A bit more on geomorphons, varying the 'inner search radius'

I had originally posted maps of DEMs of Cornwall and the Aberystwyth area of mid-Wales processed using Tomasz Stepinski et al.'s geomorphon method as descibed in the papers of: Stepinski+Jasiewicz 2011, Jasiewicz + Stepinski 2013.

Since then I have become aware of the 1 arcsecond SRTM data being available, so I have recalculated the geomorphons and here show the map of the same area of mid-Wales that the landcover assignment from my MSc was concerned with.

Using the search distance L=1500m, as did T. Stepinski in his geomorphon map of Poland (user guide), I varied the 'inner search radius' which determines the minimum distance the software looks, using 0, 2 and 4 cells. 1 cell is ~ 24m in the SRTM 1 arcsec data after conversion to OS grid reference coordinates.

Inner search radius 0m

In some of the flatter areas results are difficult to interpret. full-resolution image.

Inner search radius 2 cells (~50m)

Looking in Borth Bog, more of the area is shown as 'Flat' full-resolution image.

Inner search radius 4 cells (~100m)

As the inner search radius is expanded to 4 cells, most of the possibly spurious peaks in the flat area of Borth Bog disappear. full-resolution image.

I expect it would be interesting to combine it with the segmentation and classification with RSGISlib and to examine correlations between the landscape position of a location, and the land-cover.

Revisiting geomorphons - with 1 arcsecond SRTM data

I had originally posted maps of DEMs of Cornwall and the Aberystwyth area of mid-Wales processed using Tomasz Stepinski et al.'s geomorphon method.

Since then I have become aware of the 1 arcsecond SRTM data being available, so I have recalculated the geomorphons and present some maps of Cornwall here.

I've also become more aware of the various layer blending methods in QGIS which has allowed me to blend it with a hill-shaded topography layer.

In these maps, the search distance L = 1500m, flatness threshold 1 degree.

West Cornwall. Full resolution image.

SE Cornwall, Tamar valley and Dartmoor. Full-resolution image.

North Cornwall. Full-resolution image.

Minimum search radius

It is possible to specify an 'inner search radius' that ensures a minimum search radius for the geomorphons. In the maps below, using a minimum search radius of 2 cells (about 50m for this data), this makes the map look a little less noisy:

West Cornwall. Full resolution image.

SE Cornwall, Tamar valley and Dartmoor. Full-resolution image.

North Cornwall. Full-resolution image.

A few more geomorphon maps

One of the problems with the maps in my last post is that the vertical scale of the landforms is completely lost, it is impossible to tell whether a slope is shallow or steep.

A possible way to bring this information back is to grey out the shallowest slopes which I have done progressively for slopes less than 5 degrees in the maps below:

Geomorphons - processing British Isles SRTM DEM with GRASS7 r.geomorphon extension

Some time ago I mentioned Geomorphons, a machine vision based approach to topographic analysis promoted by Stepinski and Jasiewicz.

If you don't know what a geomorphon is, here's a link to the Space Informatics Lab at University of Cinncinati, see also the papers: Stepinski+Jasiewicz 2011, Jasiewicz + Stepinski 2013 

At the time I first came across this I had some trouble getting GRASS7 to install and get the extension working in Linux so I tried running GRASS7 in windows inside a VM but it didn't really work well.


Now however I have a working GRASS7 installation on Linux and here's a few demonstration maps, which show the broad categories of landscape position. Geomorphons use a line of sight approach best explained by the links and papers above.

I have used a search distance of L = 1000m in these maps, using Shuttle Radar Topography Mission data for the British Isles. I've set a threshold for flatness as 0.2 degrees, which is probably too small because in flat areas I'm not confident of the reality of these interpretations.

Geomorphons have a different kind of scale-dependence than the usual differential geometry based approaches used, though it is not independent of scale, it is more so than a simple slope analysis.

Co-ordinates are OSGB all-numeric throughout.

West Cornwall

SE Cornwall

N Cornwall

 

Aberystwyth area in mid-Wales

Mid-wales the same as above but with flatness threshold of 1 degree

Changing the threshold for flatness to 1.0 degrees, flat areas such as Borth Bog are shown as such on the map.

A closeup with 5m data:

Nevertheless I keep search distance L=1000m.

Analysis of Crater Greg using LandSerf

I had some trouble with geomorphons, because I couldn't get the GRASS 7 extension to compile in Linux for unclear reasons, so I was limited to using the Windows version inside a virtual machine, and it decided to crash when presented with Mars data, although it was OK with Earth DTMs...

I have however been using LandSerf to generate curvature layers, and a feature classification, so here we are, for the well known (at least to Martian glaciologists) crater "Greg":

I hope you're not too confused by the use of transparency...

Units are in metres.

Some of the glaciers form a channel as they are constrained by topography, but not all channels are glaciers.....

Using longitudinal curvature calculated in LandSerf, showing negative curvature where a slope starts off steep and shallows as it goes down, which the Martian glaciers in Greg and elsewhere generally do, (Bryn Hubbard et al.):

Using RSGISLib to segment a layerstack consisting of nadir image, elevation, slope, aspect, and curvature layers:

Selecting all slopes above 15 degrees

Geomorphons of mid-Wales

Following the machine-vision based method of topographic analysis, "Geomorphons" by Stepinski and Jasiewicz, here's an example of the method applied to mid-Wales:

The authors argue that the method can make a useful analysis of topography at differing scales, in a better way than typical differential-geometry based methods do.

A close-up of the Aberystwyth area:

The method is available either as an extension to version 7 of the open-source GRASS GIS software, or, for small datasets, as an online interface.


After a number of hours, a wider search area can be made to match the scale of Stepinski and Jasiewicz's map of Poland in their paper , in which the authors used a DEM at 30m resolution at a search radius of 50 cells.

Hopefully I'll have something useful from this method applied to Mars in time to go in my dissertation....

Website: Space Informatics Lab - University of Cincinnati