Creating appropriate Slope from SRTM data

1178
1
04-27-2011 07:29 PM
FahmiHaris
New Contributor
Hello, I am now generating a slope map on scale of 1:25.000.

The method that used:
1. ArcGIS 9.3
2. Source of the data from SRTM 90
3. SRTM data in the form of ASCII converted into contour intervals of 12.5 meters (as crosc check material).
2. SRTM data above also be converted directly into the slope, using per cent (%) as unit.
3. The Slope data then being reclassified in accordance to the office-need into 8 classes: 2,8,15,20,25,30,40,100.

Problem:
When the class of slope being cross-checked with the contour, it's not fullfilled with distance between contours as it categorised. Let's say  the 2-8% slope class, should the distance between the contour (d) is between 156-625 meters (real distance) but when it measured with an ArcGIS ruler, the distances between contour line shows only 100 meters that should one class below.

Question:
1. How to overcome this problem?
2. I use another method as plan B, that convert the contour of the SRTM to TIN and then I change it to slope. But the result is still not accurate. Is there a process that is more accurate?


As an emergency solution at this time, I use manual methode by delineating contours manually on paper. This manual process takes much longer time to be done. So your help would be so much lovely matter...

Thans in  advance.
-
FCM Djoko Widodo
BPN Regional Office of Central Java Province
INDONESIA
E-mail : fahmicmdwidodo@gmail.com
1 Reply
JakubSisak
Occasional Contributor III
Is this the "3 arc second SRTM" data? Do you mean 90 because the approximate resolution is 90m?

Are you trying to verify whether Spatial Analyst produces the correct result?

I would say the problem lies in comparing the slope raster which was generated from the DEM itself and the contour lines which are too dense for this type of DEM. I'd say there is a certain degree of interpretation that goes on in the contouring algorithm to calculate the contours from the DEM this course.

Also you will probably have better luck posting this in the Spatial Analyst forum.




0 Kudos