Welcome

Welcome to the Land Transformation Model web site. The LTM combines geographic information systems, artificial neural networks, geostatistical and remote sensing technologies to forecast land use change.

What is the LTM?

The Land Transformation Model is a land use forecasting model as well as a tool that can be used to examine the spatial and temporal aspects of driving forces of land use change. The model uses a set of spatial interaction rules and machine learning, through neural net technology, to determine the nature of spatial interactions of drivers, such as transportation, urban infrastructure and proximity to lakes and rivers, that have historically contributed toward land use change in the past. This information is then used to conduct forecasting studies.

Example US output

Forecasts of future urban locations were made to the year 2020 and 2040. Each of the over 1300 townships were modeled and the output "stitched" together to create a statewide map of future land use.

You can learn more about how the model was developed and the economic consequences this might have for Michigan's land based industries by selecting projects and the Michigan Land Based Industries project.

Graphics at the right show urban expansion for the Detroit Metropolitan Area. Data from this project were obtained from the South East Michigan Council of Governments (SEMCOG) and from the Michigan Department of Natural Resources.

Download LTM output for Michigan, Wisconsin and Illinois in ArcGIS GRID format.

Get the model

You can acquire the model under a gnu licence agreement. Executables are located in the following zip file. A second version (warning: 126 MB zip file) contains sample data, "cheat data" (data that has been processed already by the model), a multimedia tutorial and the code. In order to learn how to use the model and to complete a tutorial on sample data, go to our LTM Tutorial Multimedia Web site that includes video captures of how to work with ArcGIS to process spatial data and how to work with the other modeling tools. A voice over is provided by the author of the tutorial, Amelie Davis.

A brief explanation of artificial neural networks are found at this link.

Where we've used it

We've been able to model urban expansion, forest expansion, deforestation, and agricultural expansion in the Upper Great Lakes of the United States, Albania and Moldova in Eastern Europe, parts of Western and East Africa, Brazil and Kuala Lumpur, Malaysia. Standard inputs generally are: roads, rivers, elevation, soils, population and two land use maps. These are processed using ESRI's ArcGIS or ArcView software.

The LTM is useful for simulating land use/cover changes across large regions. It can be used to simulate land change in areas that contain several million to even a few hundred million cells. It is thus a useful tool to couple to regional climate, hydrologic and carbon sequestration models.

NSF Biocomplexity Project - CLIP in East Africa

We've been able to model agricultural expansion in East Africa as driven by climate, soils, towns, national parks, roads and elevation. Here is an example "probability map" produced by the LTM for the likelihood for the presence of rainfed agriculture across Kenya and parts of northern Tanzania. The red colors are areas of high likelihood for rainfed agriculture, blue are low, and black represents areas that are non-candidates for rainfed agriculture (parks, current urban). Note the areas around Mt. Kilimanjaro, where rainfed agriculture is currently prevalent along the northeastern slopes of the mountain.

Two types of errors can be mapped and then their importance assessed in the coupling to RAMS. Here is a map of the errors from an LTM simulation (1km cells) and correct location of rainfed (1) and non-rainfed (0) agriculture cells. The black boxes represent the RAMS climate grid (at 36km).

Peer-Reviewed LTM Publications

Robert Gilmore Pontius Jr., Wideke Boersma, Jean-Christophe Castella, Keith Clarke, Ton de Nijs, Charles Dietzel, Zengqiang Duan, Eric Fotsing, Noah Goldstein, Kasper Kok, Eric Koomen, Christopher D. Lippitt, William McConnell, Bryan Pijanowski, Snehal Pithadia, Alias Mohd Sood, Sean Sweeney, Tran Ngoc Trung, and Peter H. Verburg. In press. Comparing input, output and validation maps for several models of land change. Annals of Regional Science.

Pijanowski, B., K. Alexandridis and D. Mueller. 2006. Modeling urbanization in two diverse regions of the world. Journal of Land Use Science (1):83-108.

Tang, Z., B. Engel, K. Lim, B. Pijanowski and J. Harbor. 2005. Minimizing the impact of urbanization on long-term runoff. Journal of the Water Resources Association. 41(6): 1347-1359.

Tang, Z., B. A. Engel, B.C. Pijanowski, and K. J. Lim. 2005. Forecasting Land Use Change and Its Environmental Impact at a Watershed Scale. Journal of Environmental Management. 76: 35-45.

Pijanowski, B., S. Pithadia, K. Alexandridis, and B. Shellito. 2005. Forecasting large-scale land use change with GIS and neural networks. International Journal of Geographic Information Science. 19(2): 197-215.

Wiley, M. J., B. C. Pijanowski, P. Richards, C. Riseng, D. Hyndman, P. Seelbach and R Stevenson. 2004. Combining valley segment classification with neural net modeling of landscape change: A new approach to integrated risk assessment for river ecosystems. Proceedings of WEF 2004 Specialty Conference Series: Watershed 2004, Dearborn Michigan. Water Environment Federation.

Shellito, B. and B. Pijanowski. 2003. Using Neural Nets to Model the Spatial Distribution of Seasonal Homes. Cartography and Geographic Information Systems 30 (3):281-290.

Pijanowski, B.C., D. G. Brown, G. Manik and B. Shellito. 2002. Using Neural Nets and GIS to Forecast Land Use Changes: A Land Transformation Model. Computers, Environment and Urban Systems 26(6) 553-575.

Wayland, K., D. Long, D. Hyndman, B. Pijanowski, and S. Haack. 2002. Modeling The Impact Of Historical Land Uses On Surface Water Quality Using Ground Water Flow And Solute Transport Models. Lakes and Reservoirs 7: 189-199.

Pijanowski, B.C., B. Shellito and S. Pithadia. 2002. Using artificial neural networks, geographic information systems and remote sensing to model urban sprawl in coastal watersheds along eastern Lake Michigan. Lakes and Reservoirs 7: 271-285.

Pijanowsk, B., D. Hyndman and B. Shellito. 2001. The application of the Land Transformation, Groundwater Flow and Solute Transport Models for Michigan's Grand Traverse Bay Watershed. Proceedings of the American Planning Association, New Orleans, Lousiana, March 14, 2001.

Pijanowski, B.C., S.H. Gage, and D.T. Long. 2000. A Land Transformation Model: Integrating Policy, Socioeconomics and Environmental Drivers using a Geographic Information System; In Landscape Ecology: A Top Down Approach, Larry Harris and James Sanderson eds.