Marine Data Literacy 2.0

Providing instruction for managing, converting, analyzing and displaying oceanographic station data, marine meteorological data, GIS-compatible marine and coastal data or model simulations, and mapped remote sensing imagery





Home >2. Marine GIS > 2.18 Navy Waves

2.18 Discontinued US Navy Marine Atlas: Global Wave Grids

  • Exercise Title:  Discontinued US Navy Marine Atlas: Global Wave Grids

  • Abstract:  Until this writing, the author has never been able to find a free or cost-of-reproduction, usable global ocean wave climatology.  The relevant OceanTeacher Academy course lesson on wave climatologies lists many commercial or otherwise restricted possibilities, and one very strange collection too poorly documented to be useful.  After much trial-and-error testing, the out-of-print Marine Atlas (below) has yielded up its secrets, and the average wave height grids can now be plotted in Saga.  [You must find a copy of 1992 original CD. There are many other data analyses on the CD, and you can use the same methods for them.]  The method below uses the 1-degree grid products, but the CD also includes 5-degree versions.  These methods and data are provided only for training purposes, and the products (statistics and/or graphics) must not be used for navigation or any other private or public purpose.  Specific grid registration is only approximate due to lack of documentation.

  • Preliminary Reading (in OceanTeacher, unless otherwise indicated):

    • N/A

  • Required Software:

  • Other Resources: 

  • Author:  Murray Brown, using materials originally published on the referenced CD-ROM.  Due to the general unavailability of this resource, and to its almost unique nature, the author has transgressed into the murky area of republishing data.  He hopes you will forgive him for this lapse.  [Only one of the data layers on the CD is provided here.]

  • Version:  8-12-2014

1.  Open the CD with Windows Explorer, to find this toplevel folder set.
2.  Open the 1DEG folder to find these 12 monthly folders.
3.  Open the January folder (01) to find these products.  Nearly all of them come in pairs with similar names.  The difference is either the letter A or the letter U.  See below for a possible explanation.

Most of the DAT files are "128 KB".

The files with WVH in their names look suspiciously like wave height products.


4.  If you have the CD, you can work directly from it.  If you don't, then save the "A" files above to the folder DATA > OCEAN > NAVYATLAS, and use them whenever the U files are discussed.
5.  An old data manager's trick is to open binary files in an ASCII editor to see if they contain an initial, human-readable header.  In this case, in file MWVHU01S.DAT, there is a header, and it confirms our suspicions about the contents:
  • The grid dimensions are apparently 361 X 181
  • The data are obviously binary after the initial ASCII header
  • The lower left latitude appears to be -90; lower left longitude appears to be -180.  But the actual "edges" could be 0.5 degree "outside" depending on cell geometry.  Later experiments might be needed to find out.
  • The data resolution ("dim") is apparently 0.01 meter
  • The type says SMOOTHED, but this will be refuted later by other data.
6.  You can use any good ASCII editor to identify the other grids.  For example SWVH (in the final two files) indicates standard deviation of the wave height.
7.  Run Saga


8.  This options menu is very big, and complicated.  We'll take it one step at a time below:
  • For RAW DATA FILE, navigate to and select the U file (i.e. MWVHU01S.DAT.  [If you don't have the CD, then use the A versions of the files, i.e. the ones available above.]
  • For CELL COUNT (X), insert 361, because there are 361 longitudes between -180 and +180, if you count 0 as a separate line
  • For CELL COUNT (Y), insert 181, for similar reasons
  • For CELL SIZE, insert 1, because we already know this is a "1DEG" data product.
  • For UNIT NAME, insert METERS, which is the usual unit for reporting waves
  • For Z MULTIPLIER, insert 0.01, because we think that is the meaning of DIM in the ASCII header
  • For NO DATA VALUE,  we have to leave that alone for now, because there's no help in the header
  • For DATA OFFSET (BYTES), we need to insert the exact size of the header.  What is it?
9.  To calculate the exact size of the header, first we need to use Windows Explorer's PROPERTIES function to find the exact size of the file.  It is 130,850 bytes.

NOTE:  You cannot use the rough value of 128 KB or the size on disk of 131,072 byte.  Only the exact value.

10.  Now we need to find out what type of digital number has been used to store the grid.  You can get a rough estimate by dividing the size of the file by the size of the grid:
  • 130850/(361*181)=2.00257

So the grid values are probably 2-byte values.  The slightly larger number here is due to the presence of the tiny little header.

11.  If the grid values are 2-byte numbers, then here is how you calculate the exact size of the header:
  • File size minus grid size
  • 130850-(361*181)*2=168

So we have 168 bytes of ASCII text at the beginning of the file.

12.  Now we can enter the DATA OFFSET (BYTES) value of 168.

For DATA TYPE, we can select 2 BYTE INTEGER (UNSIGNED) as a first guess, because we already know it has to be 2 bytes, and there can't be negative wave heights.

The other items, LINE OFFSET, LINE ENDSET, BYTE ORDER and LINE ORDER, we can leave unchanged and just hope for the best.  If these settings don't work, then you would have to perform experiments with each one to optimize the settings.  In our favor, these are extremely common selections, and should work 99% of the time.

Now click OK.

13.  A new grid object appears in Saga, and you can use ADD TO MAP to display it.
14.  It looks OK, except for some issues.
  • If you move your cursor over the dark brown continents, all "wave height" values are 65535, so this must be a NO DATA VALUE
  • The wave height values at sea are in the hundreds of meters, instead of the expected 0-6 meters, so our Z MULTIPLIER is too large by a factor of 100
15.  Make these change to the setup for the module, then run it again.
16.  Now you have perfectly reasonable wave height values, and the land areas are white (i.e. blanked out).

The figure LEGEND has also been shown, activated by clicking the tab at the bottom of the properties panel.

17.  Select the new data grid and select HISTOGRAM to see the distribution of values.
18.  This is the histogram for the January data in the file labelled with a U.  Notice that it is "noisy", with lots of highs and lows.

Experiments with the original data file labeled "A" indicate that it might mean "averaged" and it looks a bit smoother.  So "U" might mean "un-averaged" You can explore this for yourself later.

23.  Select SAVE AS and then navigate to the folder PRODUCTS > SAGA > GRIDS and save this final product grid as mean_wave_ht_global_january_navy_atlas_ncdc.sgrd.  Take the time later to make a full 12-month collection of these average grids (either U or A, as you wish) for climatological work.
24.  Protect yourself and your organization by making sure that all products made from this source are labeled "Should not be used for navigation or any private or public purpose, other than data management training.  Specific grid registration is only approximate due to lack of documentation."