Land Use Optimal Allocation Based On Land Degradation Risk Of Menglian County

Land resource is an important carrier of the human survival and development, also a kind of non-renewable resource. Since this century, unreasonable land use, leading to severe degradation of land resources in our country, has become an important limiting factor of regional social and economic development. How to improve the land use ways, allocate each land using type reasonably, improve the efficiency of land use is of great significance to solve the problem of land degradation, and have become one of the highlights in agro-scientific research in the current regional sustainable development.Menglian, a mountainous county in southwest frontier of Yunnan Province, recently, due to economic interests, rubber, tea, coffee and eucalyptus rapidly develop on a large scale, replaced the cultivated land and natural forest and other land types, which may cause a series of land ecological problems. Therefore, selecting Menglian county as the research object, evaluating land degradation risk and implementing the target of land degradation risk minimization of optimal allocation of land use research has strong representative, and is also an urgent realistic need, which can provide a reference for the land use planning of Menglian county.Choosing Menglian County as the research object, adopting remote sensing image, social and economic data, land use data and other data as data source, land degradation risk evaluation index system was built and spatial distribution characteristics of Menglian was analyzed based on the method of literature analysis.Then, based on RS and GIS technology, using grey linear programming model to realize land use quantity structure optimization aimed at land degradation risk minimization;What’s more, optimum projects must be based on cultivated land protection and food security, so we have set up3sets of solutions under two situations, namely the cultivated land "policy oriented" and "demand-oriented", eventually adopting CLUE-S model to allocate the selected optimal quantity structure project to the space.The main conclusions are as followings:(1) Based on the method of literature analysis, six indexes, namely, vegetation cover, soil organic matter, slope, soil moisture content, Soil K value(representative of soil texture) and annual rainfall were selected as land degradation risk evaluation index system of Menglian county, and five levels of land degradation risk of Menglian county were divided. The results indicates that:①Level IV is the largest type of land degradation risk level, accounting for30%of total area of the study area; secondly is level Ⅲ, taking proportion of26.36%of the total area; While, Level I took the smallest proportion of the five levels, only accounting for7.32%of total area of the study area.②As a whole, forest land, grassplot and natural unused land are at the degradation risk of level Ⅲ, while cultivated land, construction land and garden plot are at the degradation risk of level Ⅳ. Overall, the land degradation risk situation of Menglian County should be come into notice.(2) Based on land degradation risk assessment, he results of land use quantity structure optimization calculated with gray liner programming model indicated that:The comprehensive degree of land use degradation risk of three optimization projects dropped and land use rate were enhanced comparing with current situation and The Overall Plan for Land Utilization of Menglian County.Project3under the "policy oriented" situation was chosen as the preferred quantity structure of CLUE-S model due to the superiority in the sustainability of cultivated land compared with Project1and Project2under the "demand-oriented" situation, and the improvement in economic, ecological and social benefits as well as land use rate comparing with The Overall Plan for Land Utilization of Menglian County.(3) The results of land use spatial allocation based on CLUE-S model indicated that:①The results of ROC(Receiver Operating Characteristic) test aimed at Logistic regression showed that the cultivated land, garden land, forest land, construction land and waters, nature unused land of ROC values were0.752,0.857,0.783,0.922,0.688and1.0respectively. On the whole class the Logistic regression results between land use types and their driving factors were quite well. It indicated that it’s appropriate for this study to choose8natural driving factors, namely, slope, aspect, elevation, annual rainfall, soil organic matter, the distance to the main road, distance to rivers and distance to the polygon water, and3human-related driving factors, namely, distance to the rural residential areas, distance to the villages and towns and distance to the county seat, to analyze the suitable probability of land use types spatial distribution with Logistic regression method.②The spatial optimal allocation results of Project3shows that:As to total areas of the whole county, error rate between the actual areas allocated by CLUE-S model and the set up areas is only-0.1%. As to each land use types, the error rate is quite small, even construction land, which occupied the biggest error rate, is only1.12%. This suggests that look from quantity allocation the simulation precision is quite high. The main transfer of land use mainly showed a trend of garden land and natural unused land transferring to cultivated land, forest land and construction land, while the transfer strength and characteristics between villages and towns performed somewhat difference.