Digital Terrain Modeling and Mapping Digital Elevation Model

Generate 3D Digital Terrain Models and Digital Terrain Maps using free digital elevation model (DEM) data from the USGS, NIMA, ASTER, Landsat and other sources. Learn to render them using POV-Ray, MicroDem, Terragen, 3DEM, Global Mapper, Arc Explorer, Multispec and other free or low cost applications. Convert data using GEOTIFF, BANDPASS, DEM2TGA7, DEM2POV2, MDEM2DEM, SDTS2DEM, BLACKART and other applications offered for free download.

News and Information

July 17, 2003 - BLACKART v1.04.02 This is the windows version of BLACKART replacing the very short-lived v1.03. It is a beta version that does not include many necessary features such as exact output file alignment and graphics capability. However, I think it is worth releasing as it is a lot better than the previous console application, having a proper GUI and some input error checking. This version fixed a couple of bugs discovered on deployment, like turning off the RTL and RT Package DLL switches and fixing the window sizing attributes. This version also has a compiler optimization that allows the code to run much faster than the console version. Input files of size 1201 by 1201 elevation nodes are now no problem. See the related Blackart Interpolation article.

July 9, 2003 - BLACKART v1.02 I corrected an error in the initial estimate vector computed by BLACKART. As a result of the correction the new version converges much more quickly than the previous version. See the related Blackart Interpolation article on this site. (Coming soon: BLACKART GUI version).

July 6, 2003 - BLACKART v1.01 I recently revised BLACKART to make the coefficient matrix data structure and associated vector data structures more compact. As a result, the new version will run input files of at least 900 by 900 elevation nodes without excessive disk thrashing (assuming your computer has at least 256MB RAM). This makes processing times faster than the previous version, although once the file size gets to 1201 by 1201 nodes processing is still slow. The computational kernel is at least as fast as the MATLAB version that I used during my research. However, I still have a couple of ideas to speed it up a little more.

June 23, 2003 - BLACKART Franklin Approximation Conjugate Gradient Iterative Contour Line Interpolator Those of you who have been breathlessly following my Master's Degree project at RPI Hartford have probably been wondering when I would convert all the developmental and research code into a useful application. Well, here it is, another piece of research-grade code written in C++, currently featuring a sub-optimal implementation of the Saunders-Paige conjugate gradient iterative least squares solver. (To be more precise, solving large input files (greater than around 160,000 nodes) at present is slow because of the representation of the sparse matrix as a simple index file. A more efficient representation would allow faster computations for large data sets by reducing excessive memory paging. I should have an improved version soon.)

BLACKART in its present state takes a space-separated ASCII elevation contour file as input and outputs either a USGS ASCII DEM or ASCII xyz interpolated surface file. The program combines the sparse matrix index file module, fast Laplacian central difference solver, Saunders-Paige CG solver and USGS ASCII DEM writer into a single console application.

So what is the point? BLACKART represents the second (and arguably more difficult) half of a general purpose application to extract high quality DEMs from topographic maps. The first half contains the raster-to-vector and image processing algorithms necessary to prepare the contour elevation grid file for input to BLACKART. Once I have the second half fully developed, I plan to start working on the first half and eventually put together a useful application for extracting DEMs from topos. In the mean time you can download the prototype code and experiment with generating DEMs from the sample contour files included or ones that you prepare yourself. See the complete Blackart Interpolation article for details.

Baghdad Area Satellite Images Examine high resolution satellite images of Baghdad at the Digital Globe website. The site offers a couple of full-resolution 0.9m and 0.6m images of the city including a color composite image taken in September, 2002 and a grayscale taken March 27, 2003 along with several subset images showing recent bomb damage. Additional photos can be seen at the War Images of Baghdad article. For the most comprehensive source of military information on the internet, including maps, satellite images, order of battle, and a great deal of additional political, geographical and social information about Iraq, visit the invaluable Global Security site.

USGS ASCII DEM Specification Periodically, people ask me for a link to the USGS ASCII DEM File Specification. The last time I tried I could not find one, so here is a copy in case anyone needs it.

