|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | GISList: Re: SRTM Data and ellipsoid-geoid problem |
| Date: |
11/09/2002 09:21:36 PM |
| From: |
Peter Shary |
|
|
Aditya,
I worked with SRTM DEMs in my scientific research on comparative
land form analysis.
I do not know your experience with SRTM, so I list below well-known
things together with those originated from my research.
There are two kinds of SRTM DEMs:
1) "PI_Processor" elevations that are relative to the WGS84 ellipsoid.
2) "GDPS" elevations that are relative to the WGS84 geoid.
For descriptions (and for data), see files at this FTP thread:
ftp://edcsgs9.cr.usgs.gov/pub/data/srtm/
For links to SRTM and other DEMs see this URL:
http://members.fortunecity.com/eco4/giseco/id8.html
Ellipsoid essentially differs from geoid, and this might result in
essential mismatch in elevation values. Geoid is the best choice, but
my colleague told me that such elevations are normally secret in many
countries including USA and Russia (he commented that the difference
between ellipsoid and geoid reaches 70 meters even in non-mountainous
territories). I do not know USA National Elevation Dataset
specifications (is it relative to ellipsoid or to geoid?), but there is
a little sense to compare ellipsoid-based and geoid-based DEMs. Note
this remark by Thomas Harris at "Gislist" listserver: "I received USGS
DRG's that are in UTM NAD27. Overlaying our Hawaii roads coverage
reveals a uniform offset in the DRG data of over 100 meters. We have
confidence in our road data because of field validation with GPS."...
Comments were like by Mike Flannigan for "Gislist": "It sounds very
familiar"...
Does someboby knows better the problem of geoid-ellipsoid differences
in national standards? I will sum.
SRTM DEMs have several disadvantages, at least the following ones:
(1) Elevations are rounded to whole number of meters. Because of this,
you will encounter problems in some applications, such as hydrological
ones. Essentially, such topography "works" good (for hydrological
modelling, oil spatial re-distribution after pipeline damage, etc.)
only in mountainous terrains.
(2) "PI_Processor" elevations contain non-existing peaks and sinks
(sinks up to 100 m in depth) that are due to low radar irradiation
reflectance from water, and to elevation mismatch at ground control
points.
(3) "PI_Processor" elevations contain "holes" (areas with unknown
elevations) related mostly to surface water and to mountain peaks.
Water surfaces and wet surrounding, such as riverbeds, are too "noisy"
(amplitude of tens of meters).
To my knowledge, "GDPS" elevations became available only couple of
month ago at FTP link above, and this data quality is not quite clear.
Washington officials promise to make 90-m SRTM DEMs available worldwide
by 2004 (for 80% of drylands), see this website:
http://usinfo.state.gov/cgi-bin/washfile/display.pl?p=/products/washfile/latest&f=02090401.clt&t=/products/washfile/newsitem.shtml
Hope this helps.
Peter Shary
**********************
Scientific researcher and GIS developer
Institute of physical, chemical, and biological problems of soil science
of the Russian Academy of Sciences
142290 Poushchino Moscow region Russia
Tel.: +7 0967 733604
Fax: +7 0967 790595
Research URL: http://members.fortunecity.com/eco4/research_shary/
**********************
>From: Aditya Agrawal <Aditya.Agrawal@NOAA.GOV>
>Reply-To: Applied GIS and RS <APPLIED-GIS-RS@LISTES.ULAVAL.CA>
>To: APPLIED-GIS-RS@LISTES.ULAVAL.CA
>Subject: [APPLIED-GIS-RS] SRTM Data
>Date: Fri, 1 Nov 2002 11:18:07 -0800
>
>Hello. I was wondering if anyone has used the new SRTM elevation data
>that is available through the USGS (http://srtm.usgs.gov). Has anyone
>come across a comparison of the level of accuracy between the SRTM and
>the NED (National Elevation Dataset, also through USGS)? I'm trying to
>get some indication of how much these two datasets differ. If you have
>any suggestions or
>insight about these data, please let me know. Thank you for your time.
>I will sum the replies.
>
>Cheers,
>Aditya
>
>--
>Aditya Agrawal
>Research Associate/GIS
>NOAA Fisheries
>Southwest Science Center
>Santa Cruz Laboratory
>110 Shaffer Rd.
>Santa Cruz, CA 95060
>phone 831.420.3916
>
>-----
>Applied-GIS-RS home page : http://www.matox.com/agisrs
>
>Archives :
>http://listes.ulaval.ca/listserv/archives/applied-gis-rs.html
>
>To unsubscribe :
>http://listes.ulaval.ca/cgi-bin/wa?SUBED1=applied-gis-rs&A:=1
>( or http://www.matox.com/agisrs and click on 'Leave list' )
_________________________________________________________________
Internet access plans that fit your lifestyle -- join MSN.
http://resourcecenter.msn.com/access/plans/default.asp
To unsubscribe, write to gislist-unsubscribe@geocomm.com
__
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
