Ocean Color Time-Series Project: 2006 Annual Report

Major Accomplishments

1. Data Merger Report.  An international group of ocean color data merging experts convened and produced a report for the IOCCG on issues, challenges, and recommendations for data merging from concurrent ocean color satellites mission data.  The report will be published by the IOCCG in 2007.
2. Giovanni.  This user-friendly data analysis tool has generated wide acclaim for ease of use and versatility.  It continued to be enhanced in 2006, including access to L1 and L2 SeaWiFS data, 8-day L3 SeaWiFS data (including ASCII output), MODIS-Aqua SST data, and search-and-order and subset-and-get functionality,
3. LOCUS.  The Laboratory for Ocean Color Users (LOCUS) and Giovanni were added to the National Marine Educators Association online resources web site, “Bridge”.  It was the Teacher’s Top Pick in December 2006.
4. CZCS L1A Data.  The entire CZCS L1A data archive was cleaned of redundant data, placed into orbit format (instead of scene format) consistent with modern L1A ocean color data, and re-created.  It is available to the general public through the Ocean Color Web.
5. 9 Presentations at National Meetings, 1 Paper Accepted.  The meetings included AGU, AMS, Ocean Color Research Team, Western Pacific Geophysics Meeting, National Marine Educators’ Association, IGARSS, Ocean Optics.  The paper was in Eos, published Jan. 9, 2007.
6. SeaWiFS/Aqua Merged Data Products Completed and Acquired.  The output products coming out of the merging process are daily global 9-km resolution maps of: Chlorophyll concentration, colored dissolved and detrital absorption coefficient at 443 nm, particulate backscattering coefficient at 443 nm.  The data products were received by the DAAC.
7. Identified Biases in SeaWiFS and Aqua Global Mean Data.  These biases produce a false trend in the time series and suggest that consistent processing alone is insufficient to produce Clime/Earth Science Data Records in ocean color.  The results were presented at the AGU/ASLO/TOS meeting in Honolulu.
8. CZCS Re-processed 5 Times.  This involved processing from L1A through L3.  The second re-processing was made available to the general public at the Ocean Color Web.  Several errors in the data set have been found, and an additional 3 re-processings were performed.
9. Obtained Complete Archives of NODC and NASA In situ Chlorophyll Data.  These data sets are being used for data set quality and consistency evaluations.  Additionally, AMT underway data have been acquired but have not yet been formatted.
10. OCTS Striping Removed.  Thanks to consultation with JAXA scientist Hajime Fukushima, an error in processing code was found and corrected.
11. OCTS Processed Twice.  The first full processings of OCTS using modern algorithm methodologies were accomplished.  The data were made available to the general public at the Ocean Color Web site.
12. Continued ESDS Participation.  Consistent and active participation by team members in general ESDS activities and in the Reuse sub-group has been maintained.
13. Consistent OCTS-SeaWiFS/MODIS-Aqua Data Series Produced.  A new method was developed for producing C/ESDRs for sensors in the modern era (1996-present).  Using data assimilation methods and removing sensor biases, a climatically consistent time series was developed, using OCTS, SeaWiFS, and Aqua when it became available.  The time series agrees with global trends of SeaWiFS, even when using Aqua data instead of SeaWiFS (before this analysis, substitution of Aqua for SeaWiFS produced a false trend).  This represents the first climatically consistent time series of ocean color data using 3 different sensors.  The results were presented at the MODIS Science Team Meeting.

Specific Activities and Accomplishments for 2006, GSFC

Data Processing - CZCS

There were 5 re-processing events for the CZCS from L1A through L3 in 2006.  The first was an exercise to facilitate future evaluations.  The second was made available in the Ocean Color Web.  Unfortunately, several flaws were later detected in this data set:  invalid satellite angle calculation (causing the nadir location to be erroneously placed 100 pixels away), incorrect use of a constant aerosol type (preventing separation of aerosol from chlorophyll variability), a new calibration methodology that makes a priori assumptions of chlorophyll and radiances rather than measured radiances (that when used for SeaWiFS produced 20% differences in the global mean), Band 2 behavior (inconsistent hemispheric patterns beginning in fall 1982 and excessively high chlorophyll retrievals), electronic masking (using the poorly behaved Band 2, and also saturation effects continued to be allowed), artifacts arising from uncharacterized stratospheric aerosols from El Chichon in 1982.

A combined chlorophyll data archive from NODC (1955-present) and NASA (1996-present) was assembled to evaluate the performance of the CZCS.  The CZCS had log RMS=0.477, log bias=0.11, and r2=0.31, with a regression slope of 0.44.

