11/17/00 Changes for Version 8.2 In versions through 7.2 we used the filters and inverse Nettleton step as described in our 1994 paper, with "polishing" as described in our 1997 paper. Polishing means that a cell with one or more soundings is set to the median of soundings in the cell, land data are set to their median, shorelines and reefs are set, etc. In short, the solution from gravity is found by 1994 method and then it is perturbed to fit the constraints. In version 8.2 we began testing a new way of getting the solution from gravity, a step before polishing. After this new way, we polish as usual. The new way is this: Assume a standard density for the topography (2.67 gm/cm**3) and do an iterative inversion (a la Oldenburg, 1974; Muller and Smith, 1993) to invert the non-linear Parker (1973) expression, obtaining a band-pass-filtered topo, h, which fits the band-passed and downward-continued gravity, g, under the assumption that the density is 2.67. The purpose of this is to capture the non-linearity which is important at tall seamounts, and so to get the amplitude of these seamounts more correct. The old way this was sort of fudged later because the nettleton step would get a scale factor which would boost their amplitude, but this was ad hoc and not very correct. Now test the fit of this h to the observed band-passed soundings under three cases: Case 1: use this h as is; Case 2: use this h but scaled by s; h = h*s; find best-fit s; Case 3: use h scaled by sp if g > 0, sn if g < 0; find best sp and sn; Now do an F test for significance of improvement of fit, and test also against case that s = 0 (i.e., no prediction at all). The solution is the case that gives the most significant improvement. If prediction is not significantly different than no prediction at all, predict nothing. The reason for doing this step is three-fold: First, as explained under non-linearity, the old way coped with non-linearity by having an ad hoc scale factor which had nothing to do with physical density. This way handles non-linearity and makes a physically plausible density, and then only rescales this if a different density would fit significantly better. Second, the old way had only one scale factor, but as the 1994 paper showed, there are situations where g > 0 has exposed ridges in topo, but g < 0 has troughs buried under sediment, so two scale factors are needed, depending on the sign. This method allows for that, but only if it would fit significantly better. Third, this method tries to cut down on the "orange peel" texture in abyssal hills, by fitting a non-zero model only where it is significant. The idea is to avoid fitting noise in the gravity. Eventually there should be a paper on this. We were going to wait until we had a 1-minute version. Oh, yes, the version 8.2 also starts from version 9.1 gravity, not 7.2 gravity. 9.1 grav has more short-wavelength amplitude and fixes an edge effect problem where profiles approach land. 09/17/00 Changes for Version 7.2 The main difference between Version 7.2 and Version 6.2 is that about 40% more ship soundings were added in the final polishing step. The complete and new ship soundings are summarized next. Also DBDBV the grid on the west coast of North America was not used. SUMMARY OF DATA HOLDINGS VERSION 7.2 and 8.2 BATHYMETRY **** marks data added since Version 6.2 SHIP PROFILES: 1) WS data base ws_all_legs - 2334 all of the legs of WS data for V6 ws_good_legs_V6 - 2165 good legs of WS data ws_good_legs_V7 - 2145 good legs with some editing from V6 2) SIO data base all_sio_legs - 1589 all legs of SIO data for V6 sio_good_legs - 1410 good legs of SIO data 3) NGDC data base all_ngdc_legs - 2066 all legs of NGDC data for V6.2 ngdc_good_legs - 1244 good legs of NGDC data ****ngdc_new_unique - 1759 legs added to NGDC since V6 ngdc_good_legs_V7 - 2778 good legs through Dec 1999. 4) BB data base bb_good_legs - 118 good legs of BB not in WS bb_legs_not_in_ws - 973 legs of BB in WS not used for V6 5) PROP data base prop_legs - 99 legs of PROP data prop_good_legs - 57 good legs of PROP data used in V6 ****prop_legs_new - 23 some new ones from SIO ****prop_polar_legs - 23 new legs from polar programs prop_good_legs - 94 good legs MISC XYZ DATA 1) found_block.xyz - Foundation seamount data from multibeam Sonne 2) ****ifremer_data - 92 new cruises of Atalante data from Ifremer, center beam only. 3) ****coral_new - a directory with an ascii file of reef locations 2 min. 4) W09.xyz and W10.xyz Indian ocean data from the Phipps-Morgan Orcutt cruise to the AAD. 5) ****sebazil.yxz - a short file of brazil depth soundings 6) ****bahamas.xyt - soundings from bahamas 7) ****Knorr_bathy - a directory with two matlab files of Knorr bathymetry for an area in the North Atlantic 8) ****NGDC_CRM - ship soundings from the east coast of the US from a 1 minute grid of soundings. From Lincoln Pratson. Probably very good. 9) ****NIWA - ship soundings from New Zealand used by NIWA to make grid. 10) ****roest - 11 files of Vining Meinez Lab data from the North Atlantic. 