中国汶川Ms 8.0级地震前NDVI变化特征及其佐证

 利用归一化植被指数（NDVI）数据，探测了汶川Ms 8.0级地震前植被变化情况。对震前9年的NDVI数据先做整体变化分析，没有发现明显的植被退化现象；再对时间序列数据进行逐年比对，发现两个比较明显的植被退化现象。其一，2006年与2005年相比，龙门山断裂东部的四川盆地大部分地区6—9月NDVI均值明显减小，幅度为0～20％。推测原因为震前四川盆地受挤压变形，产生热弹效应放热致使地表升温，进一步与大气耦合导致长期干旱少雨，破坏植被正常生长条件，使植被退化。因此，2006年四川盆地大范围植被退化现象可视为震前异常。其二，2008年5月上旬与2007年相比，龙门山断裂带上发现明显的植被退化现象，成条带状分布，NDVI下降幅度超过20%。这可能与临震前龙门山断裂带上发生某些物理化学变化有关，即临震前龙门山断裂带地壳岩石的破裂发育、断裂之间以及岩石之间的摩擦滑移，导致地下热物质上涌和摩擦生热释放，造成短期（近）地表高温，破坏了植被的正常生长条件，使其退化。因此，可以看作震前短期异常。此外，沿着时间尺度和空间尺度，找到一些震前中期和短期异常佐证，以增加植被异常判断的可信度。

NDVI Variation Features Before Wenchuan Ms 8.0 Earthquake

By using NDVI data, the continuous change of vegetation prior to the Ms 8.0 Wenchuan earthquake is investigated in this paper. Firstly, according to a comparison year by year, it was found that, to the east of Longmenshan Fault, in Sichuan Basin, the vegetation was degraded by 0～20% from 2005 to 2006. The phenomenon was caused by drought in 2006. Spatially, the movements of Sichuan Basin plate played an important role on the Wenchuan earthquake. Therefore, according to the relationship between drought and earthquake, the degradation could be regarded as one of precursory anomalies prior to the Wenchuan Earthquake. Secondly, the vegetation in early May in Longmenshan Fault was degraded significantly from 2007 to 2008. Moreover, the same phenomenon did not happen in other adjacent faults, like Xianshuihe Fault. As is well known, prior to earthquakes, there may be changes of electric and magnetic fields, gas emissions, groundwater levels, temperature, surface deformations, and so on. These changes are all related with the vegetation degradation to some extent, and they could be seen as a short-term anomaly prior to the 2008 Wenchuan Earthquake. Furthermore, the two anomalies developed spatio-temporally from distant regions to near regions with respect to the epicenter. Additionally, some evidences were found to re-confirm the effects of vegetation anomaly. For example, the anomaly of Earth resistivity was observed 2 years before the earthquake. However, there are some limitations for detection of vegetation anomaly prior to earthquake due to other influencing factors, such as growth periodicity and topography. Further research is needed.