Development of ocean color algorithms in the Mediterranean

Development of ocean color algorithms in the Mediterranean Sea Rosalia Santoleri 1, , Gianluca Volpe 1, Simone Colella 1, 3, Salvatore Marullo 2, Maurizio Ribera D’Alcalà 3, Vincenzo Vellucci 3 1 2 3 ENEA -CR Casaccia – Sezione Modellistica Oceanografica Stazione Zoolologica ‘A. Dohrn’ Laboratorio di Oceanografia Biologica

ISAC Contribution to Ocean Color activity Mediterranean high resolution surface chlorophyll mapping • Use available bio-optical data sets to estimate the uncertainties of the existing ocean colour algorithms and to define an optimal chlorophyll algorithm for the Mediterranean Sea. • Adapt the OC processing software to include selected regional algorithms and validate satellite chlorophyll estimates on the basis of in situ data. • Evaluate the uncertainties of all current global satellite chlorophyll products available from public archive (e. g. DAAC) in the Mediterranean • Reprocessing Sea. Wi. FS dataset using the selected Mediterranean algorithm • Prepare Mediterranean gridded data compatible with model requirements.

Why Mediterranean needs regional algorithm ? Chlorophyll concentrations over the oligotrophic waters of the Mediterranean Sea are systematically overestimated when global algorithms (e. g. OC 4 v 4) are used to convert blue-to-green reflectance ratios in to chlorophyll-a concentrations: • Gitelson et al. (Journal of Marine System, 1996) • D’Ortenzio et al. (SIMBIOS meeting January 2001) • D’Ortenzio et al. (Remote sensing of the Environment, 2002) • Bricaud et al. (Remote sensing of the Environment, 2002) • Claustre et al. (Geoph. Res. Letters, 2002) From these works it results that global algorithms cannot be applied to-court to the Mediterranean Sea but a specific cal/val activity is needed.

Mediterranean Ocean Color CAL/VAL DATA SETS 10 Mediterranean cruises from 1998 up to now were organized by ISAC in the framework of Italian National Projects Bio-optical stations Bio-optical measurements: (143 chl/opt measurement points) to define the Mediterranean regional algorithm • • • In water downwelling irradiance and upwelling radiance profiles using SATLANTIC SPMR above water measurements using the SIMBAD and SIMBADA radiometer In the bio-optical stations phytoplankton pigments distribution (HPLC and spectro-fluorometric analysis) and ancillary biological data were also acquired following NASA protocols. In situ chlorophyll-a data: (872 chl profiles) to validate Sea. Wi. FS, Polder, MODIS, MERIS chlorophyll products and merged level 3 binned data produced by Mersea

Cruise Period Zone # of profiles Chlorophyll Range MIN MAX # of Satlantic # of SIMBAD(A) MATER 4 25/4/1998 15/5/1998 Sardinia Channel Sicily Channel 57 0. 025 0. 105 - - MATER 5 20/10/1998 27/10/1998 Sicily Channel 50 0. 047 0. 085 - - EMTEC 99 20/4/1999 7/5/1999 Ionian Sea 125 0. 039 0. 137 20 - MATER 6 14/5/1999 30/5/1999 Sardinia Channel Sicily Channel 100 0. 003 0. 135 - - PROSOPE 04/09/1999 14/10/1999 Western Basin Ionian Sea 16 0. 020 0. 112 16 - SYMPLEX 99 21/10/1999 6/11/1999 Sicily Channel Ionian Sea 212 0. 039 0. 176 12 - NORBAL 1 26/3/2000 19/4/2000 Gulf of Lions 81 0. 113 2. 289 - - NORBAL 2 5/12/2001 20/12/2001 Gulf of Lions Tyrrhenian Sea 65 0. 088 0. 386 13 14 NORBAL 4 6/3/2003 26/3/2003 Gulf of Lions 115 0. 428 7. 061 16 28 NORBAL 5 18/4/2003 25/4/2003 Gulf of Lions 40 0. 605 2. 096 4 7 DINA* 29/3/2001 28/8/2001 Gulf of Naples 11 0. 079 0. 455 11 - DYFAMED* 5/2/1998 25/11/2002 Liguro-Provencal 55 0. 042 2. 366 - - All cruises 1998 -2003 Mediterranean 872 0. 003 7. 061 92 49 141

