Forest landscapes with diverse life forms are complex interactive ecosystems that support economy evident from the estimate of 4.7 trillion dollars annually through goods and services from global forest ecosystems (Krieger, 2001; de Groot et al., 2012). Unplanned developmental activities have led to the destruction of pristine forests and grasslands at regional as well as global scales (Lambin and Meyfroidt, 2011) evident from barren hilltops, reduced duration of stream flow, etc. Anthropogenic forces have modified over 83% of the Earth's land cover threatening sustenance of biological diversity (Sanderson et al., 2002 Fischer and Lindenmayer, 2007).
The conversion of natural ecosystems is the second largest driver of human-induced climate change, which accounts for 10% of anthropogenic carbon dioxide (CO²) emissions (Le Quere et al., 2013). Changes in forested landscape alter composition and configuration, affecting respective ecosystem functions. Fig. 1 illustrates various drivers of landscape degradation and its role in climate change. LULC change alters the homogeneous landscape into a mosaic of heterogeneous patches, which leads to the habitat fragmentation with smaller and more isolated patches (Ramachandra et al., 2016a). Consequences are reduction in natural resources availability (Vinay et al., 2013), biodiversity (Hansen et al., 2004), and changes in local climate (Chase et al., 2000; Bharath et al., 2013; Ramachandra 2014) as well as to global climate (Bagley et al., 2014; Val Martin et al., 2015; Halder et al., 2016). This emphasizes the need for conservation of forests to mitigate changes in the climate and to sustain water and food.

Fig. 1. Consequences of LULC changes.

Protected areas (PA), national parks (NP), sanctuaries, nature reserves, wildlife refuges, wilderness areas have been created through policy initiatives in order to protect the native habitat of endemic species and to reduce the magnitude of land conversion. Thus, PA system has evolved strategically to protect and maintain biological diversity, cultural resources at local to global scales. Maintaining ecological integrity in the protected area with buffer region is essential as most of protected landscapes are open systems that face anthropogenic and other biotic threats from adjacent areas. Alterations in landscape structure with a reduction in contiguous forests would increase the likelihood of invasive plants and animal range expansions, alter hydrologic regime (water availability), which leads to the erosion of integrity of the protected ecosystems. These are supposed to be managed through legal or other effective means from extinction especially those on the brink of extinction (Gaston et al., 2008). Globally, establishing PAs has gained impetus for conservation and according to the International Union for the Conservation of Nature (IUCN), nearly 13% of the global land surface is now under some form of protection. Land use land cover (LULC) mapping and monitoring of PAs can serve as important indicators of landscape and environmental status, distributions, and patterns. PA networks, conservation reserves, sanctuaries are created under various policy initiatives (Wild Life (Protection) Act 1972, Forest Conservation Act 1980, Environment (Protection) Act 1986, Biodiversity Act 2002) in order to conserve the forests from unregulated exploitation. The landscape surrounding protected areas also known as buffer region, influences PA's ability to maintain ecosystem functions and achieve conservation goals. The health of any PAs indirectly depends on surrounding buffer regions (Kintz et al., 2006). Drastic LU changes in buffer regions around PAs can reduce their effective size and limit their ability to conserve biodiversity because of alterations in ecological processes and the ability of organisms to move freely among protected areas (Hamilton et al., 2013). LULC changes in buffer zones surrounding protected ecological reserves will have serious implications in the management and conservation of protected areas.
LULC changes involving conversion of land with vegetation cover to other land uses (built up, agriculture, barren land) have also increased land surface temperature (LST) (Mallick et al., 2008; Pal and Ziaul, 2016) due to alterations in the degree of absorption, surface temperature, evaporation rates, storage of heat, energy and water balance processes (Li et al., 2016). LULC induced LST changes play a vital role in many environmental processes affecting microclimate through associated biophysical changes (D'Odorico et al., 2013; Bharath et al., 2013). Moreover, warming aggravates tree mortality, regeneration of species with the induced water stress, which would further reduce forest cover (Allen et al., 2015). Vegetation die-off will rapidly alter ecosystem nature and properties at a regional scale. Thus, it is essential to monitor LULC changes to mitigate the impacts of changes in climate at a global scale. Multi-resolution (spatial, spectral, radiometric and temporal) remote sensing (RS) data with Geographical Information System (GIS) and modelling techniques provided a precise evaluation of the spread and health of forest ecosystem and forecasting. Accurate up to date LULC change information helps in addressing environmental consequences, protecting natural resources and planning (Ramachandra et al., 2016b). Modelling and visualization is considered as a conceptual, mathematical approach to account driving forces of change in a landscape such as socioeconomic, political, natural and cultural factors (Verburg et al., 2004; Batty and Torrens, 2005; Ramachandra et al., 2014). Cellular automata (CA), Markov chain models, and agent-based modelling approaches have been implemented for comprehensive projection of a region (Matthews et al., 2007; Bharath et al., 2014). Since forest land use changes are non-linear process, CA-Markov LULC models are efficient in prediction compared to the traditional mathematical models for decision making and planning (Mondal et al., 2016). It is simple and provides a better understanding of how nonlinear process can aptly be simulated with less complexity (Pontius and Malanson, 2005; Grinblat et al., 2016). CA-Markov modelling technique exhibits better capabilities of descriptive power and simple trend projection of spatial changes as compared to other techniques (Chen et al., 2013). Objectives of the study are
(i) to analyze current status and trends in temporal LULC patterns of forest cover in and around select PAs of central Western Ghats (Karnataka) from 1973 to 2016,
(ii) assessment of temporal variations in temperature with change in LULC,
(iii) modelling and visualization of spatial patterns of forests in 2026.