基于MODIS数据辅助的太湖叶绿素a空间插值应用  

Spatial Interpolation of Chlorophyll-a in Taihu Lake Based on the MODIS Data Aided

作　　者：王玉磊[1] 钱新[1] 高海龙[1] 叶瑞[1] 张云燕[1] 

机构地区：[1]南京大学环境学院污染控制与资源化研究国家重点实验室,江苏南京210023

出　　处：《环境科学与技术》2014年第1期168-173,共6页Environmental Science & Technology

基　　金：国家水专项课题资助(2012ZX07506-007;2012ZX07101-014)

摘　　要：通过运用Modis太湖遥感数据反演得到的太湖叶绿素a浓度,利用克里格与加权反距离这2种插值法,首先进行全湖的插值方法的比较和插值样点数的选取;并对太湖8个湖区进行插值点分区研究;最后运用克里格对太湖进行全年不同季节的应用效果评价。结论如下:(1)在插值方法的选取上,克里格法更适合太湖叶绿素a的插值应用,全湖合理的取样数为68个。(2)各湖区蓝藻空间分布相关性由大到小依次为:南部沿岸区>湖心区>贡湖>东部沿岸区>东太湖>西部沿岸区>梅梁湖>竺山湖。(3)对于蓝藻不同生长阶段来说,在蓝藻休眠中后期(12-次年2月)、复苏期(3-4月)、生物量增加前期(4-7月)的预测效果较好;而在蓝藻生物量增加中后期(7-9月),上浮、积聚期到休眠前期(7-12月)预测精度较差。总体来说,冬季、春季插值的效果好于夏季,秋季插值效果最差。Based on the chlorophyll-a concentration obtained from the MODIS remote sensing data in Taihu Lake, two interpolation methods including Kriging and IDW （Inverse Distance Weighted） were used to study the divisional interpolation for 8 regions of Taihu Lake, and the method of Kriging was used to evaluate seasonal application effects for the whole Lake. Results showed that in the choice of chlorophyU-a interpolation method in Taihu Lake, Kriging is more suitable, with the whole lake reasonable sampling number as 68. The spatial distribution correlation of cyanobacteria between different lake regions is apparent, with cyanobacteria spatial distribution correlation from high to low as the southern part of lake 〉 the middle of lake 〉 Gonghu Lake 〉 the eastern part of lake 〉 East Taihu 〉 the western part of lake 〉Meiliang Lake 〉 Zhushan Lake. For the different growing stages of cyanobacteria, including the-dormant period of December-February, the recruitment period of March-April, and the earlier stage of biomass increasing of April-July, the accuracy of the prediction is good; while for the later stage of biomass increasing （July-September）, from the cyanobaeteria floating, accumulating to the earlier stage of sleeping （July-December）, the accuracy of the prediction is poor. Overall, winter and spring have a better interpolation effect than summer, and autumn is the worst.

关 键 词：空间插值 MODIS数据 叶绿素A 分湖区 预测误差 

分 类 号：X824[环境科学与工程—环境工程]