The entire data set was reprocessed to rectify these problems, which are severe enough to prevent hope of an ESDR using CZCS data.  First we re-created the Gregg et al. (2002; Applied Optics) data set, which alleviates many of the errors by blending with in situ data.  It also eliminates problems with El Chichon aerosols by use of a stringent reflectance threshold.  Confirmation of this has been achieved through comparison with in situ chlorophyll, which indicates no degradation during the presence of the stratospheric aerosols.

Then there were 2 additional re-processings to attempt to rectify the remaining problems.  They differed in calculation of aerosols.  One used previously derived reflectance ratios in low chlorophyll and objectively analyzed in high chlorophyll regions.  This provided the spectral characteristics of the aerosols.  The aerosol amount was derived from known water reflectance at low chlorophyll, and used a bio-optical model at high chlorophyll.  The second version used he aerosol reflectance ratios as before, but additionally computed aerosol amount from objective analysis in high chlorophyll regions.  This had the practical value of avoiding assumptions associated with bio-optical models.

In both cases modifications were made to solve the errors identified in the public data set.  The satellite angle calculation was discarded and an older adaptation of spherical trigonometry was applied.  This solved the nadir offset problem.  Both of the above methods for aerosols provide variability in the aerosol fields, enabling clear observation of chlorophyll variability.  The older application of calibration was validated through comparison with in situ data and was re-applied.  Band 2 was discarded for use based on it anomalous behavior.  This immediately eliminated the excessively high chlorophylls that were previously obtained.  Discarding Band 2 also had the value of improving electronic masking.  Additionally we tightened the mask to completely eliminate any effect of saturation.  Finally the El Chichon effects were eliminated using a threshold on surface reflectance.  These results are currently under evaluation.

Data Processing - OCTS

The pervasive image striping problem in OCTS data was solved when JAXA scientist Hajime Fukushima found an error in the processing code.  The OCTS was the re-processed for its entire 8-month lifetime to L3.  The data are available at the Ocean Color Web

Data Access

• The search-and-order function for SeaWiFS data, connecting Giovanni to the OBPG archive, was promoted to operational status.
• MODIS-Aqua SST data (nighttime, 4 and 11 micron) were added to the Giovanni OPS system.
• Completed addition of the full-mission SeaWiFS 8-day data products to the ocean color Giovanni. This addition included the search-and-order capability provided by a link to the OBPG archive.
• The SeaWiFS 8-day data interface in Giovanni was enhanced with functions which had previously been available only for the SeaWiFS monthly data interface.  The two primary functions added are the "subset-and-get" connection to the OBPG data browser (allowing data orders for Level 1A and 2 subsetted data), and the frequently-requested ASCII output option for SeaWiFS 8-day data.
• Daily NOBM data from W. Gregg was added to the Ocean Color Giovanni interface. 
• The aph, bbp, and merged chlorophyll data from David Siegel and Stephane Maritorena (UCSB) were acquired.  S. Shen fixed the GrADS incompatibility in preparation for addition to Ocean Color Giovanni.  The data was placed in TS1 mode in preparation for public release, pending reprocessing, at the end of 2006.

Laboratory for Ocean Color Users

• LOCUS Tutorial Research Project 5, "Sea Surface Temperature vs. Chlorophyll Scatter  Plots, Part I" was promoted to the public Web site.   
• LOCUS Tutorial Research Project 6, "Sea Surface Temperature versus Chlorophyll Scatter Plots,  Part II: Where in the World (Ocean)?" was written and added to the LOCUS Web site.
• LOCUS Tutorial Research Project 7: Time-Series Analysis with Significance Testing Using Giovanniand Excel" was published on the LOCUS Web site.
• LOCUS and Ocean Color Giovanni were added to the "Bridge", National Marine Educators Association online resources Web site, on November 2.  LOCUS was the "Bridge" Teacher's Top Pick in December.
• "Locus Tutorial Research Project 8:  The NASA Ocean Biogeochemical Model (NOBM) Part 1: Examining the North Atlantic Bloom with Total Assimilated Chlorophyll" was published on the LOCUS Web site.

