Quantitative Monitoring Of Soil Erosion Based On3S Technology In Granite Red Soil Region Of Western Fujian Province

Soil erosion is a complex land surface process driven by the interaction of natural and human factors. The accelerated erosion can undermine surface soil, soil fertility and organic carbon storage, and in turn affecting human survival and sustainable social development. China is one of the most severely eroded countries, and has been fomed typical eroded regions:Loess Plateau, Three Gorges reservoir region and Southern hilly area.Therefore, to carry out the multi-directions and multi-angles monitoring work is particularly significant. Which can both helpful in ascertain the status of changes in region soil erosion, and also provide important theoretical basis and decision making reference for comprehensive management and ecological restoration construction. The3S technology can meet the requirements of timely, convenience and accuracy attributed to its advantages in dealing with spatial-temporal data.This paper, supported by sub-project of11th Five-Year National Science and Technology Program:Study and Demostration of Key Technologies for Controling and Remediation of Red Soil Degradation and Developing High Quality Ecological Agriculture (2009BADC6B), selected a typical eroded hilly region in granite distributed area in Southeastern China:Changting County, Fujian Province. By using3S technology, the paper studied the change of soil erosion based on two concepts:erosion intensity and erosion degree, and at two spatial scales:the macroscopic sacle based on RS image and the microcosmic scale based on measured samples’soil properties and hyperspectral characteristics. Finally, the concept of soil erosion extent was proposed to characterize the comprehensive hazard of soil erosion, which could provide the effective theoretical basis for decision making of erosion controlling management. The main results and conclusions were as follows:(1) The optimized integration model based on GeoCA can simulate spatial and temporal change of soil erosion better than erosion intensity assessment algorithm and traditional GeoCA model:The traditional GeoCA was optimized to simulate spatial and temporal change of soil erosion by integrating the algorithm of soil erosion, cellular states transition functions and neighborhood transition rules. The accuracy of optimized GeoCA can reached70%. The result showed that controlling soil erosion had been improved in the past30years, mainly reflected in the increased non-apparent eroded area and decreased extreme eroded area. Currently, three different erosion intensity zones was identified, including slightly eroded area in nouthwest, moderate eroded area in southeast and extreme eroded area in central. The highest erosion intensity is existed in central of Changting County. The imporving tendency for erosion controlling was slow down since2000, indicting that erosion intensity still limited the social and economic development.(2) The different erosion degree soil can be distinguished with the minimum data set of hyperspectral inversion including soil organic matter (SOM), complex iron (Fep) and non-crystalline Al (Alo):In extreme eroded area in central of Changting county,35soil samples had been collected according to residue status of soil genetic layers.Results showed that there was a similarity in morphology of spectral between different soil erosion degree, while obvious differences in soil properties(SOM, Fep, Alo et al), which resulted in that the overall reflectivity and characteristics of specific absorption peak differed in various erosion degree soil.Therefore, the minimum data set of hyperspectral inversion was built including SOM, Fep, Alo was identified, the ersion degree soil should be distinguished based on inversion effect and curve shapes of correlation coefficient between minimum data and corresponding hyperspectral reflectivity. Result showed that the inversion effect of moderate eroded soil samples was significantly lower than the other2eroison degree soil samples, and according to curve shapes of correlation coefficient between minimum data(SOM, Fep, Alo) and corresponding hyperspectral reflectivity, the non-apparent and exereme eroded soil samples can be identified effectively.(3) Soil erosion extent can effectively characterize the comprehensive hazard of soil erosion which was the integration of soil erosion, erosion degree and land risk response:The concept of soil erosion extent integrates the concept of erosion intensity and erosion degree, which can characterize the comprehensive hazard of soil erosion for erosion evaluation based on GIS. The key to spatial representation is the scaling of erosion degree from sampling point to region. However, the result is not satisfactory when the model was created directly between RS image and hyperspectral curves. The spatial distribution of soil erosion was fuzzy categorized by using fuzzy adjacent algorithm based on the inversion results of spectral and image of soil properties (SOM, Fep, Alo).Result showed that the accuracy can reached71.4%, which indicated that the spatial representation used by fuzzy adjacent algorithm was feasible. On the other hand, characteristics ofconcentration and contiguous is existed in erosion extent distribution, the total area of moderate and slight eroded soil were significantly higher than the severe or extreme eroded soil, but in spatial, they existed mainly in the surrounding area, not entried into the center of county, which indicated a imporving tendency of erosion controlling, however the task of erosioncontrolling can’t be ingored.(4)3S technology can promote the research of spatial decision for erosion controlling. In this paper, the research system of soil erosion extent partition-erosion controlling mode partition-eroison controlling meatures partition was proposed in this paper. The erosion extent partition was the basis of controlling mode partition and controlling meatures partition, the controlling mode shoulde be divided into two classes:economic priority and ecological priority, and the diagnose creteria standard was the composition of transportion, environment and economic development. And the controlling measures including conservation of forest and grass, watershed management, farmland protection and bio-engineering practices. This paper conducted the controlling mode partition and eroison controlling meatures partition based on3S technology. The result showed that planted broadleaf and dredging was the basic measures, and various meatures covered the entire area of the center of county and distributed discretely in spatial.To sum up, the entire paper formed the complete research system of ’erosion intensity monitoring-erosion degree identify-erosion extent represented in spatial-erosion controlling decision ’with the support of GIS, mathematical statistics, hyperspectral analysis of soil properties and non-linear mathematical models (cellular automata, fuzzy mathematics). The research provided good theoretical and pratical basis for further study of soil erosion and3S technology.