2

I have a DEM raster file downloaded from the USGS site. The file can be downloaded here.

I am trying to create a slope raster in percent with the DEM tool in QGIS. The GDAL command is the following:

gdaldem slope /Users/Downloads/n41w107/imgn41w107_13.img /Users/Downloads/slope.tif -p -s 1.0 -b 1 -of GTiff

The tool runs fine, but the created file slope.tif has strange data. The statistics for the band are the following:

STATISTICS_MAXIMUM=29834248
STATISTICS_MEAN=1845376.751065
STATISTICS_MINIMUM=0
STATISTICS_STDDEV=1658350.0283881

So they are obviously way to big for percent. What am I doing wrong?

Update: I calculated the Slope in degrees and got the following statistics:

STATISTICS_MAXIMUM=89.999809265137
STATISTICS_MEAN=89.583139343302
STATISTICS_MINIMUM=0
STATISTICS_STDDEV=5.9592219763749

A mean of 89 degrees seems really high to me. Is that reasonable?

Improve this question
2
  • 3
    I am not familiar with USGS-data and capabilities of slope calculation with QGIS. But I would check if xy-units are equal to z-units. If xy-units (e.g degree) are not equal to z-units (meter, feet) the calculation of slope will be wrong.
    – Jens
    Apr 3, 2014 at 14:07
  • @jens thanks. I ended up here and that was the answer for me.
    – BenjaminGolder
    Apr 3, 2014 at 17:38

1 Answer 1

Reset to default
1

Be wary of stating, out the gate, that something is producing "wrong" results. The phenomena of slope percent approaching infinity as slope degrees approach 90 is well known.

You could just truncate slopes > 100 == 100.

ESRI actually provides a very nice description on how slope is calculated. You could try the calculation in degrees to make sure that the model is producing values that are sensible.

Improve this answer
1
  • I calculated the Slope in degrees and updated my question.
    – ustroetz
    Apr 3, 2014 at 13:55

Not the answer you're looking for? Browse other questions tagged or ask your own question.