120 likes | 226 Views
Comparisons of Argo profiles and nearby high resolution CTD stations. Megan Scanderbeg. Motivation. Investigate how much recent data is in the reference database used for DMQC and/ or how much recent CTD data is available Find out how paired Argo and CTD profiles compare
E N D
Comparisons of Argo profiles and nearby high resolution CTD stations Megan Scanderbeg
Motivation • Investigate how much recent data is in the reference database used for DMQC and/ or how much recent CTD data is available • Find out how paired Argo and CTD profiles compare • Increasingly important to have recent, high quality CTD data for delayed mode quality control considering recent pressure offset concerns and increasing use of Argo data in global change analyses • Comes from an Action item from AST-10: M. Scanderbeg will work with S. Diggs to find number of collocated shipboard CTD from 2004-2008 and Argo profiles
Where to get CTD data? • Started with current reference db (CTD_for_DMQC2008V01), but very few stations in 2004 or later (~200) • Worked with S. Diggs at CCHDO to get access to most recent CTD data for Argo DMQC use only • Downloaded all publicly available CTD data on CCHDO website from 12/2003-12/2008 • Combined for a total of 1543 stations Graph from C. Coatanoan’s presentation on the ref db from DMQC3 showing temporal distribution of CTD data in ref db
Other sources of CTD data • Requested WOD data from NODC’s Search and Select website (http://www.nodc.noaa.gov/OC5/SELECT/dbsearch/dbsearch.html) on 7/23/09 using the following criteria: • Ocean station data and high resolution CTD/XCTD • 01/01/2004 through 04/31/2009 • Temperature and salinity data • All casts deeper than 1000 meters • Received 17,849 stations, refined to 15,169 CTD only stations • Several of the stations are repeats of data from CCHDO • Ignored overall qc variable assigned to temperature by WOD as several (1125) profiles failed monthly, seasonal, and/ or annual checks resulting in “bad” overall qc variables even though individual flags on individual measurements were good
CCHDO stations not in WOD • Searched for duplicate CCHDO data in WOD database by latitude • +/- 0.1, longitude +/- 0.1 and day +/- 1 • at least 236 stations from CCHDO only
Comparing Argo profile to CTD profile • Looked for matches within “scaled distance” of 100 km and 10 days • Found 2361 matches • Comparison method: • Moved Argo profile up and down by 40 dbars compared to CTD profile • For each shift of the Argo profile: • For each Argo temperature measurement, interpolated to corresponding smoothed CTD temperature and found the difference • For each pair: • Minimized the sum of the squares of the temperature differences scaled by the number of differences for each pressure shift • Found the corresponding pressure offset based on the results of minimizing the sum of the squares • Included the pressure offset only if the minimum was greater than 5 dbar from the end of the shift to try to include cases where a true minimum was found (609 cases where within 5 db of end)
Conclusions on CTD data • Recent CTD data is available both through CCHDO (public and private) and WOD/NODC to be added to the QC reference database • Ideally, an update for the database would be available each year with the cruises from both centers in the past year. • Desires for the yearly update include: • One format for all data that is easily incorporated into current database • List of the cruises included in the database • How best to accomplish this goal?
Conclusions on comparing profiles • Decent number of matched Argo-CTD pairs, especially in 2004-2007. • Coverage in time (2008) and space could still be improved • 2007 mean is a bit larger in magnitude than others years. Need to wait and see if this continues for 2008 as more CTD data becomes available • Could do same investigation by float type to look for differences • Method used to find pressure offset does have the potential concern that the thermocline could be dominating the calculation