Data Policy and Sea BASS Evolution Giulietta S

Data Policy and Sea. BASS Evolution Giulietta S. Fargion CHORS April 11, 2006

Sea. BASS Background • To facilitate the assembly of a global bio-optical data set, the Sea. Wi. FS Project developed the Sea. Wi. FS Bio-optical Archive and Storage System (Sea. BASS), a local repository for in situ radiometric and phytoplankton pigment data used regularly in scientific analyses (Hooker et al. 1994, Werdell and Bailey 2002 ). • The system was expanded to contain oceanographic and atmospheric data sets collected by the NASA Sensor Intercomparison and Merger for Interdisciplinary Biological and Oceanic Studies (SIMBIOS) (1997 -2004). • To develop consistency across multiple data contributors and institutions, the Sea. Wi. FS & SIMBIOS Projects defined and documented a series of in situ data requirements and sampling strategies that ensure that any particular set of measurements will be acceptable for bio-optical and atmospheric correction algorithm development and ocean color sensor validation (Mueller and Austin 1995, Fargion et al. 2001, Mueller et al. 2002; 2003, 2004 & 2005).

Under SIMBIOS • Community collected in situ data with an agreed upon protocol and a “set of required” measurements for validation match-ups and algorithm development and validation • The Program in 2 nd NRA targeted some ocean critical areas • The submission rate was high (every 3 -6 months). Was this due to the funding vehicle (contracts vs. grants) ? • Community participated in large international experiments. Bio-optical and chemistry expertise was achieved by the integration of individual PIs in a “SIMBIOS” team • Community could use an instrument pool • The Project supported calibration RR’s, protocol development and instrument R&D with the community

Principal in situ observations for satellite ocean color system validation, and algorithm development and validation. The right-hand column identifies and classifies measurements as: (a) required for minimal validation match-ups; (b) highly desired and important for general algorithm development and validation; (c) specialized measurements of important, but restricted, applicability to algorithm development and validation (for the present); and (d) calculated or derived quantities Table from Mueller et al. 2004.

Table from Mueller et al. 2004.

NASA Policy Requires Data Submission to Sea. BASS • Submission: Ocean color algorithm development and new products are essentially observation limited, and rapid turnaround access to such data are crucial for progress. “You can’t keep coming in here and demanding data every week!”

• Formats and Metadata: Data should be provided in the currently agreed-upon format, along with relevant information describing collection conditions, instrument specifications, instrument performance and calibration, and statements of data accuracy. The currently used data format specifications and examples are posted on the Sea. BASS web site. The provider should use FCHECK, which is an automated format checker program, to test the format validity of Sea. BASS data files via return e-mail. • Use Conditions: Prior to the three-year data collection anniversary, users of data will be required to provide proper credit and acknowledgment of the provider. A citation should also be made of the data archive. Users of data are encouraged to discuss relevant findings with the provider early in the research. • More information ? – http: //seabass. gsfc. nasa. gov – jeremy@seawifs. gsfc. nasa. gov

Current Sea. BASS Use • Historically, the data archived in Sea. BASS have predominantly been used for satellite-data-product validation activities and bio-optical algorithm development (O’Reilly et al. 1998, Bailey et al. 2000, Hooker and Mc. Clain 2000, Moore et al. 2001, Maritorena et al. 2002, Schwarz et al. 2002). • Under SIMBOS however, these data have also been used in support of international protocol workshops, data merger studies, and time series analyses. • To facilitate the ocean color satellite validation analyses the NASA Ocean Biology Processing Group used Sea. BASS to compile a large set of coincident radiometric observations and phytoplankton pigment concentrations for use in bio-optical algorithm development. This new data set is called the NASA bio-Optical Marine Algorithm Data set (NOMAD) (Werdell, P. J. and S. W. Bailey, 2005)

Impact on Algorithm Development We need to define a minimum set of parameters. Under SIMBIOS, we where following the Protocol tables shown before…. .

NOMAD geographical coverage of the US continental shelf waters

What’s next ? • NASA Research Themes: – Carbon and Biogeochemistry, Ecosystems and Diversity, Habitats, Hazards and Health • Do we need a “new-revised” priority list of in situ parameters? Also, we should consider “what” will be able to be sampled in the near future (i. e. , with the advance of technology). • The in situ parameters must be “powerful”, i. e. , get us the most for our money • We will need agreed upon and defined protocols for these parameters, including QC

Complexity of Coastal Optics

. . “water color is determined by IOP, and Chl is just one of the active components that determine the IOP, therefore Chl can be determined only with a relative large uncertainty from ocean color remote sensing. ”…. . IOCCG Report 05

Issues in the continental shelf waters • Optically complex waters Need for separating the chlorophyll, CDOM, detritus, organic, inorganic particles, etc. • Spatial variability in coastal waters: – coherence length is much shorter – spatial sampling – requires adaptive sampling – concerns regarding sampling the same water mass • Bottom reflectance contamination • Coastal aerosols (atmospheric correction) • Vertical subsurface structure – surface signature is not coupled with profile – stronger color signatures - strong absorption in blue requires increased sensitivity

Some Remote Sensing Future Challenges • Atmospheric corrections – Absorbing aerosols & aerosol heights – “bright waters” in the NIR • Optically active components – c. DOM vs. pigment absorption – Particle concentrations--i. e. , backscattering coefficient of particles (bbp), beam attenuation coefficient of particles (cp) – Particulate inorganic carbon (PIC) and detritus vs. living particles – Particle size spectra – Taxonomic/functional groups

from Balch 2004.

