Data  and Methods

An integrated approach involving compilation of both spatial and non-spatial data from government agencies and institutions, application of spatial and temporal analyses using Remote Sensing data, GIS techniques and conventional field survey (ground truthing) were adopted in this study.  The main sources of primary data were from field (using GPS), the Survey of India (SOI) toposheets of 1:50,000, 1:250,000 scale and multispectral sensors (MSS) digital data of the IRS (Indian Remote Sensing satellites) -1C and IRS -1D (1998, 2002, 2006). LISS-III MSS data scenes corresponding to the district for path-rows [(100,63) (100,64) and (101, 64)] and cloud cover less than 2 % were procured from the National Remote Sensing Agency, Hyderabad, India (http://www.nrsa.gov.in). LISS-III (band-2, band-3, and band-4) and PAN images provide a spatial resolution of 23.5 m and 5.8 m respectively. The secondary data was collected from the government agencies (Directorate of census operations, Agriculture department, Forest department and Horticulture department). Multi spectral sensor data (G, R and NIR bands) was used for land cover and land use analyses. Panchromatic data of 5.8 m spatial resolution (IRS-PAN) was used for village level fuel wood species mapping. Primary data for supervised classification of MSS data was collected through stratified random sampling using GPS. This collection of primary data (Ground control points-GCPs) helped to correlate the ground information (attribute information) with the remote sensing data. This entailed:

1. Generation of geo-refernced base layers like district boundary, district with taluk and village boundaries, road network, drainage network, contours, mapping of waterbodies, etc. from the SOI toposheets of scale 1:250000, 1:50000 and revenue maps of scale 1:6000.
2.  Extraction of bands (LISS3 with resolution 23.5m and PAN with resolution 5.8m) from the data (in BIL and BSQ format) respectively procured from NRSA.
3. Geo-correction of bands through resampling using ground control points (GCP's). GCP’s were chosen such that they can be easily identified in the image (like road and railway crossings, edges of large fields and approximate middle points of small lakes). It was taken care that the GCP’s are of adequate number and equally spread throughout the image.
4. Cropping and mosaicing of data corresponding to the study area. – Kolar district  
5. Histogram generation, Bi-spectral plots, Regression analysis.
6. Computation and analysis of various vegetation indices.
7. Generation of FCC (False Colour Composite) by saturating 2.5% from each end of the gray scale using linear contrast stretch method and identification of the heterogeneous patches regions for training data collection (ground truthing).
8. Collection of attribute information from field corresponding to the chosen training sites using GPS.
9. Classification of remote sensing data- land use analyses (both district wise and taluk wise), Statistical analysis and report generation.
10. Fusion of LISS3 and PAN data using fusion algorithm such as IHS (Intensity, Hue, Saturation).
11. Village wise vector layers of specieswise tree  distribution was prepared using calibrated Global Positioning System (GPS) through an extensive field survey, taking the village as a sampling unit. Villages were selected so as to represent the entire taluk / district. The pilot surveys were undertaken to ascertain the sample size – the number of villages needed to be surveyed in order to obtain a reliable picture of bioresources in the region and their growing stock (wood volume in case of trees). The villages to be surveyed in different taluks were decided based on the extent of vegetation cover in respective taluks.
12. Stratified (based on land holding) random sampling of households in select villages to assess the fuel wood demand.