TITLE: Presentation of 2002 SBI Field Year Hydrographic Data AUTHORS: Charles N. Flagg Bldg 490D Brookhaven National Laboratory Upton, NY 11973 Ph: (631) 344-3128 e-mail: flagg@bnl.gov I'm moving to the State University of New York at Stony Brook as of March 15, 2004. There my address will be: Marine Science Research Center Endeavour Hall, Room 203 Stony Brook University Stony Brook NY 11794-5000 Ph: (631) 632-3183 e-mail: cflagg@notes.cc.sunysb.edu (tentative) Lou Codispoti Horn Point Laboratory, 2020 Horns Point Rd P.O. Box 775 Cambridge, MD 21613 Ph: (410) 221-8479 e-mail: codispot@hpl.umces.edu Jim Swift Scripps Institution of Oceanography Physical Oceanography Research Division: 0214 9500 Gilman Dr La Jolla, CA 92093 Ph: (858) 534-3387 e-mail: jswift@odf.ucsd.edu FUNDING SOURCE: National Science Foundation grant to Scripps Institution of Oceanography, OPP-0125399 DATA SET OVERVIEW: In the SBI 2002 field year there were three cruises to the Chukchi and Beaufort Seas between May 5 and August 26 two process cruises on the USCGC Healy and one mooring cruise on the USCGC Polar Star. During these cruises a total of 375 hydro casts were made covering the continental slope in a series of cross-isobath sections from east of Herald Canyon to the Beaufort shelf and over part of the eastern Chukchi Sea (Figure 1). In order to get a first comprehensive look at this large amount of chemical and physical data an effort was instigated to provide profiles, property/property plots, and vertical sections. The results include some 2000 color and black and white plots. Figure 1a, Cruise track and station plot for the Shelf Basin Interaction process cruise HY0201 on the USCGC Healy from May 8 to June 17, 2002. Figure 1b, Cruise track and station plot for the Shelf Basin Interaction process cruise HY0203 on the USCGC Healy from July 16 to August 26, 2002. Figure 1c. Station plot for the Shelf Basin Interaction/Arctic West Summer 2002 mooring cruise AWS02 on the USCGC Polar Star from July 15 to August 13, 2002. The CTD and bottle data from all cruises were exhaustively quality checked by the SBI hydrography service group prior to inclusion in the plotting process. The CTD data were collected using Sea Bird 911 CTDs with dual pumped temperature and salinity sensors, auxiliary fluorescence and transmission/turbidity sensors, and on the Healy a single pumped dissolved oxygen sensor, all mounted on a Sea Bird rosette. The bottle data were collected in 30 liter Niskin bottles during HY0201 and HY0203 and 10 liter bottles during AWS02. The ASCII data files were then scanned, decoded and loaded into MATLAB files with all the subsequent processing was done in MATLAB. The variables available from each cruise are shown in Table 1. The variable, N**, is a derived parameter that shows excess nitrogen above the Redfield ratio and is a useful water mass indicator. The relationship is given by: Vertical profiles were generated from downcast CTD and bottle and downcast CTD data for each variable measured, including all the casts made at each station. The data were then divided into from 5 to 7 logical groups of stations to form sections as shown in Figure 1. The section plots include "waterfall" plots showing vertical profiles usually from the deepest cast at each station in the section. These plots use a logarithmic depth scale that expands the upper part of the water column. Property/property plots are shown for the primary nutrients and salinity, for N** versus salinity and for temperature versus salinity. The vertical section plots were produced using kriging to interpolate the irregularly distributed observations onto a uniform grid for contouring. These plots show the color (or gray scale) and line contours of the variable in the upper of the two panels and the kriging standard deviation in the lower panel. Kriging is a linear interpolation scheme whose aim is to minimize the prediction error variance ((Isaaks and Srivastava, 1989). This is accomplished by calculating the sample spatial covariance, or variogram, and determining an analytic function that best fits the observed dependence. The covariance function is then used to determine the weights that will minimize the overall interpolation error. The plots of kriging standard deviation indicate areas where, under the assumptions of the methodology, there is greater or less confidence in the prediction. The kriging was performed using the EasyKrig software developed by Dezhang Chu for the NW Atlantic GLOBEC program which can be found at http://globec.whoi.edu/software/kriging/easy_krig/easy_krig.html. Table 1. Tabulation of variables measured/calculated and plotted for each cruise Variable HY0201 HY0203 AWS02 CTD Data Temperature X X X Salinity X X X Sigma-t X X X Sigma-Theta X X X Pot. Temp. X X X CTD Oxygen X X Fluorescence X X X Haardt Fluor. X X Transmission X X Turbidity X Bottle Data Bottle Oxygen X X Nitrate X X X Nitrite X X X Urea X X Ammonium X X N** X X Phosphate X X X Silicate X X X DATA FORMAT: The plots are available in PNG format. PNG is a successor of GIF as a plotting format that gives better resolution than either TIFF or JPEG while taking less storage space. The extracted hydrographic data is available in two mat-files per cruise, one each for the bottle and CTD data, in which the variable names are fairly self-explanatory. REFERENCES: Isaaks, E.H . and R.M. Srivastava, 1989. An Introduction to Applied Geostatistics, Oxford University Press, New York, NY, 561pp.