| Among all global environmental changes, Land Cover Change is the most significant one in the areas of combined effects of natural and human activities. The impact of Land Cover Change on surface hydrological processes is a hot research topic at present. The resolution and classification of land cover data have been one of the main factors of the hydrological model accuracy. The rapid development of RS and GIS, which has promoted the development of Hydrological Sciences and played an important role in the hydrological study, has a scientific significance. This paper puts forward some new ideas and methods based upon:the current research focus of hydrology, integrated exploratory study of the impact of Land Cover Change on the hydrological process simulation and the regional surface water resources assessment. Thus this paper has a strong academic exploration and scientific significance.Taking the example of the Case Study of Northern Jiangsu Province, the paper studies the effect of the resolution of land cover data on regional surface water resources evaluation; Taking Qinhuai Watershed as an example, it researches the impact of the classification of land cover data on hydrological simulation process. In time scale, it studies the impact of different periods, different resolution land cover data on hydrological process simulation in Taihu Lake. The main conclusions are as follows:(1)Variable characteristics of land cover dataWith the decrease of remote sensing image resolution, the area changes of land cover are mainly shown in paddy field and dryland. When the paddy field area increases, the dryland area will decrease and the area of water surface and urban will change in different extent. The increasing of urban area shows the impact of urbanization. It is apparent that with the urban area increasing, the paddy area decreases remarkably and that with the Drought woodland increasing, the water surface area decreases. In addition, the speed of urbanization seems faster in some places.(2)The impact of the resolution of surface coverage remote sensing classification data on the surface water resources assessment.The resolution of Land cover data has a great impact on water resources estimates of water resources regionalization, with the change in the proportion ranging from0.7%to105.2%. Especially in dry years (P=95%).The results of the runoff in the two resolutions have a significant difference. The differences of the multiple water conservancy partitions are more than50%and the difference between normal water years (P=50%) and general drought years (P=75%) is less than20%.Comparing the simulated results of the surface water resources amount of the land cover based on different resolution remote sensing image with the years average amount of surface water resources results in Jiangsu Province Water Resources Assessment, the simulated results based on10m resolution have higher degree of fitting. The results caused by different resolution of drought years and special drought years are coincident. Besides, based on the same resolution, the difference between stimulated results in normal years and yearly mean results is the smallest. The second is that of general dry years, the fitting of special dry year is the poorest.Under the condition of the same assurance, with the surface coverage data remote sensing image resolution decreases, the simulated surface water resources seems to be smaller; The main production flow change is that the paddy field and urban runoff deceases with the decrease of dry land and surface runoff. Among them, the increase of runoff in paddy field and the decrease of dry land runoff are more significant. The rainfall influences production flow a lot. The statistical change of underlying surface based on resolution is the main factor of runoff generation. Rainfall and land cover data affect surface water resources amount significantly.(3)The impact of the classification of surface coverage remote sensing data on hydrological simulation process.Runoff simulation results in different classification system show that the feature classification system of remote sensing images affects hydrological modeling less than the resolution remote sensing image. The influence degree of the feature classification system is diverse in the different position of the sub-basin. The differences change most obviously in the10th sub-basin, while no significant differences in more than half of the sub-basin. According to day simulation results of many years, with the change of classification system, the flow change of the outlet section is small in two-thirds of year and is more than5%in one-tenth of year, especially the biggest difference to31.6m3/s, which are amount to12.64%of the outlet section flow.Secondly, the influence degree of regional runoff caused by the feature classification system is positively correlated with hydrological cycle. From comparative study of feature classification divided into four categories, five categories, and twelve categories, difference of simulation results of the feature type divided into four categories is lager than that divided into five categories and twelve categories. In addition, the results are significant differences caused by different classification system with the changes in the classification. For the tenth sub-basins of the study area case, monthly average flow in the condition of feature type divided into twelve categories in most cases are larger than that divided into five categories, with the range of0.06%to0.34%in2008. In2005, monthly average flow in the condition of feature type divided into twelve categories compare with that divided into five categories, both have50%increase and decrease in a month, with the range of0.07to0.1%. And compared monthly flow in the condition of feature type divided into five categories compare with that divided into four categories, monthly average flow change ratio is from0.07%to0.1%, with the range of0.14%to1.41%.(4) The effect of different resolution land cover data on the result of hydrological simulationThe changes of land cover data come from different resolution remote sensing images and have a significant effect on watershed level. It appeared a tendency of the higher water level with the higher classification resolution. The higher the remote sensing images resolution is, the greater difference of the simulative water level will be. If the differences of resolution increase with the rainfall, there will be a total superposition effect.In the same hydrological and meteorological conditions, water level of the basin increases with the increase of the RS image resolution. The difference of30m resolution RS image results and statistical results is biggest. In the high flow year, the maximum difference is up to the8.7%of the normal water level, such as Changzhou station. The difference of water level caused by the land coverage data come from10m resolution remote sensing images and that of300m is less than it, up to7.43%of the normal such as Changzhou station in high flow year. Next is the difference of water level caused by the land coverage data come from30m resolution remote sensing images and that of300m, up to5.49%of the normal such as Wuxi station in high flow year. On condition of the same resolution change, the water level difference in high flow years is bigger than that in other normal years. Therefore, in order to be closer to the real situation, the simulation results should be adjusted when we do some related researches.Land cover change and rainfall both have an effect on watershed hydrological process. The amount of rainfall is the main influence factor for the overall level of water level process. On the overall trend, the whole water level is the highest in high flow year (P=20%), higher in normal flow year (P=50%) and the lowest in dry years (P=90%). The water level fluctuation is the biggest in high flow year, and the water level is higher than the other two represent years, especially in the flood season. The water level fluctuation is much less in normal flow years, and the water level is lower than that in dry years. The water level in dry year is the lowest, while it is higher than the other two represent years in fewer periods. Thus, in addition to the effect of rainfall, other factors also have certain effect on water level process. In the condition of the high flow year, firstly, hydrological response is more significant than that in dry year. Secondly, the changes of land cover data have an effect on water process. It shows the trend that the water level increases with the change features of land cover data with the development of city. The fluctuations in each station are different, which is approximately from0.33%to6.45%.This paper first discusses the impact of remote sensing image resolution of land cover data from share on surface water resources evaluation and surface hydrological process simulation, and then makes up gap in the fields of research, and suggests new ways for surface coverage data to be better used in hydrological research. It studies the impact of classification system of different land cover data change on hydrological simulation accuracy and gives some main indexes that will be applied in features classification standard of hydrologic remote sensing image. It is based on the hydrological situation with different assurance rate and hydrological process simulation of land cover change in different period. In addition, the paper respectively discusses the rainfall and surface coverage change on watershed hydrological process, putting forward new research idea and method for watershed hydrological elements. It is an exploratory study with high scientific meaning. |
PDF Full Text Request