As part of a research work, I am modelling Land Use/Land cover changes over the span of 4 decades on a 40 km² watershed in a rural area located in sahelian climate (west africa). I am mainly interested in three types of land units, namely bare soils, cultivated soils and vegetated areas. I have been able to map them accurately enough using remote sensing analysis of Landsat 5 TM/Landsat 8 OLI Images. As such, I now have land cover maps for the watershed at four dates, 1985, 1995, 2007 and 2017. I am now looking into modelling the changes using a set of driving factors (yet to be identified) and assess land cover changes. I ran into land use simulation models mostly based on Cellular Automaton (CA) concepts such as Land Use Sim (http://www.landusesim.com/landusesim-land-use-modeling-simulation-software/) but it seems to be a paid software. Are there any other free simulation software/packages one can point me at ? Preferably, something that can be easily tight to a GIS environnment (ESRI ArcGIS for example) for easier raster/vector processing.
Thanks in advance.
Roland.
A free software for land use modelling is "Modules for Land Use Change Simulations - MOLUSCE" (https://plugins.qgis.org/plugins/molusce/), which is a plugin for QGIS (https://www.qgis.org/en/site/). You should first download and install QGIS, and then install MOLUSCE from the repository.
Documentation on MOLUSCE can be found on the following links:
http://nextgis.com/blog/molusce/
https://github.com/nextgis/molusce/tree/master/doc/en
http://wiki.gis-lab.info/w/Landscape_change_analysis_with_MOLUSCE_-_methods_and_algorithms
For a contemporary resource you might try the Global Ecological Land Units mapping of the world. It's a 2015 product of Esri and the USGS. Impressive and available thorugh Esri, which had a white paper on it, and the USGS (http://rmgsc.cr.usgs.giv/outgoing/ecosystems/Global.
If you can't find it (that may be an old URL) you can message pub. info at USGS or at Esri.
A free software for land use modelling is "Modules for Land Use Change Simulations - MOLUSCE" (https://plugins.qgis.org/plugins/molusce/), which is a plugin for QGIS (https://www.qgis.org/en/site/). You should first download and install QGIS, and then install MOLUSCE from the repository.
Documentation on MOLUSCE can be found on the following links:
http://nextgis.com/blog/molusce/
https://github.com/nextgis/molusce/tree/master/doc/en
http://wiki.gis-lab.info/w/Landscape_change_analysis_with_MOLUSCE_-_methods_and_algorithms
There are several LUCC models for free. One quite widespread and famous is Dinamica Ego (http://csr.ufmg.br/dinamica/). You also have the CLUE model (http://www.ivm.vu.nl/en/Organisation/departments/spatial-analysis-decision-support/Clue/index.aspx) developed by a research team in Netherlands.
Also in this paper ( An open and extensible framework for spatially explicit land use change modelling: The lulcc R package) you can find the presentation of a R package which includes two LUCC models, being CLUE one of them if I am remember well.
For a general overview, you have short presentations of several LUCC models in the last part of this book: http://www.springer.com/us/book/9783319608006
I have given you above the links to some of them. However, I am sure there are another ones different which are also for free.
it is known many free software such as QGIS, R language , paython an other a lot
You could try Netlogo! You can use it for both Cellular Automata and Agent-Based Modelling, which is also increasingly used to model LUCC.
Netlogo may be less suitable for large or very complex simulations than other options, but it's free and very easy to learn. That's also why it's widely used, which in turn leads to the availability of a lot of documentation.
QGIS and GRASS GIS are the most popular free and open source software for spatial analysis and modeling.
GRASS GIS contains stimulation models and many more features and functions.
For example FUTURES Urban regional simulation model
Please refer following url.
https://grass.osgeo.org/grass72/manuals/addons/r.futures.html
Niroshan
It's a bit late answer, but I hope it could help others.
According to the previous answers, there are some useful tools for this purpose such as MOLUSCE and Semi-automatic Classification Plugin (SCP) plugins in QGIS (below links).
However, as you know land change detection procedures is highly depend on the accuracy of your images classification. The more appropriate image classification method is, the higher accuracy change detection has.
Change detection can be easily obtained from the difference of two classified images. We have many image classification methods but in my point of view, they can divide into two main parts: traditional and new trend methods. The first part divide into supervised and unsupervised classification algorithms (such as maximum likehood, minimum distance, etc.) and the second part is machine learning algorithms (ANN, SVM, Deep learning, etc.).
Thus, I suggest you to focus more on your image classification method and find out which algorithm is more suitable for your case (There are many open source codes in python for various image classification algorithms as below links). Then, classified your images from time t1 to time t2 based on selected method. Finally, the difference of images will give you change detection map from t1 to t2.
Regards
1. https://fromgistors.blogspot.com/p/semi-automatic-classification-plugin.html
2. https://www.machinalis.com/blog/python-for-geospatial-data-processing/
3. https://www.machinalis.com/blog/obia/
LuccME is an open source spatially-explicit modeling framework available at:
http://luccme.ccst.inpe.br/en/home-2/
You can choose existing components of allocation/potential/demand or create new ones. The package mentioned above in R was based on it, but does not have all the options. The core components are based on the clue model (continuous and categorical land use variables). For non-programmers, it has an easy to use GUI. It has documentation in English and people available to help.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images