Assessment and development of algorithms to detection of oil spills using MODIS data
Assessment and development of algorithms to detection of oil spills using MODIS data
复制标题
DOI：
10.1007/s12524-010-0007-9
复制
发表时间：
2010-08
影响因子：
2.5
通讯作者：
H. Srivastava;T. Singh
中科院分区：
文献类型：
--
作者：
H. Srivastava;T. Singh
溢油检测和随后的监测是沿海地区管理的主要关切，因为它们形成了潜在的海洋污染物。本研究的基础是研究中分辨率成像光谱仪(MODIS)高分辨率波段(250米和500米)用于检测溢油的可行性。2003年1月18日、19日和20日的MODIS-AQUA被用于研究委内瑞拉马拉开波湖的漏油事件。对L1B和L2数据的检查表明，由于大气改正错误，L2产品，如SST、RRS和BRDF并不是很有用。目视检查250米和500米空间分辨率的原始辐射数据，即L1B数据，被认为是最简单但可行的溢出检测方法。这项研究进一步表明，大气中469、555和645纳米处的未经校正的辐射显示出明显的漏油特征。因此，尝试执行比率运算来增强该特征。结果表明，645 nm和555 nm处的差值与总和之比经469 nm归一化后效果最好。通过与已发表文献的比较，验证了结果的正确性。这一结果清楚地表明了MODIS-AQUA高分辨率数据在溢油监测中的潜力。因此，对于此类活动，每日覆盖和高分辨率的MODIS-AQUA数据可以是可靠和经济有效的。
Oil spill detection and subsequent monitoring are of major concern for coastal zone management as they form potential marine pollutants. The present study is based on examining the feasibility of Moderate Resolution Imaging Spectroradiometer (MODIS) high resolution bands (250 and 500 m) for detection of oil spill. The MODIS — Aqua for 18, 19 and 20thJanuary 2003 were used to study the oil spill in Lake Maracaibo, Venezuela. The examination of L1B and L2 data revealed that L2 products such as SST, Rrs and BRDF were not very useful due to erroneous atmospheric corrections. Visual examination of raw radiance data i.e. L1B data in the 250 and 500 m spatial resolution was found to be the simplest yet feasible method for spill detection. The study further revealed that atmospherically uncorrected radiances at 469, 555 and 645 nm were showing significant signature of oil spill. Therefore an attempt was made to perform ratio operations to enhance the feature. The study concluded that the ratio of difference and sum at 645 and 555 nm, normalized by 469 nm, provides the best result. The result was validated by comparing with the previous published literature. The result clearly indicates the potential of MODIS-Aqua high resolution data in oil spill monitoring. Therefore, MODIS-Aqua data with daily coverage and high resolution can be reliable and cost-effective for such events.