Presentations

• A poster on Giovanni, including references to the Ocean Color Time-Series Project interface, was presented at the American Meteorological Society Annual Meeting in Atlanta, Georgia in January.
• J. Acker presented the poster "Spatial and Temporal Coherence of SeaWiFS Chlorophyll Concentration Anomalies in the North Atlantic Bloom (1998-2005) Examined with Giovanni" at the AGU/ASLO/TOS Ocean Sciences meeting in Honolulu, Hawaii, on February 23. 
• Poster presented at the AGU/ASLO/TOS Ocean Sciences meeting in  Honolulu on February 22 entitled: "Global and Regional Trends in  Ocean Chlorophyll as Viewed by SeaWiFS and Aqua" (Casey/Gregg).
• J. Acker presented a demonstration of Giovanni at the Ocean Color Research Team meeting and discussed Giovanni with several attendees at the meeting.  These discussions resulted in the creation of an "ad hoc" ocean color Giovanni science advisory group, and submission of a presentation proposal to the National Marine Educators Association national conference, which was subsequently accepted.  This presentation took place in mid-June and was a "hands-on" demonstration in a computer laboratory environment.  The AGU Oceans poster on North Atlantic Bloom climatological chlorophyll anomalies was also presented at the OCRT meeting.
• The poster "Observing Increased Chlorophyll-a in Storm Wakes for the 2005 Atlantic Hurricane Season Using 8-Day Data Products in Giovanni" (Suhung Shen, James Acker, Greg Leptoukh, Hualan Rui, Steve Berrick, and Steve Kempler) was presented at the AGU Joint Assembly in Baltimore, MD.
• On July 17, J. Acker presented "Giovanni and LOCUS: Innovative Ways for Teachers and Students to Conduct Online Learning and Research with Oceanographic Remote Sensing Data" at the National Marine Educators' Association annual meeting in Brooklyn, NY.  This presentation resulted in positive comments from the attendees, as well as a potential collaboration with the University of Southern Mississippi Marine Education Center and Scott Aquarium.  This collaboration may also lead to increased use of marine remote sensing data at Centers of Ocean Sciences Education Excellence (COSEE) workshops held there. Prior to the meeting, a message to the Giovanni/LOCUS mailing list offering an electronic version of the presentation resulted in over 130 requests.  Subsequent to the meeting, an animated and non-animated version of the presentation was placed on the LOCUS Web site.
• Suhung Shen presented "Seasonal Variation of Chlorophyll-a at Luzon Strait and South China Sea" at the Western Pacific Geophysics Meeting in Beijing, China, on July 25.  This presentation highlighted the use of Giovanni and SeaWiFS 8-day data.
• G. Leptoukh presented "Terra, Aqua and Aura Data Exploration, Visualization and Analysis via Giovanni.", including Giovanni ocean data visualizations, at IGARSS '06 in Denver, Colorado.
• The Ocean Optics XVIII extended abstract entitled "Use of Giovanni with Ocean Color Time-Series Project Data for Trend Detection in the Coastal Zone"  by J.Acker, W.Gregg, G.Leptoukh, S.Kempler, N.Casey, G.Feldman, C.McClain, W.Esaias, and S.Shen was augmented with p-significance tests of each generated time-series provided by N. Casey.
• J. Acker presented the poster "Use of Giovanni and Ocean Color Time-Series Project Data for Trend Detection in the Coastal Zone" at Ocean Optics XVIII in Montreal, Canada.  Useful discussions took place at the meeting, particularly discussions of the potential for adding a euphotic zone depth or Secchi depth product to Giovanni.  Several ideas for Giovanni research collaborations were explored.
• The initial draft of the poster "Using Giovanni to Examine Patterns and Trends in the Coastal Gulf of Mexico" was composed and provided to GES DISC colleagues for review.  The poster will be mailed for display at The Gulf of Mexico Alliance Nutrients Criteria Conference, Gulf Breeze, FL, January 16-18, 2007.

Data Merging

A workshop on ocean color data merging was held in Bethesda, MD, USA on May 10 and 11, 2005.  The workshop was sponsored by the IOCCG.  Watson Gregg (NASA/GSFC) and Paula Bontempi (NASA/HQ) chaired the workshop.  The participants were: 

• Jim Aiken, Plymouth Marine Laboratory 
• Ewa Kwiatkowska, NASA/GSFC 
• Stephane Maritorena, ICESS/UCSB 
• Hiroshi Murakami, JAXA/EORC 
• Frederic Melin, JRC  Claire Pottier, LEGOS/CLS  
• Simon Pinnock, ESA

The final report from the workshop was delivered to the IOCCG on Dec. 30, 2006.

ESDSWG Participation

Participated in conference calls and E-mail discussions of the NASA Earth Science Data Systems Working Group (ESDSWG) Software Reuse Working Group.  Completed third paper review for possible listing on the Reuse Portal.  Reviewed Reuse Enablement System requirements document and made comments.  Suggested ideas for speakers/invited presenters for the the upcoming Joint Working Group and AGU meetings.