of floods. However, owing to the special environment and over investment of the establishment of hydrological stations, there are less hydrological stations than meteorological stations. Therefore, in the research of frequency and intensity of floods and climate change in the future, the missing runoff data are added on the hydrological models as the fundament of estimated climate data in the future. At present, although the main distributed hydrological model can achieve the quantitative simulation of hydrological data, the simulation effect is often not ideal because of the specificity of the runoff.The study area of the research is Changtaiguan basin which is located in the beginning of Huahe River basin. The data concludes the meteorological data and Underlying Surface data with different map scales, and the meteorological data are the daily observed data of ten encryption stations covered the period from 1980 to 1990.Based on these, we improve the methods which are the drawbacks of SWIM distributed hydrological model, and build the model input data information database that are based on the Underlying Surface data with different map scales and spatial resolutions. The database concludes 56 kinds of Underlying Surface data matrix, and they are all inputted into SWIM. According to the Nash efficiency coefficients, the data with best map scale and spatial resolution will be confirmed.The content of the research is concluded as followed:(1)Improve the D8 algorithm. The D8 algorithm deliberately fills the depression grids which are existed in the DEM data, so the results of doing this is going to lead to create some unreasonable river network such as parallel ones. Therefore, the S-D8 algorithm which is improved from the D8 algorithm cancels the step of filling depression grid, instead of handing them one by one with new methods. At the meanwhile, expanding the calculation range(from 3*3 range to 5*5 range), improving distance weight gap formulation, using the way that cope with depression region from the margin to the centre. The aims of these steps are establish ing the new flow analysis method. After improving D8 algorithm, the S-D8 algorithm can avoid the drawbacks of D8 and keep the advantages of D8, restoring the real river network with more correctly.(2)Confirm the best drainage area threshold. In the process of SWIM working, the drainage area threshold determines the density of river network, and it will let SWIM create different subbasin information which is an important parameter of the Model. By now, the main method confirms the threshold is constant threshold method. It depends on the individual’s experience to conduct qualitative analysis. In the research, three quantitative methods and one improved method based on that are introduced, they are average slope method, drainage network density, fitness index method and improved fitness index method. The calculation range of drainage area threshold is : 30km2、60 km2、90 km2、150 km2、200 km2、250 km2、300 km2、400 km2、500 km2、600 km2、700 km2、800 km2、900 km2、1000 km2. According to build the relation curve between indexes and different drainage area thresholds, the best of them should be confirmed. Besides, based on the spot investigation into the Changtaiguan basin, we quantitative measure the real channel width of the mainstream and tributaries for improving the fitness index method. At last, according to the four quantitative threshold methods, the best drainage area threshold is confirmed, and it is the 200km2.(3)Create Underlying Surface database with different map scales and spatial resolution. Based on the land use data with the map scales 1: 100000 m and 1:4000000m, the soil data with 1:1000000m and 1:4000000m and the DEM data, we divide the spatial resolution range is : 30m、60m、90m、120m、150m、200m、300m、400m、500m、600m、700m、800m、900m、1000m. With the different map scales and spatial resolution, the 56 data matrix is formed, and putting them all into the SWIM Model to simulate the runoff. After parameters setting, we use Nash efficiency coefficients to evaluate the effect of simulation. Finally, after the comparison and analysis of Nash and the curve between Nash and spatial resolution, the best data matrix is confirmed, and they are land use data with map scale of 1:100000m and the spatial resolution of 120 m, the soil data with map scale of 1:1000000m and the spatial resolution of 120, the DEM data with the resolution of 120 m. The SWIM Model simulates the best runoff data with these Underlying Surface data.The innovative ideas of this research are that: Improving the basin characteristic information, the older one is extensively used in the main distributed hydrological model, Arc GIS software and Hydrological analysis tools. And we quantitative confirm the best drainage area threshold and the Underlying Surface data with best map scale and spatial resolution. The purpose of this research is to improve the simulation process of hydrological model with more correct and refine, and obtain the simulation results with more realistic, making the scientific data base and algorithm support for supplementing the loss hydrological data and conducting the runoff forecast and intensity and frequency analysis of extreme flood events.