Simulation Of Land Use Change And Analysis Of Landscape Ecological Effects In The Mainstream Of Shaxi River Based On The CLUE-S Model

Based on the remote sensing images of 1992,2001 and 2013 in the Mainstream of Shaxi River of Fujian, the paper discussed the characteristics of land use change, using the CLUE-S model, the land use in 2001 and 1992 was simulated based on 2001, and land use was simulated in 2013 based on 2001, and then the land use situation in 2025 under different situations were simulated based on 2013. Finally, based on the historical and simulated datas, the regional landscape ecological effects were evaluated, which were of great significance to the future land use planning and landscape ecological construction, and the main conclusions were as follows:(1)The number of forest land, farmland, grassland and unused land decreased, while the garden, residence, traffic land increased, the water remained stable, and residential and industrial and mining land showed an increase by 85.61% and 398.05%.The main transfer flows were the transfers among forest land, farmland, garden, grassland and unused land, and the forest land, farmland, and garden land mainly transfered to residence and industrial and mining land. The main transfer within the 6000 buffers from forest to garden at earlier period turned to transfer from forest to industrial and mining land at later period, especially from 1000m to 3000m.With the increase of buffer the transfer flows increased, but the amount of transfer flow increasing amount and increasing amplitude decreased, the buffer from 2000m to 3000m was an inflection point.(2) The single dynamic degree of land use from large to small from 1992 to 2013 was as follows:the land of mining and transportation, the residence land, the unused land, the garden land, the grass land, the farm land, the forest land and the water. The overall dynamic degree of earlier period was more than later period, and the overall land use dynamic degree reduced with the time scale extending to 21 years.The areas with high value of comprehensive index of land use degree expanded to the edge, the expansion of the city and county built were the most obvious, while low value area shrinked, the overall level of land use degree increased.(3) Driving factors and the various land use types had good correlation, especially terrain factors and urbanization rates. Through scale screening, it was concluded that the ROC values of 100* 100m were higher, so 100m* 100m was suitable for the simulation. With the simulation scale from 9 to 12 years, the Kappa coefficient showed from 0.93 to 0.89, when the simulation scale expanded to 21 years, Kappa coefficient decreased down to 0.86, it showed that the CLUE-S model simulation has significant time scale effect. In addition to other land (grassland and unused land), the accuracy of the orther land use types were more than 0.75, the CLUE-S model simulation of the Mainstream of Shaxi River had a good effect.(4) In the 4 situations of this research, compared with 2013, the number of forest land, farmland and other land has decreased, the area of residence, industrial and mining land and garden showed an increasing trend. Under the natural growth scenario, the garden showed the largest increase by 27.69%. In the economic development scenario, the farmland declined by 22.29%, residence and industrial and mining transport land increased espectively by 54.45% and 76.14%. In the case of farmland protection scenario, the farmland showed the lowest decline only by 1.75%. In the integrated development scenario, the forest land and other land showed the largest decline respectively by 5.9% and 13.35%.(5)From 1992 to 2013, the degree of landscape fragmentation, landscape shape complexity, landscape diversity and decentralization increased, while the landscape connectivity decreased, the landscape ecological effect index of land use change increased. To 2025, high value of landscape ecological effects areas will constantly expand, median and low value areas will constantly shrink, which indicates the disturbance degree of human activities on the landscape pattern will increase continuously. In the situations of 2025, the landscape was seriously influenced by humanity under comprehensive development and economic development scenarios, human disturbance of farmland protection scenario was greater than that of natural growth scenario. The increase of landscape heterogeneity and homogenization will lead to the decrease of the internal stability of landscape. With the increase of population, the ecological environment vulnerability will increase, while the regional ecological risks will finally increase.