CCD遥感数据研究马尾松毛虫灾害空间扩散规律

Application of CCD Data in Prediction of Masion Pine caterpillar Distribution and Damage

摘要: 以2009广西重大灾害为实例，以广西宾阳和横县交界的这个重要灾害区域为研究区，主要利用GIS的空间分析技术和CCD遥感数据，运用红边dVI₆₆₀和dVI₄₇₅参数、植被指数NDVI、RVI分析马尾松毛虫灾害范围和程度，探讨了松毛虫灾害的发生发展规律。结果表明：CCD数据的红边dVI₆₆₀和dVI₄₇₅参数对受灾区域较为敏感；比较多时相各指数受灾前后变化，其变化趋势一致，各指数值增加表明未受马尾松毛虫危害，而其下降则表明是受灾区域，其中以红边dVI₆₆₀值变化幅度最大；结合地面数据，初步确定了不同虫害程度的指数值，初步确定了不同受灾程度的面积，其中以红边dVI₆₆₀分析反演的马尾松毛虫灾害最适合实际灾害情况，为以后的研究提供了线索。本研究还证实了应用遥感可以尽可能早地发现灾害初始发生的论断。

Abstract: Based on the Masion Pine caterpillar outbreak data in 2009 near the border areas of Binyang and Hengxian counties in Guangxi, GIS spatial analysis technology and CCD remote sensing data of HJ were studied for the prediction of Masion Pine caterpillar distribution and damage. Various vegetation indexes of CCD data including dVI₆₆₀, dVI₄₇₅, NDVI, and RVI were examined. The results showed that dVI₆₆₀ and dVI₄₇₅ were sensitive to reflecting the damage areas of different degrees, so the developing rules of Masion Pine caterpillar hazard could be shown with different temporal point imageries. The vegetation indices prior to and after the damage by Masion Pine caterpillar showed that the change of these indexes was consistent with increasing values of these indices indicating lack of insect infestation and decreasing values of these indices correspond to insect outbreak and damage. The critical values of these indices reflecting different degrees of damage of the corresponding infested forest areas were determined by integrating CCD remote sensing data with the ground survey data. It was showed that dVI₆₆₀ was the most sensitive index among the examined indices with the highest accuracy of predicting Masion Pine caterpillar damage to Pinus massoniana forest. The current research proved that remote sensing data could be used to detect initial occurrence of Masion Pine caterpillar earlier than traditional methods.