农业工程学报2017，Vol.33Issue(10)：271-277,封4,8.DOI:10.11975/j.issn.1002-6819.2017.10.036
沙化土地整治背景下的沙漠化监测预警
Desertification monitoring and early warning in context of sandy land consolidation
摘要
Abstract
In order to ensure China's food and ecological security, it's particularly necessary to perform desertification monitoring and early warning during the land consolidation projects execution. The study area was located in Kezuohouqi which is famous for the sandy land. Based on Landsat 8 OLI image data, desertification monitoring was achieved by the decision tree classification method. The desertification early warning system was constructed by PSR (pressure-state-response) model; the natural and human influencing factors of desertification were regarded as "pressure", the desertification monitoring results were regarded as "state", and the land consolidation projects were regarded as "response". The weight of each factor in the PSR model was set by expert scoring method, and DESI (desertification ecological security index) was adopted in this paper; the greater the index, the lower the risks of desertification. Based on the DESI, the results of desertification early warning were divided into 4 different levels through natural breakpoint method, namely, no desertification warning, slight desertification warning, medium desertification warning and severe desertification warning. In the end, the validity of land consolidation could be evaluated through calculating the area of different desertification early warning results in the land consolidation projects and conducting the comparison. The results showed that: 1) The area of desertification accounted for 88% of study area. To be exact, the area of medium and severe desertification accounted for 61% of the total area, and the extremely severe desertification was mainly concentrated in the western part of the study area whose area accounted for 6% of the total area. 2) The zonal differentiation of desertification early warning results was clear; the spatial distribution from east to west in Kezuohouqi in turn was no desertification warning, medium desertification warning, slight desertification warning, medium desertification warning, severe desertification warning, and then medium desertification warning. The area of medium desertification warning was the largest, and the distribution laws of severe warning desertification and extremely severe desertification were similar, but the severe warning desertification tended to gather in the northwest pastoral areas of Kezuohouqi, the slight warning and no warning areas were located in the eastern part of Kezuohouqi, and the land use pattern was mainly agricultural land. 3) In the land consolidation projects, high quality capital farmland construction and arable land consolidation were mainly distributed in the areas with no desertification warning. The potential land suitable for agriculture and the project of land reclamation were mainly distributed in slight desertification warning area and medium desertification warning area. There was the sever desertification warning land among the farmland consolidation, the potential land resources suitable for agriculture, and forage grass base construction projects, whose distribution style was sporadic. However, the severe desertification warning area in high quality capital farmland was not up to the mark and should be eliminated from the projects. Overall, in the context of land consolidation, the risk of further desertification is small. The severe desertification warning area would be a key point for land consolidation projects, and the government must put a high premium on the sever warning area. The results of desertification early warning in this paper can provide the reference for land consolidation and adjustment.关键词
土地利用/整治/遥感/沙漠化监测预警/决策树分类/PSR模型Key words
land use/consolidation/remote sensing/desertification monitoring and early warning/decision tree classification/PSR model分类
农业科技引用本文复制引用
韩已文,刘忠,于学谦,王佳莹..沙化土地整治背景下的沙漠化监测预警[J].农业工程学报,2017,33(10):271-277,封4,8.基金项目
国土资源部公益性行业科研专项(201411009) （201411009）
