The SeaWiFS Bio-Optical Algorithm Mini-Workshop (SeaBAM)
The original Bio-optical Algorithm Mini-Workshop group was born out of informal meetings conducted during the Ocean Optics XIII conference, held in Halifax, Canada from 21-25 October 1996. The SeaWiFS Bio-optical Algorithm Mini-Workshop (SeaBAM) was held at the University of California, Santa Barbara from 22-24 January 1997. Results from the latter are posted below.
Purpose:
- Finalize the operational SeaWiFS chlorophyll a and CZCS-pigment algorithms.
- Realize the SeaWiFS bio-optical algorithm performance goal of 35% accuracy over a range of 0.05-50 mg m-3.
- Establish final definition of CZCS-pigment (present definition = [chl a + phaeo]).
- Establish final definition of accuracy and identify the appropriate statistical parameter(s) for its quantification.
- Establish criteria for final selection of "best" algorithm for both pigment parameters.
- Define algorithms to be compared and identify probable reasons for differences.
Participants:
- Ken Carder
- Sara Garver
- Steve Hawes
- Mati Kahru
- Stephane Maritorena
- Chuck McClain
- Greg Mitchell
- Gerald Moore
- Jim Mueller
- Jay O'Reilly
- Brian Schieber
- Dave Siegel
Candidate algorithms:
- All empirical algorithms are described in the linked table.
- Garver and Siegel semi-analytical algorithm (a.k.a. UCSB inherent optical properties model)
- Carder and others (USF) semi-analytical algorithm
Evaluation dataset:
Definitions:
CZCS pigment: A fluorometric pigment concentration (chlorophyll a + phaeopigments) that can be calculated using bands comparable to the CZCS wavelengths (443, 520, 550). Note that the SeaWiFS protocols need to be more detailed on this topic. The purpose of generating this product is to provide a means of comparing products that can be derived from the CZCS to those from later missions for examining decadal scale variability. Restricting the algorithm to the CZCS wavelengths minimizes biases introduced in the products that are artifacts of the algorithm form. It is assumed that the global CZCS data set will be reprocessed using an updated pigment algorithm that is consistent with the SeaWiFS pigment algorithm. Maritorena will evaluate the assumption that the differences between fluormetric and HPLC bio-optical data sets are indistinquishable using the clear sky data set. The issue of how to validate the reprocessed CZCS products using simultaneous measurements was discussed. Given that algorithms being developed at this time are based on different pigment measurement methodologies which yield different values, validation using historical data will require some adjustment in the historical values.
Chlorophyll a: Any fluorometric or HPLC concentration identified as chlorophyll a by the provider. While there are differences in the values obtained by the two techniques, globally the difference has been shown to be of the order of 10%. Debate continues as to what pigments should or are being summed and reported as "chlorophyll" concentration in the data sets being submitted to SeaBASS. Because other sources of variability in the bio-optical data sets, e.g., data processing methods and calibration have been found to be as great, and in order to have enough data over a large dynamic range to develop and evaluate algorithms, this definition was adopted. Action item: Maritorena will separate the fluorometric and HPLC data to see if the differences are distinquishable and/or color code the two in the global comparison to see is the two overlap. Status: There are 416 HPLC and 673 fluorometric bio-optical stations with about 156 with both. However, 73 of those stations with both are ±92-'93 BATS data with 565 nm, but not 555 nm optical measurements. Therefore, the comparison data set is too small for a valid comparison. Dennis Clark' MODIS bio-optical algorithm ATBD contains a comparison by Chuck Trees that indicates HPLC pigments are about 7% smaller than fluorometric pigments (same size > 2300; concentrations < 10 mg m-3). Of course, this difference varies regionally. It was decided that the difference for global algorithm considerations is small enough that blending the two methods is acceptable.
Contributions to the SeaBAM Technical Memorandum:
Many of these documents are in PostScript form
- SeaBAM overview (mcclain)
- SeaBAM evaluation data set (maritorena_oreilly_schieber)
- SeaBAM algorithm evaluation (oreilly_maritorena)
- SeaWiFS algorithms and CalCOFI data (mitchell)
- UCSB inherent optical properties model (garver)
- USF semi-analytical model (hawes)
Final SeaWiFS chlorophyll a algorithm:
Ocean Chlorophyll 2 (OC2)
- Chl = -0.040 + 10^[0.341 - 3.001 * X + 2.811 * X^2 - 2.041 * X^3]
- X = log10 [Rrs(490)/ Rrs(555)]
Additional reading:
- O'Reilly, J.E., S. Maritorena, B.G. Mitchell, D.A. Siegel, K.L. Carder, S.A. Garver,
M. Kahru, and C. McClain, 1998: Ocean color chlorophyll algorithms for SeaWiFS.
J. Geophys. Res., 103, 24,937-24,953
- Maritorena, S. and J.E. O'Reilly, 2000: "OC2v2: Update on the initial operational
SeaWiFS chlorophyll a algorithm." In: J.E. O'Reilly and co-authors, SeaWiFS
Postlaunch Calibration and Validation Analyses, Part 3. NASA Tech. Memo.
2000-206892, Vol. 11, S.B. Hooker and E.R. Firestone, Eds., NASA Goddard Space
Flight Center, Greenbelt, Maryland, 3-8.
- O'Reilly, J.E. and co-authors, 2000: "Ocean color chlorophyll a algorithms for SeaWiFS, OC2, and OC4: Version 4." In: J.E. O'Reilly and co-authors, SeaWiFS Postlaunch Calibration and Validation Analyses, Part 3. NASA Tech. Memo. 2000-206892, Vol. 11, S.B. Hooker and E.R. Firestone, Eds., NASA Goddard Space Flight Center, Greenbelt, Maryland, 9-23.
Author/Site Curator: Jeremy Werdell
Responsible NASA Official: Charles McClain
Last revised 6 June 2002