Study On The Desertification Monitoring And Early-Warning In Xinjiang

Desertification is one of the most serious Land degradation problems in globaltoday. The premise to prevent and control desertification is monitoring andearly-warning desertification. However, due to the complex causes of desertification,the lack of observation data, and so far, the monitoring of desertification is not sufficient,and the early warning of desertification is just beginning, it needs to further research.This paper aimed at the key scientific problems of the monitoring and earlywarning desertification. Through using the basic theory of the monitoring and warning,utilizing the advanced technology of GIS, the paper constructed the monitoring andwarning model of desertification, and this model has be used in Xinjiang, as thelong-term monitoring and early-warning of desertification research. The majorconclusions obtained from the paper as follows:(1) Through sorting the connotation of desertification, summarizing and choosingthe monitoring indices of desertification, comparing the classification systems ofdesertification degree, and contrasting the early warning methods and models used indesertification, the paper produced the initial monitoring and early warning theorysystem of desertification.(2) The paper manufactured a set of automatic sand-moving monitoring instrument.The instrument can automatically gain the data of sand flux, wind speed, wind directionand temperature obtained from the instrument. It can remotely transmit the date fromfield to the monitoring center, and the data can be automatically saved to the data-baseat the monitoring center, and the data can be automatic analysis. Through the instrument,the sand activities of desertification realized the distance automatic monitoring.(3) By selecting six typical land types that is desert, fixed dune, gobi, grassland,forest, and cultivated land, this paper researched the MODIS NDVI data stability thatinfluences the monitoring desertification feasibility. The research results showed that the NDVI stabilities of low vegetation coverage were very high monthly and yearly, theNDVI stabilities of high vegetation coverage were shifty with the different year and thedifferent month, the primary data did not satisfy the requirement of monitoringdesertification changes. Then it composed the yearly biggest NDVI with the monthsNDVI from April to September, and testing the composed data stability, it founded thatthe data stability improved, it met the requirements of monitoring desertification.(4) It utilized the yearly biggest MODIS NDVI data composed by the monthsNDVI from April to September, and it established the flow sand land, the half flow sandland, the fixed sand land, the gobi land, and the saline-alkali land as the monitoringdesertification ranges. Using the binary inversion model to invert the vegetationcoverage, it constructed the annual conventional monitoring desertification mode, andrealized the annual monitoring desertification in Xinjiang, China.(5) By using the grey GM(1,1) prediction model and combining with rasteroperation function of GIS, the paper created a grid gray GM(1,1) prediction model.Used the grid gray GM(1,1) prediction model, it forecasted the key factor (NDVI) ofdesertification changes. The forecast results showed that the total model predictionaccuracy was within5%. Using the model, it forecasted the NDVI in2012,2013,2014,2015,2020, Xinjiang.(6) The study created the grid accumulated desertification early warning model.Used the model and the monitoring and predicting desertification results in2012,2013,2014,2015,2020, it published the desertification warning degrees in Xinjiang. Itfounded that the desertification warning degrees in southern Xinjiang was greater thanthe degree in northern Xinjiang. The largest alert of desertification was the area in themiddle reaches of Tarim River, especially in region from Yingbaza to Qala. In theDabancheng, Urumqi, in the northwest and southwest sides of the Ulungur Lake, in themiddle reaches of the Qarqan River, the north of CeLe county, the north of Yutiancounty, in the west of Qiemo county, in the west of Makit county, and in the west ofBachu county, all appeared different alerts of desertification. (7) Based on the perspective of human beings selves, considered the hazard factorof desertification and the output value of land, and integrated the grid accumulateddesertification early warning model, this study also founded the desertification loss riskearly warning model. The desertification loss risk early warning model was used toalarm the desertification in Xinjiang. The results showed that the warning determined bythe loss risks model of desertification more tended to human activities than the gridaccumulated desertification early warning model. The results were more suitable forprevention and control of desertification.(8) Used GIS technology, WEBGIS technology and internet technology, the studyconstructed the desertification monitoring and early warning system in Xinjiang. Itrealized the automatic monitoring and early warning of desertification in Xinjiang, theresults of the desertification monitoring and early warning was published in network. Itrealized the sharing research results.