Prediction of Land Use/ Cover Change by Using LCM Model

Authors

1 Faculty of Natural Resource and Marine Science, Tarbiat Modares University Noor, Mazandaran

2 Department of Environmental Sciences, Gorgan University of Agricultural Sciences and Natural Resources

3 Department of Soil Sciences, College of Agriculture, Isfahan University of Technology

Abstract

Population growth leads to urban and industrial expansion, cropland expansion and land use/cover
change. Present research aims to model and predict land use/cover change by using LCM, based on
three management scenarios which call continuity scenario, agriculture-conservation scenario and
changes-limitations in Deylaman region located in Guilan Province. LCM uses Markov chain model,
Multi-Layer Perceptron and MOLA for producing transition submodels, transition probability map
and finally land use change prediction. Validation of model was investigated by calculation standard
kappa index of agreement, kappa for quantity, kappa for location and proportion of Hits to the total
pixels. The mentioned indices were equal to 0.79, 0.82 and 0.85, respectively and proportion of Hits
to the total pixels was equal to 22 percent. Modeling results based on continuity scenario showed 9%
and 23.4 %decrease in forest and rangeland, respectively in the next 40 years. Also the results showed
that management of the area based on changes-limitation scenario would result in more conservation
of natural ecosystems and prevent the irregular expansion of cropland.
 

Keywords