基于卫星降水的重庆市旱涝监测及其可靠性分析

为了弄清重庆地区旱涝变化规律，为该地区旱涝监测、农业生产等提供参考，以处于长江流域上游的重庆市为研究对象，将TRMM 3B43月降水数据运用到单站点旱涝监测Z指数中，构建基于遥感数据的TRMM-Z指数，从而对重庆市2000—2011年间的旱涝情况进行年和月尺度上的监测，并利用同期气象站点Z指数对TRMM-Z指数进行可靠性分析。结果表明：（1）TRMM-Z指数监测方法在区域旱涝监测中具有很好的适用性和精度。TRMM-Z指数与气象站点Z指数年尺度相关系数为0.87，月尺度为0.85，都通过了P＜0.01的显著性检验。（2）年旱涝特征，在2000—2011年际间重庆地区共发生2次干旱（大旱、极旱）、2次洪涝（大涝、极涝），波及范围基本覆盖整个重庆地区。其余年份TRMM-Z指数在正常的范围，但也有局部区域受到一定程度旱涝灾害。（3）月旱涝特征，TRMM-Z指数围绕0呈锯齿状增大减小交替变化，其中，2011年4月干旱程度最强，TRMM-Z指数达-1.878，2007年7月洪涝程度最强，TRMM-Z指数达1.930；旱涝统计显示，洪涝发生频次略高于干旱，1年12个月里，平均约有2.6个月发生洪涝，2.5个月发生干旱；从旱涝变化趋势来看，1月、9月、10月呈下降趋势，其中1月下降幅度最大，其余各月均呈上升趋势；2006年的7月和8月表现为最干旱月份。

Floods/Drought Monitoring Based on TRMM Precipitation Z-Index and Its Reliability Analysis

In order to clarify the variation of floods/drought and provide references for floods/drought monitoring and agricultural production in Chongqing, the author regarded Chongqing which was in the upper of the Yangtze River as the research object. And through applying the monthly TRMM 3B43 precipitation data to the single site Z-index method, the author monitored Chongqing floods/drought on the scale of year and month from 2000 to 2011 by building TRMM-Z index based on remote sensing date. At the same time, the author analyzed the reliability of the TRMM-Z index by using the meteorological station Z Index. The results showed that: (1) TRMM-Z index monitoring method had good applicability and accuracy in the regional floods/drought monitoring. The correlation coefficient between TRMM-Z index and meteorological station Z index was 0.87 on year scale and monthly scale was 0.85. Both indicators are passed P<0.01 significance test. (2) In floods/drought characteristics, 2 droughts (drought, extreme drought and 2 floods (large floods, extreme floods) in Chongqing have occurred in 2000 to 2011. (3) Month floods/drought characteristics, TRMM-Z monthly index fluctuated up and down around the value 0, alternating droughts and floods. During the time, the strongest degree of drought was in December 2003 and its value was -1.878. The strongest degree of flood was in August 2006 and its value was 1.930. And the floods/drought statistics showed that floods (2.6 months in an average of a year) happened more frequently than drought (2.5 months in an average of a year). From the trends of floods/drought, January, September and October showed a downward trend and the largest declined in January. However, floods/drought showed an upward trend in the rest of the month. In 2000-2011, the driest happened in July and August in 2006.