Phytoplankton taxa display distinct long term patterns of variability Allen’s data 1916 -1950 Cross-correlation analysis of monthly anomalies of Southern Oscillation index (SOI) and SST (a); SST and phytoplankton abundances (b)

In situ coastal data • We expect higher uncertainties in these complex coastal waters because of bio-optical composition, instrument shadowing, optical signal extrapolation to just beneath the surface, increased atmospheric variability (aerosols) and increased BRDF variability. • High quality data are needed for both the vicarious calibration and product validation. These data must follow sampling, analysis, and protocol methods approved by the community.

Next “Sea. BASS” • Bio-optical calibration quality data sets (including atmospheric data? ). • Broad and diverse validation data sets – Better coverage of US continental shelf – Other critical areas • Should Sea. BASS data set contain information regarding the estimated uncertainties of the various variables ? • Should Sea. BASS be seen also as the long term archive ? If so could we envision a distributed system for initial QC development of new in situ parameters done by universities, etc. ? • Critical is the coordination among the distributed systems to agreed upon formats and standard fields, along with relevant information describing collection conditions, instrument specifications, etc.

Next “Sea. BASS” Sea. BASS database will then be used by a wider community a) Ocean color community (since 1997): – Calibration – Development & validation of algorithms – Other b) Ecosystem-based management community: – – Natural hazards: harmful algal blooms Pollution: estuaries, coast Biodiversity: coral reef Other c) Habitats, Hazards and Health, Modeling communities (1 and 3 -D): gathering community input….

• Marine biological and biogeochemical models remain, at a fundamental level, data-limited. This is particularly true at basin and global scales. • Most common in situ data sets used are JGOFS, NODC and recently, the underway p. CO 2 from COSP; satellite data (Sea. Wi. FS & MODIS ocean color) • Modelers need “gridded” data (available tools? resolution? ) • Present model parameterization and evaluation strategies: – Focus sites (HOTS, BATS, etc. ) – Ocean data assimilation (general circulation model and historical observations)

First modeling community input • All carbon variables (PIC, POC, DIC, DOC and p. CO 2) • Gross (GPP), net (NPP) or net ecosystem (NEP) • Backscattering data and particle size distribution • IOP’s • Fluorescence parameters • PAR, mixed layer depth • Phytoplankton functional groups

Road Map • Continuing the inventory of the oceanic parameter collected under NASA grants & data submission (AOPs, IOPs, HPLC, etc. ) to Sea. BASS Note the big bag of BROWNIES !

Road Map • Four workshops planned with the ocean color community on: 1) Future parameters for the Sea. BASS/NOMAD database: – Volume scattering function (“Backscattering”) and particle size distribution – Phytoplankton functional groups (What about physiological parameters ? ) – Carbon data set measurements (dissolved inorganic carbon (DIC), dissolved organic matter (DOM), particulate organic carbon (POC), particulate inorganic carbon (PIC), biogenic silica concentration (BSi), calcite, alkalinity, T, S, O 2, and related tracers such as CFC’s, 14 C, p. C 02, etc. ) – Sediment trap data (if export production is a future product) – Primary production (GPP, NPP), and PAR, mixed layer depth? – Sea surface temperature (SST) ? Nutrients ? – Other ? ? ?

Road Map 2) Revision of the “required” and “highly desired” in situ observations for ocean color system validation, and algorithm validation through an open workshop; 3) Measurements issues, any community agreement on: - processing (software # versions, definitions, etc. ) - protocols (quality assurance, QC, etc. ) - instruments (estimate of uncertainties, etc. ) - sampling packages with # instruments (# sampling frequency, etc. ). 4) Modeling (1 and 3 D) workshop to discuss and agree on: » » Parameterization of what parameters ? Evaluation of what parameters ? Gridded data and available tools Other Email: Paula or Giulietta (gfargion@chors. sdsu. edu)

Other issues • Team collection model or individual PIs ? • Some critical data sets not being measured adequately (IOPs); do we need an instrument pool ? • Protocol development: does it need to be continued? If so how? What topics remain open or inadequately addressed? – Coastal turbid water issues? – Protocols for the “new-revised” Sea. BASS parameters? – PP adopt of established protocols such as JGOF? • RR: should we have them again ?

Last point….