Research On Landscape Dynamincs And Effects On Runoff In Yiluo River Basin

In recent years, landscape dynamics and theirs environmental effects had aroused great attention. Under the impacts of landscape dynamics and climate changes, hydrological environment, hydrological processes and ecological systems would also have significant changes. Study on the relationships between landscape dynamics and hydrological changes were important to regional water resources security, environmental and ecological issues. It would also be great valuble to undersatand the characteristics of water and the rule of runoff producing and to improve the research on the safety of water resources protection, sustainable use of water resources.In this study, the paper chose the upper reaches of Yihe River as the case study area. Based on the field investigation, the using of remote sensing and geographic information systems technology, the paper analysed landscape dynamics since the 1980s. Various natural factors and human social factors of landscape dynamics were also analysed, and the integrated geographic environmental factors on landscape dynamics were examined. Then response of runoff to landscape dynamics was discussed. The main outcomes were as follows:(1)since the 1980s, watershed landscape dynamics were significant. In the study area, woodland landscape was the dominant landscape type. During the whole period, the area of woodland, pond and construction landscape types showed increasing, while the area of the grassland, river and unused land landscape types showed decreasing. The area of arable land at first increases, then reduced.(2)The forest landscape had the largest Aggregation Index and the Largest Patch Index. With the area of forest increasing, Aggregation Index increased and Landscape Shape Index and Perimeters-Area Fractal Dimension Index decreased. With the development of society, grassland and unused land landscape also tended to regular, and Aggregation Index increasing, Landscape Shape Index and Perimeters-Area Fractal Dimension Index decreasing. The distribution of mountain agriculture and construction land landscape was scattered, and their Landscape Shape Index were relatively large. Landscape Shape Index of cultivated land at first increases, then reduced. Construction land landscape tended to form large patches, more regular shape, resulting an increasing Aggregation Index and a decreasing Perimeters-Area Fractal Dimension.(3)Analysing the driving mechanism of landscape dynamics showed that: landscape dynamics were largely explained by geographical environmental factors (e.g. elevation, slope, and distance to residential center distance, distance to landscape edge, distance to road and distance to rivers) and human social factors (e.g. population, economic and policy factors ). Landscape dynamics had a close relationship with the geographical environment factors.(4)By establishing model, the integrated impacts of the elevation, slope, and distance to residential center distance, distance to road and distance to rivers on landscape dynamics were analysed. The results showed that: in the whole geographic environmental factors distribution index ranges, landscape dynamics and geographical factors index showed a significant correlation. And the integrative geographic environmental factors index was valuable in describing the landscape dynamics characteristics in the study area.(5)Based on the GIS technology, field investigation, distributed hydrological model SWAT(the Soil and Water Asseccment Tool) was constructed for the research watershed. Spatial database and attribute database were established. With the support of SWAT model, the 21 sub-watersheds and 98 Hydrological Response Units were automatically delineated. And the model were calibrated and validated by adusting model runoff sensitive parameters with the real runoff data of Dongwan hydrological station. In the yearly and monthly scale, the model can accurately simulated runoff changes. In the annual scale, Nash-Suttcliffe (Ens) coefficient were 0.89 and 0.87 respectively within calibration period and validation period, relative coefficient (R2) were 0.923 and 0.899 respectively. While in monthly scale, Nash-Suttcliffe (Ens) coefficient were 0.72 and 0.64 respectively within calibration period and validation period, relative coefficient (R2) were 0.801 and 0.669 respectively. The resulted showed that SWAT model could be used in this area.(6)Since the 1980s, in the annual scale, landscape dynamics made the average runoff decreasing with differernt reduction degree at different stages. In sub-basin scale, runoff increased with the cultivated area increasing, and runoff decreased with the woodland area increasing. From the 1980s to 1990s, climate change made the annual average runoff decreased, while climate change made the annual average runoff increased after 2000. All these had a close relattion with the low annual average precipitation in the 1990s and increased annual average precipitation after 2000.