6/16/97 Changes for Version 6.2 A. Eliminated bad sounding in Indian Ocean at 14N 62E. This created a large seamount that does not exist in the gravity anomaly and thus probably does not exist. B. Added SCAR coastline constraint to replace high resolution GMT coastline around Antarctica. C. Added GEBCO contours from Weddell Sea. This fixes the incorrect location of the continental shelf on the east side of the Antarctic Peninsula. D. Added grids from "Digital Bathymetric Data Base - Variable Resolution" DBDBV Version 1.0. The areas from DBDB-V are: Mediterranean Sea Black Sea Red Sea Persian Gulf East Pacific Ocean longitude > 140W, latitude 29N-45N Baltic Sea longitude 15E-25E 30N-48N E. References for DBDB-V: Data Base Description for DBDB-V, Version 1.0, Naval Oceanographic Office, March, 1996. DoD Directive 8320.1, DoD Data Administration, Draft, 26 September 1992 (NOTAL). Naval Oceanographic Office Data Model, Hydrographic/Bathymetry, Deaft, latest applicable version. 5/29/97 This improvement of version 5.2 has high resolution land elevations derived from the 30 second topography provided by the USGS EROS data center. The 30 second data were blockmedianed to 2 minutes on a Mercator grid and then merged with the ocean and shoreline constraints. Chris Small provided cleaned up versions and guidance in the Addidtion of the GTOPO30 data. The references for the GTOPO30 is: http://edcwww.cr.usgs.gov/landdaac/gtopo30/gtopo30.html Here is a description of the data copied from the above WWW site: *********************************************************************************** GTOPO30 is a global digital elevation model (DEM) with a horizontal grid spacing of 30 arc seconds (approximately 1 kilometer). GTOPO30 was derived from several raster and vector sources of topographic information. For easier distribution, GTOPO30 has been divided into tiles which can be selected from the map shown above. Detailed information on the characteristics of GTOPO30 including the data distribution format, the data sources, production methods, accuracy, and hints for users, is found in the GTOPO30 README file. GTOPO30, completed in late 1996, was developed over a 3 year period through a collaborative effort led by staff at the U.S. Geological Survey's EROS Data Center (EDC). The following organizations participated by contributing funding or source data: the National Aeronautics and Space Administration (NASA), the United Nations Environment Programme/Global Resource Information Database (UNEP/GRID), the U.S. Agency for International Development (USAID), the Instituto Nacional de Estadistica Geografica e Informatica (INEGI) of Mexico, the Geographical Survey Institute (GSI) of Japan, Manaaki Whenua Landcare Research of New Zealand, and the Scientific Committee on Antarctic Research (SCAR). ************************************************************************************* 5/21/97 This new version has two main improvements: A. more complete shoreline B. no false islands offshore Please provide feedback on theoffshore depths and the match or mismatch of the shoreline to other grids. 9/13/96 This directory contains global mercator grids of topography based on a variety of sources. Smith, W. H. F. and D. T. Sandwell, Global Seafloor Topography from Satellite Altimetry and Ship Depth Soundings, submitted to Science, April 7, 1997. Smith, W.H.F. and D. T. Sandwell, Bathymetric prediction from dense altimetry and sparse shipboard bathymetry, J. Geophys. Res., 99, 21803-21824, 1994. The gravity grid is described in the following publications: Sandwell, D. T. and W. H. F. Smith, Marine Gravity Anomaly from Geosat and ERS-1 Altimetry, J. Geophys. Res., in press, 1997. Sandwell, D. T. and W. H. F. Smith, Marine Gravity from satellite Altimetry (poster), The Geological Data Center, Scripps Inst. of Oceanography, La Jolla, CA 92093, (digital file, Version 7.2) anonymous ftp to baltica.ucsd.edu, 1995. topo_polish_5.2.img.gz -Predicted depth using ship soundings to constrain the long wavelengths and Geosat/ERS-1 gravity anomalies to constrain the short wavelengths. Pixels that are constrained by ship measurements or coastline data are set to the measured values. Then the difference is splined using 10 iterations of the biharmonic operator to make a smooth transition from the predicted pixels to the measured values. (compressed with gzip) topo_polish_5.2.img.ers - ER-Mapper header for topo_polish_5.2.img. -rwxrwxr-x 1 sandwell 1007 May 21 18:13 COPYRIGHT -rwxrwxr-x 1 sandwell 3586 May 29 15:21 README -rwxrwxr-x 1 sandwell 16418 May 21 18:13 diskio.c -rwxrwxr-x 1 sandwell 8138 May 21 18:13 img2xyt.f -rwxrwxr-x 1 sandwell 96 May 21 18:13 makefile -rw-rw-r-- 1 sandwell 136857600 May 29 15:24 topo_polish_5.2.img -rwxrwxr-x 1 sandwell 869 May 21 18:13 topo_polish_5.2.img.ers The gridded data are stored in an integer*2 format without any header or record information. *.img A 6336 by 10800 grid of 2-byte integers = 136,857,600 bytes. Byte order is big_endian. The topography is meters above sea level. An even value signifies the cell does not have a ship or coastline measurement while an odd value signifies that it does. The Mercator projected image spans longitudes from 0 E to 360 E and latitudes from 72.006 N to -72.006 N. A spherical earth is used for the Mercator projection. The center of the upper left grid cell (i.e. the first integer in the file) is located at 72.0009 N, .01667 E. Longtiudes increase with a 1/30 degree spacing. The The center of the last integer in the file is located at -72.0009 N, 359.933 E. Note the latitude spacing is 1/30 degree at the equator but decreases as 1/cos(latitude) according to a Mercator projection on a sphere. The files can be accessed either with the program img2xyz or the GMT program img2mercgrd. In addition, it can be used with image processing programs such ER-Mapper or GIPS.