Rajan, KS 
  
Also, the model takes into account the external influences that are likely to effect shifts in the agricultural patterns. The 
main factor considered here is that the external demand generated from export policies of specific crops, like cassava in 
Thailand, lead to large-scale shifts from conventional agricultural practices. As of now, these external influences are 
exogenous variables and are not calculated within the model. Figure 5. (a) shows the simulated land use map of 
Thailand in 1990 based on the model simulation run from 1980-90; and (b) the actual land use map of Thailand in 1990. 
6.2 Migration Sub-Model 
The populations are both a cause to population changes and also affected by them. In case of developing new areas for 
development, they change the land use. But, for increasing populations with limited incomes due to the limits to the 
existing land use, they respond by migrating out to already established areas like urban centers. The changes in 
population are calculated based on the per capita income changes subjected to a maximum population density, along 
with the data on its current age distribution and the educational levels. 
6.2.1 Urban Migration: The model results for Urban out migration are dependent on the rural to urban migration and 
the natural population growth rate of the urban area. As there is an increase in population pressure in the existing urban 
grids, the density readjustment occurs leading to an out migration of the population to the neighbouring grids around the 
existing urban areas. This out migration is always accompanied by a land use change decision to urban areas. 
6.2.2 Rural Migration: The migration of households to new grids (development) or partially agricultural grids that are 
accompanied by land use conversion to agricultural lands determine the rural migration. Figure 6. shows the result of 
the total in and out-migration from the individual grids, in the Nan province during the simulation period of 1980-90. 
  
    
Out mig. sum 
0 
      
Eu 96 - 
101-138 
136-175 
EH 176 - 205 
INN 206 -235 
EH 236 -270 
EH 271-375 
BH 376 -1115 
  
  
  
  
Figure 6. Shows the result of the total in and out-migration from the individual grids, 
in the Nan province during the simulation period of 1980-90. 
7 DISCUSSION AND CONCLUSIONS 
The major feature of the AGENT-LUC model is in characterizing the behavioural aspects of the “agent” in transforming 
each grid of the land unit. The model integrates both the micro-characteristics of the land unit like the biophysical 
conditions of the grid and also takes into account the macro features that are characterized at the administrative 
boundary levels. The detailed structure of the model provides us with various outputs that help in fine-tuning the model 
to the area of applicability based on the scale of the data that is available. The model structure is modular in nature, 
allowing for the model to be expanded as and when new knowledge on related systems get available, like a macro- 
economic trade model to give the export demand and prices of the international market or a forest land cove change 
model. 
The scale and resolution of the data available vary rather widely for different regions. It is of utmost importance to 
develop a model framework that can be easily ported across the various scales and can be implemented. As the basic 
framework of this model is dependent on mostly global datasets and can derive most of its information for the national 
level datasets, it can be readily applied to areas where these data can be made available. If a particular region or 
country is able to provide the model with more accurate and finer datasets, then the agent decision model may have to 
be tuned to the local conditions, without there being a need to alter the model structure. 
  
1218 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000.