Sign in to follow this  
Guest vlada stoje

Version 2 of SRTM data available

Recommended Posts

Hi all,just stumbled across a news release at http://seamless.usgs.gov/announcements.aspThe "cleaned" version of the 1-arcesc (US only) and 3-arcsec SRTM data has been uploaded. Note, however, that "cleaned" does not mean free of voids. Here's the relevant excerpt:"The editing, also referred to as finishing, consisted of delineating and flattening water bodies, better defining coastlines, removing "spikes" and "wells", and filling small voids. This "finished" set is publicly available at two postings: 1 arc-second for the United States and its territories and possessions, and 3 arc-seconds for regions between 60 degrees N and 56 degrees S latitude.The method NGA uses to produce 3 arc-second data from the 1 arc-second set is "subsampling", namely selecting the center value from the set of nine centered on a particular posting location. This is the method that has been used to produce the 3 arc-second edited data that are available through the Seamless Data Distribution System. The "finished" data after editing still has void areas. The editing, also referred to as finishing, consisted of delineating and flattening water bodies, better defining coastlines, removing "spikes" and "wells", and filling small voids. Other elevation data, such as GTOPO30, would need to be used to fill the remaining void areas."Cheers, Holger

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

Thanks Holger, I missed the announcement trying to access the download interface the other day having been locked out for months.My problem was the XP SP2 issue and I needed to add the site to 'Trusted Sites' in IE Security settings to get it to load, in case others find the same problem.Now to see if there are any improvements.Andy

Share this post


Link to post
Share on other sites

Hi Holger,thanks for the link to the edited SRTM! I tried download data for Slovakia and for Czech Republic, in the BIL version, without any problem. Notice, when requesting square 1x1 deg, we got matrix with 1200 x 1200 pixels (not 1201 x 1201 as by research SRTM). Moreover, the coordinates of the top left pixel from the blw meta file, for example 14.0004166 and 50.9995833, seem like the symmetry positioning of the points in the square.I tried calculate the difference between edited and research SRTM in the Surfer, and I think that there is small mistake in the created blw meta file. The correct values, bringing minimum residuals, are as expected one half step (1.5 arc sec from the 3 arc sec) bigger, in this example 14.0008333 and 51.0000000:http://www.volny.cz/stoje/scen/srtm_edited_01.gifAnyway the whole project is great, big thanks to the authors and members of the team, including astronauts from Endeavour!CheersVlada

Share this post


Link to post
Share on other sites

Hi Vlada,Interesting observation! I had requested a rectangular area (not 1x1 degrees) of the new "finished" data and then overlaid a research-grade 1x1 tile on top of it in Global Mapper. I don't see any offset in the data, they match prefectly.Maybe there's a misunderstanding? Remember that the 1x1 degree tiles are actually slightly larger than 1x1 degrees. Here's the excerpt from the SRTM_Topo.txt at ftp://e0mss21u.ecs.nasa.gov/srtm/Documentation/"SRTM-1 data are sampled at one arc-second of latitude and longitude and each file contains 3601 lines and 3601 samples. The rows at the north and south edges as well as the columns at the east and west edges of each cell overlap and are identical to the edge rows and columns in the adjacent cell."In contrast, I assume that when you request a 1x1 tile at the USGS seamless data site you actually get a 1x1 degree section, hence the 1200x1200 dimension.As for the adjustments of corner coordinates: Global Mapper will give you an option of NW edge of pixel or pixel center and I always pick the latter because of this quote from the same document: "To be more exact, these coordinates refer to the geometric center of the lower left pixel, which in the case of SRTM-1 data will be about 30 meters in extent."Thus, for the 3-arcsec data the actual NW coordinates for the inf file should be (North + 0.00041666) and (West + 0.00041666) and that is exactly what Global Mapper calculates.Cheers, Holger

Share this post


Link to post
Share on other sites

Hi Holger,first I was thinking that we got another unexpected gift, the 3 arc sec sampling shifted with 1.5 arc sec, so after fusion with the research-grade 3 arc sec, the roughly 2 arc sec data could be resulted. But sorry, not this case! The edited and the research-grade data match best way, for example in the case of the square 1x1 degree, if we put the 1200x1200 edited matrix on the top right section of the research-grade 1201x1201 matrix. The left column (meridian) and the bottom row (parallel) from the 1201x1201 research-grade matrix are passed in the 1200x1200 edited matrix, when requesting square 1x1 degree.For now the coordinates in the blw metafile point to the bottom left corner of the upper left pixel in the 1200x1200 edited matrix, but I agree better would be to point the center of this pixel (point), as you described it for Global Mapper. Perhaps this phenomenon isn