GEOTIFF4 Correction: I corrected a serious problem with GEOTIFF4 that was reported by a user. When saving a subset image, the program appended a newer version of an output USGS ASCII file to an older version if the file already existed instead of overwriting the existing file as it should have. I corrected the error and reposted the program. I have no idea why I set the append flag when opening the output stream in the first place and why I did not catch this embarrassing problem during development. It's probably a result of losing my mind many years ago. I am very sorry for the confusion this must have caused users of the program.

SRTM Rocket February marked the three year anniversary of STS-99, the shuttle mission that mapped 80% of the earth's surface using an interferometric radar instrument. After three years of foot-dragging and unkept promises, NIMA (NASA's "partner") has yet to authorize release of non-US SRTM DEM datasets. In case you have not been following this sad story, the plan is to degrade non-US data from a marginally adequate 30m (instrument resolution) to a practically useless 90m resolution prior to release to the public in 2004 (we are now told ). All that our multi-billion dollar SRTM investment has returned so far, three years post-launch, is an inferior duplication of the USGS NAD and a flashy SRTM website. The "showcase" images offered by NASA are more insulting than enlightening. Earth to NASA: just give us the data!

Perhaps NIMA is worried that the bad guys will use the data to threaten us somehow. If so, they should just do what ASTER does: offer the data (essentially) unreferenced. Then at least we could ground-truth individual granules with user-supplied GCPs. NASA/NIMA have already released 30m data for the United States, so what is the problem with releasing a similar quality international product? Anyone wishing to do us harm could order much higher quality DEMs from Space Imaging or Digital Globe anyway. But this is obviously not the point. It is about turf and control, and NIMA does not want to give up either.

NASA/NIMA have also announced distribution of the data through USGS EROS. This may mean that the data will ultimately be distributed on a fee basis, as USGS EROS does not give away much for free (as a visit to their website will confirm). After years of waiting and untold tax dollars spent, will we enjoy a $60 fee for each granule like we do for ASTER L1A/B? While the growing NASA trend of pay-per-view may please commercial interests by limiting competition, it raises serious questions at a time when NASA is struggling to discover its mission. The American public truly believes in NASA despite its many faults and funds the agency with it's tax money to advance the frontiers of science. NASA betrays that trust when it fronts a second-rank remote sensing business.

NASA could easily acquire and distribute sub 1 meter DEM and image data if it wanted to, but it has conceded that scientific high ground to the commercial remote sensing industry. I have a new business model to suggest: let NASA get the data, let the USGS freely distribute the data, and let the commercial remote sensing industry apply the data to the betterment of human kind. Commercial remote sensing has never been viable collecting data anyway (at least not without government contracts as a prop). Acquiring data is far too expensive as a strictly commercial proposition. That role is precisely the one that NASA should be fulfilling. As for NIMA, maybe they should reconsider their role as well.

You may disagree with my analysis, but one fact is not arguable. If you weigh the expenditure of funds versus the paltry benefit delivered so far, SRTM is one of the worst investments in the history of NASA. The tragedy is that it should have been one of the best.

Anyone wishing to express their dismay at this situation might consider contacting Hon.Sherwood Boehlert (R NY) Chairman of the House Committee on Science to express their views.

MDEM2BSQ Update At the request of a user I recompiled MDEM2BSQ using VC++ in order to allow larger input files to be processed. The new program is called BIGBSQ and it should process input files of "unlimited" (>65MB) size.

EarthSat Link Update Thanks to Albert Groebli for sending in a dead link alert for the NASA ESAD MrSid site. The link to this great site is now corrected. (Note: some firewalls will not like this link.)

