基于马尔柯夫模型的青海共和盆地土地沙漠化预测

在遥感和地理信息系统技术支持下，以青海共和盆地干旱草原1996和2006年的TM和ETM影像为基础，结合野外路线考察，对研究区近10年来的沙漠化发展趋势进行了监测。利用马尔柯夫模型，步长为1年，预测了2007－2016年共和盆地不同程度沙漠化面积的变化。研究结果表明，1996到2006年，研究区沙漠化总面积略有减少，但中度和严重沙漠化面积有所增加。在整个预测期沙漠化面积变化明显，其中轻度和重度沙漠化面积将有所减少，而中度和严重沙漠化面积将有所增加。这说明研究区沙漠化行势依然严峻，有必要采取措施防止沙漠化程度的进一步加深。

Application of the Markov process model on the dynamic change of land desertification in the Gonghe Basin, Qinghai province

Based on the 1996 and 2006 TM and ETM images of arid and desert grassland in the Gonghe Basin and the field investigations there, supported by RS and GIS methods, this paper monitored desertification trend from 1996 to 2006, and predicted yearly desertification statue from 2007 to 2016 using Markov process model. The results showed that in the last ten years the desertification had an decreasing trend, in which the whole desertified area decreased from 7565.15 km2 to 7419.60 km2, but moderate and very serious desertification increased from 1071.28 km2 and 858.78 km2 to 1308.987 km2 and 903.07 km2, while slight and serious desertification reduced from 4933.26 km2 and 701.82 km2 to 4656.68 km2 and 550.87 km2 respectively from 1996 to 2006. The prediction indicated that the areas of slight and serious desertification would be decreased, while the moderate and very serious desertification would have a increasing trend. By the end of 2016, the very serious desertification would be 928.65 km2, or increased 24.89 km2 compared with 2006. This pointed out that desertification in Gonghe Basin would be still serious, and some countermeasures must be taken to control and lighten desertification.