A multi-functional conceptual tool for mass movement routing



A multi-functional conceptual tool for mass movement routing

r.randomwalk is a flexible and multi-functional conceptual tool for backward- and forward-analyses of mass movement propagation. Mass points are routed from defined release pixels of one to many mass movements through a digital elevation model until a defined break criterion is reached. Lateral spreading is ensured by a constrained random walk approach. Compared to existing tools, the major innovative features of r.randomwalk are:

  • multiple break criteria can be combined to compute an impact indicator score;
  • the uncertainties of break criteria can be considered by performing multiple parallel computations with randomized parameter settings, resulting in an impact indicator index in the range 0-1;
  • the results can be validated against observations and visualized (built-in functions to do so are available);
  • observed landslides can be back-analyzed to derive the density distribution of the observed angles of reach. This distribution can be employed to compute an impact probability for each pixel.

Further, impact indicator scores and probabilities can be combined with release indicator scores or probabilities. Risk indicator scores can be computed if possible objects at risk are given.

r.randomwalk is open source, distributed under the GNU General Public License. It builds on GRASS GIS and the R Project for Statistical Computing. Note that r.randomwalk currently only works on UNIX systems (Ubuntu LTS 12.04 is recommended) with GRASS GIS 7.0 installed from source.

The key concepts of r.randomwalk and various tests for specific areas are laid down in detail in the following publication:

Download links
Mergili, M., Krenn, J., Chu, H.-J. (2015): r.randomwalk v1, a multi-functional conceptual tool for mass movement routing. Geoscientific Model Development 8: 4027-4043. doi:10.5194/gmd-8-4027-2015
Mergili, M., Chu, H.-J. (2015): Integrated statistical modelling of spatial landslide probability. Natural Hazards and Earth System Sciences Discussions 3: 5677-5715. doi:10.5194/nhessd-3-5677-2015

Below you can find the model code, the manual and the data (as far as it may be freely distributed) needed to reproduce the computations performed for the test areas in Mergili et al. (2015).

Be aware that the application of computer models in the field of natural hazards is highly critical. First, all tools, data and manuals were prepared with utmost care and with the purpose to be useful - however, they may still contain errors of various types. Second, even the best models only produce a distorted and generalized view of reality. Their interpretation requires (i) extreme care, (ii) a detailed understanding of the model and (iii) complementary information such as measurements or observations. The unreflected communication of model results may lead to unwanted consequences. Whilst I am highly grateful for critics or suggestions, I clearly refuse any responsibility for any adverse consequences emanating from the use of any of the versions of the tool provided below.

Please consult the manual of r.randomwalk provided along with the code and the relevant publications before starting to use the tool. If you encounter issues not covered by the manual, you are highly welcome to join the forum on r.randomwalk at landslidemodels.org.

Note that the efficient use of r.randomwalk requires some computational skills. If necessary, please acquire the relevant knowledge on Linux, GRASS GIS, and GIS in general before starting to work with r.randomwalk or posting in the forum.

Code and manual

Extract the source code and place the folder r.randomwalk somewhere into your home directory. Do not change the internal structure of the folder or the names of the files. Refer to the manual for instructions how to install r.randomwalk.

Download links
20 March 2017
Additional functionalities added.
18 November 2015
Several bugs fixed and additional functionalities added.
25 August 2015
Bugs with regard to the two-parameter friction model fixed.
Source code
1 August 2015
Initial version of r.randomwalk.
Source code

Test data

Release files and start scripts are available for all test areas in Publication 2. GRASS Locations are provided for those test areas where the data may be freely distributed. Perform the following steps to run r.randomwalk for one of those areas:

  1. Extract the GRASS Location somewhere into your home directory and open GRASS 7.0 with this location.
  2. Save the release file and the start script into any place in your home directory. It is, however, important that both files are placed in the same directory.
  3. Within GRASS, use the command line terminal to change to the directory with the start script and the release file, and to execute the start script with 'sh scriptname'. scriptname stands for the name of the start script of the corresponding test area. All results will be placed in the current directory and in the active GRASS Location.

Instead of starting r.randomwalk with a start script, you could also use the command line or a simple graphical user interface (by just entering 'r.randomwalk' in the command line). With the given start scripts you can easily combine multiple calculations, and you should be able to reproduce the results presented in Publication 2 (even though there will be some minor differences due to randomization). You may manipulate the start scripts in order to explore how changes in the parameter settings influence the model results and the computational times. Please refer to the manual for details with regard to the flags and parameters.

Download links
Acheron Rock Avalanche
New Zealand
Prehistoric rock avalanche in Canterbury with a volume of 6.4 million cubic metres and a travel distance of 3,550 m. The start script consists in procedures for parameter testing and the generation of an impact indicator index.
Kao Ping Watershed
61.5 square kilometre test area with 207 new landslides mapped after the Typhoon Morakot in 2009. The start script includes the back-calculation of the observed landslides and, on this basis, the computation of impact probabilities. The results are validated using ROC Plots.
Unfortunately, the spatial data for this area may not be freely distributed.
Release file
Start script
Gunt Valley
2106 square kilometre area with 113 high-mountain lakes some of which are susceptible to sudden drainage. The start script is used to compute impact indicator scores related to possible future lake outburst floods.