RPI Master's Project It has been my good fortune to be able to work with Dr. William Randolph Franklin at RPI Troy in Troy, NY USA on an interesting problem in computational cartography. I am doing this work as partial fulfillment for the requirements of my MSCS degree at RPI Hartford. My project is entitled: "A Better Interpolation of Large Contour Line Data Sets using an Over Determined Laplacian PDE". This project is related to the more general problem of producing DEMs from contour line data. Dr. Franklin is a well-known applied mathematician with expertise in computational cartography. Anyone interested in following my progress can take a look at my school web page at RPI Hartford

359 unread Mail messages in your Inbox... As those of you who have attempted to contact me by email over the last few months have discovered, I am falling way behind in answering my mail,due to my school and work schedule. I hope to resume correspondence again at the end of this semester, which should be my last for now. In the meantime as I am merely scanning my email, the following would help:

1) Describe your topic in the subject line of your email if possible.
2) Have an interesting subject ( threats, bug reports, error reports and dead link reports are always interesting, as are reports of new data sources, new free software and new techniques).
3) Have you checked the FAQ?
4) Try not to offer funds from a Nigerian bank account, or Nigerian diamonds, or a chest full of Nigerian oil money.

University of Maryland GLCF! The University of Maryland's Global Land Cover Facility represents the finest source of free Landsat TM, ETM and MSS imagery on the planet and one of the most important sites on the internet. Read about this exciting data source and their recently upgraded search engine and GIS database in the new article entitled University of Maryland GLCF.

Free Aster L1B! LP-DAAC has recently opened its promised ASTER L1B Data Pool ftp free download site for ASTER and MODIS data. Although coverage is limited to the United States (and not even the entire United States) this is a welcome gesture by LP-DAAC, the USGS and NASA. Let's hope they find a way to expand this offering.

High Resolution Image from Landsat 15m Band! Learn how to combine low resolution Landsat color bands with the high resolution panchromatic band to make a high resolution HSL true color composite. See the article entitled Landsat HSL Composite from Panchromatic Band.

Free 3DEM Everyone's favorite GIS application is now freeware! This excellent program, featured in several articles on this page is available for free download at Simtel.net.

Free NASA Imagery Exchange Data! Learn where to get free NASA Imagery Exchange Data in the new article only at terrainmap.com.

Free EarthKam Data! Learn where to get free EarthKam Space Station Imagery in the new article only at terrainmap.com.

Free EarthSat Data! Learn where to get free EarthSat imagery in the new article Free EarthSat Satellite Imagery.

Regrettable Trend Continues Gisdatadepot.com has recently started charging for maps that they formerly offered for free. Gisdatadepot.com now charges for DRG (topo maps) and DOQQ (aerial photo) data. The only alternative for those desiring free topos and DOQQs seems to be state GIS websites that offer the same data for which the commercial sites are now charging. Unfortunately, only about half the states have such websites. You can find links to the free data at sites such as University of Oregon Library System State Directory, www.nb.net and www.topowest.com. Another excellent GIS resource is the state University libraries. These often offer a wealth of GIS data. For example, with a little digging I found a complete collection of Washington state DOQQs for free download at the Washington State Geospatial Data Archive. I also found a complete set of Texas DEMs, DRGs and DOQQs at TNRIS. I in turn found these links at the very useful University of Arkansas CAST page. Another possible alternative for frequent users of topo maps is TopoUSA. This program is described in the section entitled DEM Starter Kit on this website, which includes a sample topo. The included topos are not as detailed as the 24K USGS versions but you get the whole United States for under $100 USD.

Note: I am in the process of updating all the articles on this page to redirect referenced links toward the appropriate free data sources wherever possible. Most of this work is complete. Refer to the News section of this page for the latest information.

Free Landsat 7 Data! Learn where to get free Landsat 7 imagery and how to process it in the new article Free Landsat 7 Satellite Imagery.

Regrettable Trend: Sadly, effective August 12, 2002 LP-DAAC has started charging USD $60 for each formerly free ASTER L1A/L1B granule. (Thankfully ASTER DEMs are still free.) ASTER imagery was formerly the highest quality free satellite data available anywhere. United States citizens who feel as I do that after paying with our tax dollars for the satellite, launching the satellite, maintaining the groundstation, processing the data and administering the database where the data is archived that perhaps we already own the data may wish to contact LP-DAAC at edc@eos.nasa.gov to express this opinion. Maybe the costs associated with making the data available do justify the price, but as is usual from our government agencies no explanation was considered necessary.

