Modelling landscape dynamics has the history of traditional urban growth modelling approaches. Subsequently, in 1960’s to manage urban sprawl, modelling of urban sprawl was undertaken (Batty et al., 1999; Torrens &Alberti, 2000; Lowry 2001; Walter, 1975; Allen, 1979; Pumain, 1986). The approach involved linking independent to dependent variables, which were statistically significant, additive as in a linear model or a non-linear model but tractable in a mathematical way. However, these models although used mostly for policy purposes, could not be useful when processes involved rule-based systems, which in practice cannot be tractable mathematical operations (Batty & Torrens, 2001). Initial attempt to model urban dynamics was based on complexity (Forrester, 1969) involving dynamic relations represented by stocks and flows which determined the various activity volumes in the city, and were synthesized from knowledge and observation of causal factors. A key distinction of this model was its ability to represent emergent behaviour of the system originating out of complexity. However, this model could not be represented spatially. Batty et al. (1999) and Sudhira et al. (2004) provided spatially aggregate model for the urban sprawl phenomenon. Cheng and Masser (2003) report spatial logistic regression techniques for analysing urban growth pattern, considering the causal factors which was applied for a city in China. Geographically weighted regression was employed to identify spatial interaction between level of regional industrialisation and for analysing spatial non-stationarity of different factors affecting regional industrialization (Huang and Leung, 2002) as conventional regression analysis would only produce the ‘average’ and ‘global’ parameter estimates which vary over space depending on the respective spatial systems. .Allen (1986), Couclelis (1987) and Engelen (1988) model urban systems as complex systems, acknowledging the self-organisation in urban systems. Capturing urban systems as discrete models gained further momentum with the cellular automata (CA) based techniques (Li & Yeh, 2000; White et al, 1993; Wolfram, 1984, 2002). Subsequently CA based simulation of urban growth was done (Couclelis, 1987; Diana et al., 2010; Batty & Xie, 1994). Current approaches of modeling spatial dynamics are based on land cover and land use dynamics (Yang, 2003), urban growth models (Batty & Xie, 1997; Batty, 1998; Clarke & Gaydos, 1998; Clarke et al., 1996; Couclelis, 1987; Jianguan, 2002; White et al., 1993, 1997), simulation of urban growth process (Li & Yeh, 2000; Torrens, 2000; Vyasalu & Reddy , 1985; Vyasalu, 1985). There are several models considering the spatial and temporal dynamics such as LUCAS (Land Use Change Analysis System) model (Berry et al., 1996), GIGALOPOLIS (Clarke et al., 1996), and California Urban Futures (CUF-II) model (Landis & Zhang, 1996). Li and Yeh (2000) developed and demonstrate the constrained CA model for sustainable urban development modelling. Some of these models interact with causal factors driving the sprawl such as the availability of land and proximity to city centres and highway. The calibration and prediction of the CA model was done coupling with GIS for generating long-term prediction for urban growth (Clarke & Gaydos, 1998).

CA has been used for simulating urban growth quite successfully mostly considering various driving forces that are responsible for sprawl. However some issues like the impact on ecology, energy, environment and economy for taking policy decisions have not been addressed effectively. In this context, the integration of agent-based models and CA models, where agent-based models would help in capturing the externalities driving the processes. CA with agent-based models would help in identifying the location of the sprawl that help in effective visualization and understanding of the impacts of urban sprawl. However, to achieve an efficient simulation of urban sprawl, modelling has to be attempted in both spatial and non-spatial domain. Modelling urban sprawl in non-spatial domain is mainly by the application of statistical techniques while CA models and agent-based modelling are known to complement modelling in spatial domain. Integration of CA and agent-based models to simulate urban sprawl phenomenon has been done through Geographic Automata Systems (GAS) framework (Benenson and Torrens, 2004), Dynamic Geo-Spatial Simulation (DGSS) framework (Sudhira et al., 2005) the swarm optimization model (Fenga et al., 2011).

The spatial visualization capacity of CA as well as the computational effectiveness are documented by Liu (2008). The swarming methods such as PSO and ant colony optimization have only emerged in the last decade, such as the geospatial reasoning (Parunak, et al., 2006), optimal path (Li, et al., 2009). The Swarming intelligence method was also used in urban geospatial scenario based on Multiple Perspectives and agent based interaction (Parunak et al., 2010). Other endevours include Agent-based ecological modelling based on swarm intelligence (Perez and Dragi civic, 2011) spatial clustering analysis (Kuo and Lin, 2010). Wang et al., (2008) attributes the aspects of human geography, urban geography and economic geography, on a series of subjects such as shape and direction of urban expansion, spatial evolution processes, dynamic mechanisms of urban expansion, transformation of farmland to urban land and internal differentiation of urban land. It is imperative to address the landscape dynamics at various levels and through appropriate metrics or indicators for effective regional planning and sustainable utilization of natural resources. The following research issues need to be addressed

• Land cover dynamics with the optimal combination of multi-resolution data for cost effective landscape dynamics analysis and mapping (Urban, peri-urban, rural).
• Land use dynamics with the multi-resolution data over various landscape (Urban, peri-urban, rural).
• Visualizing of growth patterns of Cities.