Rb = 1- m 1+ m

1. OS52A-0530 bb Q =6.0 a+bb bb Q =4.0 a+bb bb/a bb (a+bb) b0-b1 - b0-b1- Oc4 b0-b1 – Linear Linear bb Q =3.14 a+bb Non-linear Non-linear Non-linear b2-q3-b2q4 b1- b2q3 b2q3-b2q4 b1-b2-q3 b1-b2q4 b1-b2q4 b1- b2q3 b1-b2q4 b1-b2q6 Linear Relationship Used Current Semi Analytical algorithms ? ? Remote Sensing Reflectance Backscattering bbt l Rrsl= 0.051 Absorption at l Backscattering bb ~ bbw  + bbp Absorption at ~ aw + a + ad + ag colored dissolved organic matter water phytoplankton detritus Non- Linear 2 Rb = 1- m 1+ m 1-g m = • 1+2g + g (4+5g) g = bb a + bb Remote Sensing Reflectance Conversion of Rb to RRS for non-linear Q= Changes in Water type Q= 3.14 a = 0.1778 Q = 4 a = 0.1357 Q = 6 a = 0.0950 RRS = F T 2 bb Q n 2 a+bb 0.051 typically g Use the Non-linear Rb and Convert to the RRS RRS = 1 T2 Rb Q n2 MISSISSIPPI Bight- SeaWIFS processing April 18,2001 Improved Algorithms for Retrieving Optical Properties in Coastal Waters from Ocean Color Sensors g = R. A. Arnone1, R. W. Gould1, P. A. Martinolich2, S. Ladner3, A. W. Weidemann1, and V. Haltrin1 1. Naval Research Laboratory, SSC, MS 2. Neptune Sciences SSC. MS 3. Planning Systems Inc, SSC. MS Chlorophyll OC4 SeaWIFS Imagery was processes using the NIR – which is a coupled ocean - atmospheric Correction (Gordon) . We varied the linear – and non-linear RRS with different Q parameters and determined the affect on coastal optical products. 1. Chlorophyll - NASA –OC4 algorithms 2. bb550 – Arnone algorithms 3. absorption 440 Total – Carder Algorithm The coupled ocean-atm algorithm is triggered by the Lu670 where high scattering has significant influence on the Lt765 and Lt865, which are used for atmospheric correction. Therefore, the nonlinear RRS and g will affect coastal waters products and not offshore waters. Objective : - Determine the affect of linear and non-linear relationships Which IOP’s (backscattering and absorption) have on reflectance. RRS ~ bb/a ~ bb/(a+bb) - Apply these relationships to SeaWIFS processing and determine their affect Coastal and offshore waters. Chlorophyll, backscattering (550) and Total Absorption - Determine where (water type) and how magnitude these relationships affect Ocean Color “SeaWIFS” Algorithms. We developed improved SeaWIFS coastal ocean color algorithms to derived inherent optical properties, based on relationships between absorption, scattering and remote sensing reflectance. The linear remote sensing reflectance to scattering: absorption ratio (bb/a) is the basis for open ocean algorithms where absorption (predomoninantly from chlorophyll) is greater than backscattering. In coastal waters, where backscattering from sediment can dominate absorption, a non-linear and spectral dependence occurs between the reflectance and the backscatter: absorption ratio. These nonlinear influences affect not only the in water optical algorithms, but they are also coupled with the atmospheric correction in coastal waters. The removal of water leaving radiance in the near-IR (765 and 865 nm) is especially necessary in coastal waters. The non-linear relationship is used in estimating the water leaving radiance in the near-IR through an iterative pixel-by-pixel process using the 670 nm water leaving radiance. We used SeaWIFS imagery and insitu measurements to evaluate the effects the non-linear relationships have on coastal algorithms where backscattering dominates the absorption. We compare these new products with the more standard NASA products and we highlight areas where regional differences are greatest (bays, estuaries). Changes in Q – Little Affect on Chlorophyll b2q4-b2q6 Differences bb in denominator decrease Chl in coastal waters Non linear with Q=3, Increases Chl. Little change with Q=4,6 bb_555_arnone Maxium Non linear b2q3-b2q6 Changing from Q=3-6 Decreases bb Differences Changes in bb product Are significant with Q changes. bb in denominator increases bb in coastal waters Non linear with Q=3, Decrease bb Larger decrease with Q=4,6 Changes from bb/a to bb/(a+bb) is significant in High scattering (bb) waters. Total Absorption “Carder” Origin of the Remote Sensing Reflectance? Linear Conversion of Diffuse Reflectance to Remote Sensing Reflectance Rb = 0.33 bb a+ bb Where Rb= Diffuse Reflectance Haltrin Appl. Opt. 37, 3773-3774 (1998) Changes in “a total” product Are similar with non-linear Q changes. Insitu - Comparisons Differences Currently 0.019 • Summary • A linear g –RRS relationship effectively changes the Q-value in the non-linear relationship. Q changes from +3.14 to 6.0 with increasing “g” in coastal waters.. • Insitu shows high variability of the RRS - “g” relationship associated with non-linear and bb-b relationship • Significant differences occur in SeaWIFS products (bb) in high scattering water (coastal • and shelf waters ) that are associated with Q. Small changes in CHL and “a”. • Improved algorithms will require estimate of Q in coastal waters for bb products. bb in denominator decrease chl in coastal waters Where are These Relations Used in Remote Sensing Algorithms - NIR – Iteration Coupled Ocean – Atm Algorithm - extension of Gordon- atm into Coastal waters by Lu765 = 0 - Bio – optical Algorithms bb and a(total) bb ~ RRS Insitu data from different Cruises Large variability in coastal Waters. Influence of Q variability ? bb –b conversion ?? Spectral Dependence of Insitu - 550 nm 440 nm 670 nm Red has lower “g” from attenuation Green has largest variation in coastal Waters. For Contact: Robert Arnone Head Ocean Optics Section Naval Research Laboratory Stennis Space Center, MS 39529 arnone@nrlssc.navy.mil (228) 688-5268 http://www7333.nrlssc.navy.mil Insitu Observations in Coastal Waters 90 stations - RRS – ASD (Above water) RRS - absorption – ac9 - bb – ac9 converted b to bb (Petzold) bb to b *53 (1.97%) (Source of Error) Variation in the bb/b relationship illustrates the changing Q parameter. Currently using ~2%. Volume Scattering Functions (VSF) show high variability in different water types. This is responsible for high insitu scatter.