REGIONAL Validation of Ocean Color Algorithms OC 4 v 4: R is log 10 of either the D’Ortenzio et al. 2002 (DORMA): R is log 10 of the 490/555 band reflectance ratios. 443/555 or the 490/555 or the 510/555 band reflectance ratios, depending on its value (the maximum is chosen) Bricaud et al. 2002: R is log 10 of the 443/555 band reflectance ratios for Chl<0. 2 (OC 4 v 4 is used in the other cases) GLOBAL

MEDOC 4 : R is log 10 of either the 443/560 or the 490/555 or the 510/555 band reflectance ratios. The switch from one band ratio to another one is based on the chlorophyll concentration itself (the Maximum is Chosen) NEW MEDITERRANEAN ALGORITHM

Sea. Wi. FS data validation A Match-up dataset between concurrent Sea. Wi. FS data and in situ measurements has been contructed 1. 2. 3. 4. Sea. Wi. FS L 1 A passes corresponding to in situ observation were selected The regional algorithms (DORMA, Bricout et al, and the new MED OC 4) were inserted in the Sea. DAS Code Sea. Wi. FS L 1 A passes were processed using the different algorithms A Match-up dataset is contructed.

Map of the Satellite-in situ match-up stations

Merging OC data in the MED • Define the error of global OC color products at regional scale and define a strategy to take into account the regional optical properties of the MED • Define of the suitable intercalibration and merging tecnique for OC • Define an optimal interpolation algorithm that takes in to account the different characteristics of ocean colour retrieval in case I/case II waters.

10 March 2003

12 March 2003

25 March 2003

Primary production in the Mediterranean Sea from remote sensing data: a model adaptation

Depth (m) CHL (mg m-3)

Remote Sensed data of pigment concentration Trophic States e Morel and Berthon’s (1989) Correlations Primary Production Maps

Antoine et al. , 1995 Bosc et al. , 2003 In situ C 14 Method WEST MED: 197 gc m-2 y-1 EST MED: 137 gc m-2 y-1 WEST MED: 172 gc m-2 y-1 EST MED: 123 gc m-2 y-1 WEST MED: 78 -150 gc m-2 y-1 EST MED: 55 -97 gc m-2 y-1

Chlorophyll vertical distribution in first two trophic states Morel (dashed line) This work (continuous line)

4000 chlorophyll profiles acquired in the Mediterranean Sea during 16 cruises (1996 up to now) has been use to compute the chlorophyll concetration of the first optical depth (Cpd) Chlorophyll concentration intervals: >0. 05, 0. 05 -0. 1, 0. 1 -0. 15, 0. 15 -0. 3, 0. 3 -0. 45, 0. 45 -1. 5, 1. 5 -5, > 5 mg m-3

Morel (dashed line) This work (continuous line)

Ctot=40. 6 Cpd 0. 46 = Ctot=40. 6 Csat 0. 459 Ctot=54. 679 Cpd 0. 6532 ≠ Ctot=48. 13 Csat 0. 627

Gennaio Febbraio Marzo Aprile Maggio Giugno Luglio Agosto Settembre Ottobre Novembre Dicembre 0 0. 2 0. 4 0. 6 0. 8 gc m-2 g-1 1. 0 1. 2 1. 4 1. 6

Antoine et al. , 1995 Bosc et al. , 2003 In situ C 14 method Our Estimate WEST MED: 197 gc m-2 y-1 EST MED: 137 gc m-2 y-1 WEST MED: 172 gc m-2 y-1 EST MED: 123 gc m-2 y-1 WEST MED: 78 -150 gc m-2 y-1 EST MED: 55 -97 gc m-2 y-1 WEST MED: 130 gc m-2 y-1 EST MED: 95 gc m-2 y-1