Simulation Of Land Use Changes Around Danjiangkou Reservoir In He’nan Province

The Danjiangkou Reservoir is the water source of the Middle Route Project(MRP)of the South-to-North Water Transfer project(SNWT)in China. The around Danjiangkou Reservoir Region in He’nan Province is its important water catchment.The regional land use changes have an important effect on water quality and water quantity of the Danjiangkou Reservoir. This paper takes catchment of Danjiangkou Reservoir in Henan province for example, which includes Lushi county, Luanchuan county, Xixia county, Xichuan county, Neixiang county and Dengzhou city. With the support of RS and GIS technologies, based on three remote sensing images of 1991,2002 and 2013, spatio-temporal changes of land use from 1991 to 2013 were analyzed, and a CA-Markov model about land use changes in research area was established to simulate and analyze the future land use pattern.The main conclusions in this paper are shown as below:Under the support of ENVI 4.8, ArcGIS 10.1 and IDRISI Selva 17.0 software,on this basis of interpreting three remote sensing images of 1991, 2002 and 2013 to obtain corresponding land use maps, the temporal and spatial change of land use from1991 to 2013 were analyzed from the perspective of structure, dynamic degree, the relative change rate of single land use type, transition matrices and the major land use changes types. The results showed forestland and farmland were the dominant land use types, farmland and other land decreased, while construction land and forestland increased. The main source of the increased construction land was farmland,while the main source of the increased forestland was other land.The relationship between the distribution of each of land use types and the driving factors was established by logistic regression analysis. The results showed that distance to the respective original and type and socioeconomic factors had greater impact on land use change, especially distance to the original class, distance to construction land, distance to farmland, population density and GDP.Logistic regression analysis between each of land use types and driving factors was used to generate the transition suitability maps, which were one of the rules ofCA. Using the transition suitability maps and transition matrices, a CA-Markov model was established to simulate spatial distribution of land use in 2024. The results showed that forestland and farmland are still dominant land use types in 2024.Compared to those during the period of 1991-2002 and 2002-2013, land use changes of the study area during the period of 2013-2024 keep the same trends, a continuing downward trend in farmland and other land, as well as an upward trend in construction land and forestland, water body will increase slightly.The research results provide a scientific basis and data support for sustainable land use and management and regional ecological conservation in this area.