My Apologies: for all the unanswered emails that I have accumulated this summer. I am in the final year of a demanding MSCS program at RPI Hartford and have been overwhelmed by the demands of school and my job. I formerly spent at least an hour a day answering emails. Eventually I needed that time to devote to studies and as a result I have a considerable backlog. I plan to improve my FAQ page to try to answer more questions using that forum. I apologize again but it is important to me to finish my degree work so that I can devote more time to writing better GIS applications.

GEOTIFF4 Version 1.49!: GEOTIFF4 can now read ASTER DEM I16 and FP32 HDF format as well as ASTER geotiff format. The program has graphics capability, subsetting, the ability to drop row and column markers, adjust elevations and more. The program outputs DEM files in USGS ASCII, flat binary, and terragen format. Download it for free only at terrainmap.com!

ASTER Relative DEM Vertical Errors: I have received several emails recently reporting apparent vertical errors in ASTER relative DEM data as compared to known standards. Read more on this including information from EOS-EDC in the FAQ ASTER section.

ArcView Upgrades I received a couple of emails recently indicating that the output from my USGS ASCII writer cannot be read by ArcView and ArcInfo software. I determined that the strict specification requires that the ASCII file be divided into 1024 byte blocks delimited by 0x0d0a. I rewrote the writers for SDTS2DEM, MDEM2DEM and GEOTIFF4 to incorporate this requirement. This did some good because GlobalMapper could not read the output in the past and now it can. I am hoping that ArcView can also read the output, and am in the process of finding someone with ArcView to test them for me. I am posting these versions as betas for now. Those having trouble with the outputs can get at the files by clicking GEOTIFF4, MDEM2DEM5 and SDTSDEM7.

Upgrade: Resequenced the creation and destruction of internal data structures within MDEM2BSQ so that memory usage is decreased by 33%. This should allow larger files to be processed.

Bug Fix: Modified MDEM2DEM so that it handles 10m SDTS source DEMs correctly.

Bug Fix: I was notified by a user that BANDPASS crashes when fed a very large >50mB input .tif file due to a page fault error. This is important because some DOQs needing redeye correction can be more than 100MB in size. The problem was in the Borland compiler that I normally use. I recompiled the program using Microsoft Visual C++, and this has apparently fixed the problem. I will post it as a separate program until I do more testing, at which time it will probably replace BANDPASS. I call the new program BIGBANDPASS. Anyone using BANDPASS for very large files should use the new program instead.

Bug Fix Added a line of code to BANDPASS to keep it from crashing horribly if an illegal file is input. Sorry.

Bug Fix Corrected SDTSDEM2 so that it reports the UTM zone and the four corner coordinates to the console correctly.

Converter: Convert MicroDEM DEM files to .BSQ flat binary format. This format can be used in the creation of MicroSoft Flight Simulator terrains from USGS DEM data. The new program is called MDEM2BSQ. FS terrain users: let me know if you find bugs or have suggestions. (Note: I had my doubts about the rotation of the output file but I am told that the way it is presented is correct.)

Converter SDTSTER4 converts 10m and 30m 7.5' SDTS DEM files to Terragen .ter format. This program can convert 10m BFP(32)(decimal meter) format as well as BI(16) just like SDTSDEM2 below. This program supercedes SDTSTER3

Converter: SDTSDEM2 converts 10m and 30m 7.5' SDTS DEM files to USGS native DEM format for input to legacy applications unable to read the modified SDTS profiles such as 3DEM70 . This program can convert 10m BFP(32)(decimal meter) format as well as BI(16). It can handle both new and old SDTS formats. This utility has been tested successfully against Microdem, DEMWorks, 3DEM, 3DEM70, and DEM2TGA6. This utility is the testbed for an upgraded SDTS reader that I will incorporate into all the SDTS utilities on this page. I will leave SDTS2DEM up for a while in case there are bugs in SDTSDEM2 that I do not know about. Eventually, SDTS2DEM will be superceded by SDTSDEM2. Thanks to George C. Miller of the USGS for supplying information that made the upgrade possible.

Revised Converters: DEM2TGA7 and DEM2POV2. can now handle CDED data format. I discovered a bug in the DEM reader parts of these programs recently while working with Canadian CDED data files. The file reader was not handling certain horizontal and vertical datum values properly causing the program to hang. I fixed this problem and now DEM2TGA7 and DEM2POV2. can now read CDED as well as USGS native DEM successfully. My apologies to anyone having trouble with these converters for this reason.

