Estimation of Transmittance of Solar Radiation in the Visible Domain Based on Remote Sensing: Evaluation of Models Using In Situ Data. Issue 11 (27th November 2017)

• Record Type: Journal Article
• Authors: Zoffoli, M. Laura; Lee, Zhongping; Ondrusek, Michael; Lin, Junfang; Kovach, Charles; Wei, Jianwei; Lewis, Marlon
• Is Part Of: Journal of geophysical research. Volume 122:Issue 11(2017)
• Page Start: 9176
• Page End: 9188
• Publication Date: 2017-11-27
• Subjects: attenuation coefficient -- inherent optical properties -- chlorophyll concentration -- water column -- algorithm performance -- uncertainties; Oceanography -- Periodicals; 551.4605
• DOI: 10.1002/2017JC013209
• Languages: English
• ISSNs: 2169-9275

Abstract

Abstract: The transmittance of solar radiation in the oceanic water column plays an important role in heat transfer and photosynthesis, with implications for the global carbon cycle, global circulation, and climate. Globally, the transmittance of solar radiation in the visible domain (∼400–700 nm) (TRVIS ) through the water column, which determines the vertical distribution of visible light, has to be based on remote sensing products. There are models centered on chlorophyll‐a (Chl) concentration or Inherent Optical Properties (IOPs) as both can be derived from ocean color measurements. We present evaluations of both schemes with field data from clear oceanic and from coastal waters. Here five models were evaluated: (1) Morel and Antoine (1994) (MA94), (2) Ohlmann and Siegel (2000) (OS00), (3) Murtugudde et al. (2002) (MU02), (4) Manizza et al. (2005) (MA05), and (5) Lee et al. (2005 ) (IOPs05), where the first four are Chl‐based and the last one is IOPs‐based, with all inputs derived from remote sensing reflectance. It is found that the best performing model is the IOPs05, with Unbiased Absolute Percent Difference (UAPD) ∼23%, while Chl‐based models show higher uncertainties (UAPD for MA94: ∼54%, OS00: ∼133%, MU02: ∼56%, and MA05: ∼39%). The IOPs‐based model was insensitive to the type of water, allowing it to be applied in most marine environments; whereas some of the Chl‐based models (MU02 and MA05) show much higher sensitivities in coastal turbid waters (higher Chl waters). These results highlight the applicablity of using IOPs products for such applications. Key Points: The vertical distribution of solar radiation in the visible domain estimated from models using remote sensing data as input is evaluated The best performance was provided by a scheme centered on the Inherent Optical Properties (IOPs) The mean uncertainty of the estimated transmittance for a range of 0.1–100% is ∼23% from the IOPs system and insensitive to water types