GEOPHYSICAL RESEARCH LETTERS, VOL. 15, NO. 10, PAGES1117-1120, SEPTEMBER 1988 SPECTRAL ANALYSIS OF CANADA BASIN UNDER-ICE DRAFT DISTRIBUTION AS RECORDEDBY THE USS QUEENFISH, AUGUST 1970 J.R. Key and A.S. McLaren University of Colorado, Boulder Abstract. Spectral analysis has been applied to the importance of the longer wavelengths in the 940 kilometers of under- ice thickness data Fourier analysis. recorded by the USS QUEENFISH in the Canada Basin, This paper reports the results of a spectral August 1970. Periodicities of 50-150 m most analysis of approximately 940 kilometers of under- commonly occurred throughout the track, with ice profile along longitude 155ø-00.0 ' W, between longer wavelengths commonin areas of thicker, latitudes 74ø-22.5 ' and 82ø-55.0 ' N (Figure 1). more variable ice. While some regional variations Our objectives were to identify the dominant could be discerned, unique regional patterns were periodicities in the draft distribution and to not apparent, supporting the hypothesis that ice further test the hypothesis that ice conditions in characteristics in this region are relatively the Canada Basin are relatively homogeneous. homogeneous. During early August 1970, the USS QUEENFISH / reoore u.er-oe topography of the Canada Basin along the 155 West meridian between latitudes 74-00.0 and 83-30 degrees North. Subsequent statistical analysis of the acoustically recorded under-ice draft distribution confirmed ship s observations that > i•-••-••• the under-ice topography in this region was relatively uniform and quite moderate in draft Beaufort Sea was also noted by Wadhams [1980]. McLaren [1986, 1987 and 1988] further determined s that the sea ice of this area, within the central part of the Beaufort Gyre, may be thinner and more open than elsewhere within the central Arctic Basin. examine further the under-ice profiles recorded by • • •, . QUEENFISH through the Canada Basin [McLaren, 1986] to determine what other indications of Figure 1. Track of the USS QUEENFISH in early enviro•ental forcings might be found. Visual examinations of the profile data indicated August, 1970 across theCanada Basin. possible spatial periodicities. Accordingly, spectral analysis wasused to describe the shape Data andMethods and apparent non-randomness of the under-ice draft profiles which QUEENFISH recorded within this QUEENFISH continuous analog under-ice draft particular area. recordings and supporting navigational logs were The morphology of sea ice has beenpreviously obtained from the Arctic Submarine Laboratory, analyzed statistically by Rothrock [1979, 1986], U.S. Naval Ocean Systems Center, SanDiego. To Hibler [1980], Thorndike et al. [1975], Wadhams recordthe under-ice topography, QUEENFISH used a and Horne [1980], Wadhams [1981], McLaren [1986] narrow-be•, 205 •z, upward-beamed acoustic and others. Hibler and LeSchack [1972] applied profiler of an •/BQS-8 sonar systemwith a spectral analysis to the undersea ice profile in footprint diameter of 2.68 m. •er three million the cemtral Arctic. •ey found periodicities of data points were obtained from the manually 56 and 82.5 m to be significant. Kozo and Tucker digitized analog recordings for subsequent [1974] applied Fourier analysis to sonar data in statistical analysis. •e data were then the De•ark Strait, from the ice edge to the the interpolated to 145 cm inte•als in order to Greenland coast. They found an increase tn ice ensure a more balanced representation of under-ice thickness variability with increasing distance thickness. The accuracy of the submarine's from the ice edge and a correspondingincrease in acoustic profiler is •15 cm at best. The 940 • track was divided into 5 subsections with data points averaged over 10 m. This point spacing and subsection length Coypright 1988 by the •erican Geophysical Union. combination were found to provide an adequate balance between resolution within the range of Paper n•ber 8L8094. wavelengths under study (20-2000 m), the relative 0094-9276/88/008L-8094503.00 homogeneity of subsections, and computational 1117