Converter: MDEM2TER converts from MicroDem native DEM format to Terragen .ter format.� With this utility, Terragen can now ingest any DEM format that MicroDem can, and that's a lot.� It is now possible to use the powerful file merge capabilities of MicroDem and then input the merged files to Terragen for rendering.� See the tutorial below.�

Converter:� MDEM2DEM (Revision 1) converts MicroDem DEM format to USGS native DEM format.� This allows programs that read only USGS DEMs to take advantage of MicroDem utilities.� For example, you can import two 7.5' SDTS files into MicroDem, merge them, then save the merged image as a MicroDem DEM.� Run� MDEM2DEM on the MicroDem output file and create a USGS DEM file suitable for input to another application that reads only USGS DEMs.

Converter: For those of you who need to convert from SDTS to 24-bit TGA format there is now SDTS2TV. It converts 30m 7.5' SDTS DEM files to 24-bit TGA format such that the elevation values range from zero to 16,000,000. Data is written to three bytes in the order: lowbyte, midbyte, highbyte.

Background

I started this page as a result of a project that I worked on during the Fall semester in 1999 at Murray State University in Murray, KY. We were given an assignment to write a program to generate a script file for the ray tracing program POV-Ray. This exercise introduced me to the world of 3D terrain modeling using POV-Ray's height field capabilities. My work with POV-Ray allowed me to combine my interests in cartography, programming and 3D Graphics.

In order to create my terrain models, I needed data. This caused me to write my first conversion program to convert USGS DEM data to .tga graphics format for input to POV-Ray. I soon became fascinated by the bewildering and frankly, shameful proliferation of file formats that mapping data is presented in, and resolved to gain a greater understanding of some of these formats. (Unfortunately, the profusion of cumbersome file formats is becoming worse, not better.) I created this page because I felt that others might benefit by my efforts to "mine" this data and convert it to forms useful for the creation of digital terrain models.

The sections presented in this page reflect my admittedly limited progression through the world of digital cartography and are presented in chronological order. Since the GIS world in general is changing rapidly, some of the initial sections are dated and do not represent best practice, but I leave them up for their hopefully educational benefit.

You will notice that there is a pronounced emphasis on free data sources, general technique (as opposed to specific applications), and the demystification of this field. The discipline has rightfully been dominated by the scientific and research community, which however is not an open society. I believe that the layman can benefit greatly by the work of the elite, and that GIS will impact our world the way desktop office applications, networks and CAD systems did a decade ago.

My interest has recently expanded to include free sources of high resolution satellite imagery in addition to high resolution DEMs. Both are required to make high quality digital terrain model ovelays but multi spectral satellite images offer a particuly rich source of data for other purposes as well.

Use of any software application on this page is merely one of convenience. I generally try to use whatever is available for free demo to illustrate techniques, and the use of the software does not constitute a recommendation. However, if an application appears in several sections, you can assume it is excellent because I don't have a lot of time to waste in putting up this page.

Lastly, you will notice that I make mistakes on this page from time to time. I am an amateur and a student (in the broad definition) of cartography. I appreciate the many emails (many from distinguished professionals) that I have received pointing out errors because they have greatly improved the content of the page. Also most of the latter sections were started as a result of email suggestions.

Next Page>>

Contacts

If you have any questions or comments about the contents, wish to report bugs, or to report dead links please contact me by Email.

Terrainmap.com
Digital Terrain Modeling Techniques
jchilds@terrainmap.com

Top

Site created June 04, 2000.
This page was last updated July 17, 2003.
Page content and all images � Copyright 2000, 2001, 2002, 2003 John Childs all rights reserved.
Images may be used for any purpose with acknowledgement to www.terrainmap.com.

Notice: No warranty for the author's programs is given or implied; use them at your own risk. The programs offered on this page, while believed accurate, have not been subjected to extensive quality control testing. They are not to be used in mission-critical GIS applications without the express written consent of the author.