|
| 基于半分析算法的赤潮水体固有光学性质反演 |
| Retrieval of Oceanic Inherent Optical Properties Based on Semi-analyticl Algorithm during Red Ride |
| 投稿时间：2009-11-20 修订日期：2010-11-29 |
| DOI： |
| 中文关键词: 赤潮 固有光学参数 半分析算法 QAA算法 |
| 英文关键词:red tide inherent optical properties semi-analytical algorithm Quasi-Analytical Algorithm |
| 基金项目:国家自然科学基金项目(40706060，60802089)；国家高技术研究发展计划“863”项目(2006AA09A310，2007AA092102) |
|
| 摘要点击次数: 5023 |
| 全文下载次数: 465 |
| 中文摘要: |
| 在珠江口海域海洋光学浮标实验期间，获取了赤潮生消过程的水体光学数据和相应的生化数据。本文利用该实验数据开展了QAA（Quasi-Analytical Algorithm）算法反演赤潮水体固有光学参数的精度检验和模型修正工作。（1）利用遥感反射率rrs和QAA算法反演主要水色波段（412 nm, 443 nm, 490 nm、510 nm, 560nm, 620 nm和665nm）固有光学量发现：490nm-665nm波段水体总吸收系数at和颗粒物后向散射系数bbp的反演误差优于25%，443nm波段at和bbp反演误差在8-32%之间，412nm波段at和bbp反演误差在31-57%之间。浮游植物吸收系数aph和黄色物质与碎屑吸收系数adg反演误差分别在32-157%和59-80%之间。对比两种参考波段（555nm和640nm）选取方式发现，除490nm和510nm外，QAA-640算法在其它波段的反演精度均优于QAA-555算法，其中412和443nm波段尤为明显。（2）利用实测数据进行了QAA算法的修正：对于at和bbp，QAA-555修正算法的反演精度有不同程度的提高，误差减小1-11%，QAA-640修正算法的反演精度在412nm、620nm和665nm波段有所提高；对于aph和adg，修正算法的反演精度没有明显改善。除412nm和443nm波段外，修正的QAA-555均优于修正的QAA-640。（3）利用现场实测at数据和QAA算法反演aph和adg，误差均优于24%；算法的误差敏感性分析发现，当at引入20%的误差时，adg反演误差波动不明显（<33%），而aph在560-665nm波段反演误差波动可达58-241%。 |
| 英文摘要: |
| In the optical buoy experiments in the Pearl River (Zhujiang) mouth regions, we obtained the time series of bio-optical data of red tide. By using these data, we carried out retrieving inherent optical parameters of red tide with semi-analytical algorithm, and testing, validating and modifying the model with in situ data. (1) In the retrieval of inherent optical properties at major satellite ocean color bands (412nm, 443nm, 490nm, 510nm, 560nm, 620nm and 665nm) with rrs and QAA (Quasi-Analytical Algorithm), the errors for total absorption coefficient at and particle backscattering coefficient bbp at 490nm-665nm are less than 25%. The errors for at and bbp at 443nm are about 8-32%, and about 31-57% at 412nm. The error ranges for phytoplankton pigment absorption aph, detritus and colored dissolved organic matter absorption adg are 32~157% and 59~80%, respectively. Comparison of retrieval result by the different reference wavelengths (555nm and 640nm) of QAA, it’s found that with the exception of 490 nm and 510 nm, QAA-640 performed better than QAA-555 at other bands, especially at 412 and 443 nm. (2) QAA is modified by in situ data. Modified QAA-555 model derived better at and bbp. The errors decrease by 1~11%. Modified QAA-640 returns better at and bbp at 412nm, 620nm and 665nm. For aph and adg, the accuracy of retrieval results has no significant improvement. With the exception of 412nm and 443 nm, modified QAA-555 is better than modified QAA-640. (3) We used QAA to retrieve aph and adg from in situ measured at, the errors are less than 24%; When added 20% errors to the in situ at, the retrieval error for adg has no significant changes(<33%), but the errors of aph fluctuate in the range of 58-241% at 560-665nm. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