Share this post


Link to post
Share on other sites

Hi all,I tried also the Global Mapper to confirm the results from Surfer, and after zooming the described mistake is visible:http://www.volny.cz/stoje/scen/srtm3s_v2v1_shift.gifAlso tried download fresh BIL seamless data (June 8. 2005) and the error is still there, but hope they will consider my e-mail soon (posted this problem also at SRTM forum)The correction is easy, just open the *.blw file in notepad and add 0.00041667 to both coordinates. Hope this could help!Cheers

Share this post


Link to post
Share on other sites

Hello Vlada,I am getting used to your illustrative images! As I read this thread, it seems that if I download again one of the HGT files covering 1x1 degree sampled at 3 arc of a second, I will get a matrix of 1200 x 1200 points instead of 1201 x 1201 points. With the 1201 x 1201, the border of one tile is exactly the same as the contiguous border of the tile that rests at the side. Therefore it looks that the format is changed. The sampling interval is 1 degree / 1200 spacings = 0.0008333 seconds.I assume that the matrix of altitude points changed from this:http://www.ptsim.com/downloads/hgt1201.gifinto this:http://www.ptsim.com/downloads/hgt1200.gifIf that is the case I may need to change some parts of Sbuilder.Luis

Share this post


Link to post
Share on other sites

Hi Luis,I think that the seamless BIL files, in the case for example one square 1x1 degree, have the points as in your first picture for the matrix 1201 x 1201, but the left column and the bottom row are missing, so only 1200x1200 points remain.For now the mistake in positioning of these BIL data is based on the fact, that the coordinates for the upper left point in the matrix 1200x1200 (second from the left in the 1201x1201 matrix) are taken by mistake from the upper left point in your second picture (in the case of the square N50E014 they are 50.99958333 and 14.00041667, but correctly they have to be 51.000000 and 14.0008333).But I hope this all will be fixed soon and the BIL files with the edited SRTM data will have standard properties, without the error in positioning.Cheers, VladaPS: I am crazy about the new features of your SBuilder! Thanks!!!

Share this post


Link to post
Share on other sites

Hi all,seems the thing is not that simple as I offered above. I have good experiences with different terrain data for Slovakia or Czech Republic and I can confirm very precise agreement of the research-grade 3 arc sec SRTM with different detailed sources here. Watching the shift between seamless finished data and research-grade data, I would say almost sure, that the horizontal position is very accurate in the research-grade and that the seamless need the correction. But Holger noted this phenomenon in the same way as Tom Farr at SRTM forum (the "center

Share this post


Link to post
Share on other sites

Hi Vlada,thanks for clarifying the offset issue! I was getting worried that Global Mapper might have problems or that I wasn't "getting" it ;-)It'll be interesting to hear the official feedback on this and/or whether they will upload corrected versions.Cheers, Holger

Share this post


Link to post
Share on other sites

Hi Holger,I have here perhaps the final correction to the above offered corrections for the current seamless 3 arc sec SRTM. After observation of several another samples, it seems that there are missing two data lines causing the positioning problem, one vertical at Greenwich meridianhttp://www.volny.cz/stoje/scen/srtm_3s_sea..._greenwich1.gifand the second horizontal, but not at the equator, but surprisingly at the parallel N01http://www.volny.cz/stoje/scen/srtm_3s_sea...s_paral_n01.gif So the correction for the coordinates depending at the position on the Earth should beULXMAP : = ULXMAP + sign(lon)*0.00041667 (west minus, east plus)ULYMAP : = ULYMAP + sign( lat

Share this post


Link to post
Share on other sites

Hi Vlada.I'm glad you're keeping up with this. It doesn't do anyone any good if the dataset isn't properly located.Dick

Share this post


Link to post
Share on other sites

Hi Vlada.I'm glad they are fixing it. I would have been happier if they would have not released it until it was right! Many thanks to you for making that forum aware of the problem... and sparking the effort to fix it. :)Dick

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this