Land is one of the most important natural resources, as life and various developmental activities are based on it (George et al 2016). Land use refers to the human induced changes for agricultural, industrial, residential or recreational purposes whereas land over describes features that are present on the earth‘s surface (Bharath and Ramachandra 2012). The study of land use and land cover change is one of the foremost arenas to understand the degree of interaction between man and environment. The changing nature of land use and land cover depends on the existing ecological factors and the level of socio-economic development.

Land has proven to be a key component in the development of the human population and is viewed as one of the most significant natural resources currently available. This observation brings into question the recent debates surrounding the pressures people tend to impose on the land, which have resulted in transformations in its physical landscape and usage. Mans impact on the earth in terms of the transformations in land cover and land use have rapidly increased over the years. However, as a result of these continuous land transformations, planning and designing sustainable urban development has become challenging due to the additional fact that the available mapped resources of the land

The environmental factors reveal the status of land as a key and finite resource for most human activities including agriculture, industry, forestry, energy production, settlement, recreation and water sources and storage (Thanushkodi et al 2012). The deterioration of environment due to population explosion and its interference leading to the over exploitation of natural resources, pollution and random land use for basic needs. Hence it is essential to collect information about various changes occurring in land use and land cover for proper land management practices (Attri et al 2015, Twisa and Buchroithner 2019). Remote sensing based change detection analysis enables the user to analyse the transformations of land use and land cover as it is able to provide consistent coverage at short intervals. The major advantages of remote sensing systems is their capability for repetitive coverage, which is necessary for change detection studies at global and regional scales. With the increasing ability to quantify and monitor the expansion of urban environments, remote sensing offers users a set of spatially consistent data, greater spatial precision and an overall higher resolution when compared to that of existing aerial imagery (Kulo 2018).

Change detection analysis are used to monitor the dynamic nature of biophysical and anthropogenic features on the earth’s surface. Several techniques are developed in literature for change detection analysis using post classification comparison (El-Hattab 2016), conventional image differentiation (Afify 2011), image ratio (Shaoqing and Lu 2008), image regression (Luppino et al 2018), and manual on-screen digitization of change principal components analysis and multi date image classification (Deng et al 2008). A variety of studies have addressed that post-classification comparison was found to be the most accurate procedure and presented the advantage of indicating the nature of the changes.

Rawat and Kumar, 2015 carried out a change detection analysis by employing supervised classification methodology using maximum likelihood technique for Hawalbagh block of district Almora, Uttarakhand, India in order to understand the spatio-temporal dynamics of land use land and found maximum increase in vegetation and built up areas while decrease in water body and agricultural lands. Similarly Haque and Basak, 2017, used both pre-classification and post classification change detection approach to assess the change from over the decades in rich biodiversity area Tanguar Haor, Sunamganj, Bangladesh. They implemented CVA, NDVI and NDWI techniques as a pre-classification approach to assess the change scenario. To obtain the classified data in a timely manner and also for efficient planning and management of the land resources satellites are the best resources to provide the data in a timely manner (Babykalpana and ThanushKodi, 2010).

Considering the capabilities of the remote sensing techniques present study made an effort to analyse the land use changes in rural areas as there is an increasing need for proper land use planning to control various problems in the regional areas. This helps the government and policy makers in understanding relationships and interactions between human as well as natural resources for better decision making.

The study area, Thirthahalli taluk is located in the south western part of Karnataka with an aerial extent of 1252.59 sq kms (Figure 1). The study area is bounded by the Western Ghats Mountains with an average elevation of 566 meters above the mean sea level (Figure 2). The study area is one of the "Hottest biodiversity hotspots" and has different varieties of flora, fauna, mammal species, bird species, fresh water fish species and insects. The area enjoys tropical climate throughout the year. Generally, the weather is very pleasant in the all parts of the study area. The evapotranspiration is normally high in ghat sections. Summer prevails between March to early June, the wet months start from early June to September, October and November months experience scanty rain by North East monsoon. The study area receives an average annual rainfall of around 3000mm/ season (Nischitha et al 2014) during the monsoon season. Throughout the monsoon season, the unbroken Western Ghats chain acts as a barrier to the moisture laden clouds which forcefully rises the clouds and the process deposit most of their rain on the windward side of the study area. The rainfall pattern suggests a steady decline in rainfall as we move from west to east. The average daily temperature of the area is about 24.5⁰ C.

In order to study the land use land cover changes in the study region medium scale Landsat 5 Thematic Mapper (TM) and Landsat 8 Operational Land Imager and Thermal Infrared Sensor satellite data is used for the years 1997 and 2017. The data were obtained for the month of April from United States Geological Survey (USGS) (https://earthexplorer.usgs.gov/), an Earth Science Data Interface for geo-spatial data. Specifications of the satellite data acquired for change analysis are given in Table 1. In addition to using satellite imagery, ancillary data is collected which included the ground truth data was in the form of reference data points collected using Geographical Positioning System (GPS) in the year 2017 which was used for image classification and accuracy assessment of the results. Addition to this the study used SPOT (2017) image from Google Earth for creation of the road network map and settlement map of the study area.

Initially, the image files are downloadable in Landsat Level 1 Data Products that standard radiometric and geometric correction was processed. As each band file is provided unlayered in GeoTIFF output format, the downloaded band files were layer stacked in ERDAS Imagine for the further analysis. Then the images are reprojected to match the projection of the study area boundary in order to subset the area of interest of the study region from the huge Landsat 5 TM and Landsat 8 datasets. The projection used for the images in the present study is UTM WGS 1984 Zone 43 North. The same projection is maintained for other thematic maps also. The digitization process is performed to extract the road network map and settlement map from the SPOT image downloaded from Google Earth.

Subsequently, the supervised classification method is applied for the classification of the Landsat images. It is a statistical process which can be used to extract the class descriptors. The process involves the selection of representative samples for each land cover class based on the visual interpretation of the image characteristics like color, tone, size, shape, pattern, texture, association etc. The details were then ground checked to verify the doubtful areas. The classification scheme was designed keeping in view of the management practices addressing each land use land cover parcels, the Landsat data are categorized in to Settlements, Water body, Agricultural crop land, Agricultural plantations, Forest and Waste lands under the level I classification. Supervised classification uses the spectral signature defined in the training sites and applies them to the entire image. For instance, it determines each class on what it resembles most in the training set. Afterward, the maximum likelihood classification algorithm was used for the supervised classification of the images. This is the type of image classification that is mainly controlled by the analyst by selecting the pixels that are representative of the identified classes (Bolstad and Lillesand 1991). Further, visual comparison of the classified maps is carried out to make the change detection analysis to find out the changes associated with land use land cover distribution in the study area.

Mapping and identifying land use and land cover change is the most important, as well as the most widely researched, topic in remote sensing as the information from it has been used extensively to derive a number of biophysical variables, such as vegetation index, biomass, biodiversity, deterioration of environmental quality, loss of agricultural lands, destruction of wetlands. More importantly, land use land cover pattern and its change reflect the underlying natural or social processes, thus providing essential information for modelling and understanding many different phenomena on the Earth. Knowledge of land use land cover and its change is also critical to effective planning and management of natural resources.

Present study area is one of the biodiversity hotspot and is one of the tourist attraction in Karnataka state. Here identified the study area is well connected with the road networks to the settlements identified with three types of transportation network metalled roads, cart tracks and unmetalled roads. The transportation network map of Thirthalli taluk is showed in figure 3.