Not yet, Espen. I'm hoping to get more people engaged in these files now that the football tournament is over (though there is always some other sport to replace it).
I'll post a couple rounds of Cresis profile analysis specific to Jakobshavn Isbrae here but if there's interest, start a separate forum section for ice-penetrating radar, a huge deal in Greenland and Antarctic and critical to Arctic Ocean ice thickness measurement validation (ie volume).
These profiles, in my view, are seriously under-interpreted and in some cases like Petermann Glacier, demonstrably mis-interpreted. The intermediate stratifications and their deformation are exceedingly important to ice sheet history and flow properties; it is not enough simply to read off surface elevations and bedrock depths if the goal is estimating future sea level rise attributable to Greenland by prioritizing to the fastest marine outlet glaciers.
After visiting numerous dead '404' urls, I did eventually find a 'FAQ' among the read-me Cresis files. It did not have any examples of annotated profile features but did explain their rather redundant data storage system and contained some interesting factoids about ice-penetrating radar.
<url>
ftp://data.cresis.ku.edu/data/rds/rds_readme.pdf</url>
The first thing to understand is Cresis uses MatLab formats but you probably do not: $2250 per individual license, $149 for home. The home use license does not include government, academic, commercial, or other organizational use (blog?).
There'd be a learning curve: "MATLAB is a high-level language and interactive environment for numerical computation, visualization, and programming. Using MATLAB, you can analyze data, develop algorithms, and create models and applications." <url>
http://www.mathworks.com/products/matlab/</url>
However, the data is also available as desktop-readable triples of track location maps (file names ending in _0maps.jpg), radar return profiles (_1echo.jpg), and interpreted tracks (_2echo_picks.jpg). The latter are 'manually driven processes' where a trained individual marks up surface and bottom reflections with purple and red lines respectively. (However bedrock cannot always be located for the Jakobshavn gorge.)
Those depths are captured into excel-readable cvs numerical format. Of the 9 columns, 5 are useful (lat, lon, surface, bottom, elevation), 1 is easily derived (thick = surface - bottom), and 4 can be deleted. Cresis data is carried to excessive precision -- surely the bedrock is not really measured to centimeter accuracy as numbers like 2087.12 suggest. And surely latitude is not usefully measured to 6 decimal points (0.11 m) when profiles show 5 m precision (see below and <url>
https://gis.stackexchange.com/questions/8650/how-to-measure-the-accuracy-of-latitude-and-longitude</url>
The 15 years of flight have generated a lot of files. After drilling down to Greenland, look for the Jakobshavn Isbrae specific folders such as 09_01 (transects) and 09_02 (tracks along the icestream) for the given date, here Apr 14. Each profile covers 50 km, so just two profiles suffice for the Jakobshavn gorge.
The internal dimension of profile jpegs is 931 x 734 pixels (in the example examined). The width is variable but for a 49.91 km transect, the 1 x 734 vertical slices are spaced at 53.6 meters which represents a resolution of 75 points to characterize bedrock topography for a 4 km wide gorge. A buried nunatak or pothole of smaller dimension might still be recognized by combining data from different years as the bedrock doesn't change year to year and the flight lines would be slightly different.
The depth is not measured directly but rather return time of a radar pulse (microseconds of propagation delay). Under the assumption of 3.15 dielectric for ice -- not applicable to snow, firn, englacial pockets, or wet temperate ice -- the depth is then calculated using the speed of electromagnetic radiation in a medium of refractive index sq rt (3.15), or 168,913,914 m/s instead of the usual 299,792,458 m/s.
The depth scale range has to vary from scene to scene to accommodate top elevation and bedrock depth. In the example I looked at, 2000 vertical meters was represented by 369 pixels for a resolution of 5.4 meters. This scale seems to be consistent, only modified by offsets.
Cresis also offers kml (Google Earth) files for the track segments. I found these convenient for precisely co-registering track profiles via their lat,lon coordinates to Landsat images. Simply mouse along the track to find the exact lat,lon of a radar reflection column. This is otherwise problematic because the icestream channel curves quite a bit. I found a way to lay down a precise grid in Gimp and will post that shortly.
For the two along-gorge, west to east flight lines of April 2014 (which may overlap slightly rather than butt up end-to-front), go to
ftp://data.cresis.ku.edu/data/rds/, open 2014_Greenland_P3/images_csarp-combined/ and append:
20140409_02/20140409_02_004_0maps.jpg
20140409_02/20140409_02_004_1echo.jpg
20140409_02/20140409_02_004_2echo_picks.jpg
20140409_02/20140409_02_005_0maps.jpg
20140409_02/20140409_02_005_1echo.jpg
20140409_02/20140409_02_005_2echo_picks.jpg
For the seven cross-gorge transects (some require a north/south pair) in west to east order of April 2014, append:
20140409_01/20140409_01_022_0maps.jpg N
20140409_01/20140409_01_021_0maps.jpg S
20140409_01/20140409_01_019_0maps.jpg
20140409_01/20140409_01_017_0maps.jpg N
20140409_01/20140409_01_016_0maps.jpg S
20140409_01/20140409_01_014_0maps.jpg
20140409_01/20140409_01_012_0maps.jpg N
20140409_01/20140409_01_011_0maps.jpg S
20140409_01/20140409_01_009_0maps.jpg
20140409_01/20140409_01_006_0maps.jpg
The data go back quite a few years (to 1993 for Jakobshavn) and involve other aircraft and other radars. It is not trivial to commingle older with newer data but there may be some value to it.
I've attached the 2014 flight line that goes under the calving front -- note the intriguing structure approximately at the first sill.