一次灰霾天气下的珠江三角洲NDVI反演  

The NDVI Inversion in Pearl River Delta in a haze Weather Incident

作　　者：王德辉[1,2] 匡耀求[2] 吴波[3] 周慧杰[4] 王萌[2] 

机构地区：[1]广州中海达卫星导航技术股份有限公司,广东广州511400 [2]中国科学院广州地球化学研究所,广东广州510640 [3]赣南师范学院,江西赣州341000 [4]广西师范学院,广西南宁530001

出　　处：《城市勘测》2015年第3期99-105,共7页Urban Geotechnical Investigation & Surveying

基　　金：广州市博士后基金项目(125124);江西省科技厅自然科学基金项目(20142BAB213027)

摘　　要：随着近年来灰霾天气影响区域和爆发频次的增多,灰霾已经严重影响到了卫星遥感对归一化植被指数(NDVI)的反演精度。现有针对气溶胶污染天气下的NDVI反演方法不适合非浓密植被地区和高气溶胶污染负载天气等情形。本文选取了珠江三角洲一次灰霾事件爆发前和爆发中的连续两天的MODIS(Terra)数据,构建了分区去气溶胶植被指数(ZAFRI,Zonal Aerosol Free Vegetation Index),用以反演"污染日"的植被指数。结果显示本方法可较好地消除大气气溶胶对NDVI的影响,ZAFRI比直接用NASA反射率产品计算的NDVI精度更高,可以作为灰霾"污染日"的NDVI替代指标。本方法将去气溶胶植被指数(AFRI)推广至珠江三角洲陆表各类区域上,为同类灰霾污染区域的NDVI反演提供一个新的可行途径。Haze has a strong impact on the precision of NDVI inversion by satellite remote sensing,with more regions being hit by haze more frequently in recent years. The inadequacy of existing inversion methods of NDVI in aerosol con-tamination weather is obvious in the case of non-dense vegetation and heavy load of aerosol contamination. This paper selects the MODIS（Terra）data for two successive days before and during the outbreak of a haze incident,establishing the Zonal Aerosol Free Vegetation Index to achieve the inversion of the NDVI in polluted days. The results show that this method can eliminate aerosol effects to NDVI better,and ZAFRI method has better precision than the NDVI directly cal-culated by NASA reflectivity products,and it can serve as a surrogate marker of NDVI in haze polluted days. This method generalizes the AFRI to all kinds of land surface regions,providing a new feasible method for the NDVI inversion in the same kinds of regions polluted by haze.

关 键 词：归一化植被指数 去气溶胶植被指数 灰霾 珠江三角洲 

分 类 号：TP753[自动化与计算机技术—检测技术与自动化装置]