An empirical algorithm to seamlessly retrieve the concentration of suspended particulate matter from water color across ocean to turbid river mouths

REMOTE SENSING OF ENVIRONMENT

卷:235

Article

Yu, Xiaolong^1,2,3  Lee, Zhongping¹  Shen, Fang³  Wang, Menghua⁴  Wei, Jianwei⁴  Jiang, Lide⁴  Shang, Zhehai¹ 

[1] Univ Massachusetts Boston, Sch Environm, Boston, MA 02125 USA

[2] Xiamen Univ, Coll Ocean & Earth Sci, State Key Lab Marine Environm Sci, Xiamen 361101, Fujian, Peoples R China

[3] East China Normal Univ, State Key Lab Estuarine & Coastal Res, Shanghai 200062, Peoples R China

[4] NOAA, Natl Environm Satellite Data & Informat Serv, Ctr Satellite Applicat & Res, E RA3, 5830 Univ Res Court, College Pk, MD 20740 USA

关键词: Remote sensing reflectance;    Suspended particulate matter;    Water color;    VIIRS;    Turbid waters;    Global algorithm;   

DOI  :  10.1016/j.rse.2019.111491

来源: Elsevier

【 摘 要 】

We propose a globally applicable algorithm (GAA(SPM)) to seamlessly retrieve the concentration of suspended particulate matter (SPM) (C-SPM) from remote sensing reflectance (R-rs(lambda)) across ocean to turbid river mouths without any hard-switching in its application. GAA(SPM) is based on a calibrated relationship between C-SPM and a generalized index for SPM (GI(SPM)) from water color. The GI(SPM) is mainly composed of three R-rs(lambda) ratios (671, 745, and 862 nm over 551 nm, respectively), along with weighting factors assigned to each ratio. The weighting factors are introduced to ensure the progressive application of R-rs(lambda) in the longer wavelengths for increasing C-SPM. Calibration of GAA(SPM) employed data collected from multiple estuarine and coastal regions of Europe, China, Argentina, and the USA with the measured C-SPM spanning from 0.2 to 2068.8 mg/L. Inter-comparison with several recalibrated well-known C-SPM retrieval algorithms demonstrates that GAA(SPM) has the best retrieval accuracy over the entire C-SPM range with a relative mean absolute difference (rMAD) of 41.3% (N = 437). This averaged uncertainty in GAA(SPM) -derived C-SPM is mostly attributed to the retrievals from less turbid waters where C-SPM < 50 mg/L (rMAD = 50%, N = 214). GAP(SPM) was further applied to the Visible Infrared Imaging Radiometer Suite (VIIRS) measurements over prominent coastal areas and produced reliable C-SPM maps along with realistic spatial patterns. In contrast, applications of other C-SPM algorithms resulted in less reliable C-SPM maps with either unjustified numerical discontinuities in the C-SPM spatial distribution or unsatisfactory retrieval accuracy. Therefore, we propose GAA(SPM) as a preferred algorithm to retrieve C-SPM over regions with a wide range of C-SPM, such as river plume areas.

【 授权许可】

Free