Field verification showed that newly built up areas obtained from digital classification of remotely sensed data, had the maximum number of small-scale industries, IT companies, multistoried building and private houses that came up in the last one decade. However, to account for the micro-level study of the land use with a detailed growth model, the classification process would require further refinement with other techniques like bootstrap, a computer implementation of the nonparametric or parametric ML estimation that provide direct computational way of assessing uncertainty and estimates of standard errors (Hastie, et al., 2001) along with high spatial resolution data like IRS-1C PAN (with 5.8 m) or Cartosat-1/2 (with 2 m/1 m spatial resolution). Supervised learning seeks to extract information from labeled samples. If the underlying distribution comes from a mixture of component densities described by a set of unknown parameters Ө, then Ө can be estimated by ML methods. Pattern classifiers along with the advances in geo-informatics coupled with the availability of higher spatial, spectral and temporal resolution data help in extracting spatial features of interest like land cover classes such as built up. In this context, an important application of pattern classifiers would be to estimate accurate temporal land cover statistics that are useful in monitoring the status and extent of these features. The analysis showed a rapid growth of urban pockets and consequent decline of water bodies and vegetation in Greater Bangalore.

Shannon’s entropy computed for Bangalore city for 2000 and 2006, (1.0325 and 1.0782) are closer to the upper limit of log n, i.e. 1.0986, showing the higher degree of dispersion of built-up in the city.

Urbanisation and the consequent loss of lakes has led to decrease in catchment yield, water storage capacity, wetland area, number of migratory birds, flora and fauna diversity and ground water table. Temporal analyses of waterbodies in Greater Bangalore indicate the decline of 32.47% during 1973 to 1992, 53.76% during 1973-2002 and 60.83% during 1973-2007 consequent to a linear growth of 466% of built up/urban area. As land is converted, it loses its ability to absorb rainfall. Urbanisation has increased runoff 2 to 6 times over what would occur on natural terrain in some pockets of Bangalore.

The relationship between LST and NDVI investigated through the Pearson’s correlation coefficient at a pixel level and the significance tested through one-tail Student’s t-test, confirms the relationship for all LC types. Also, increased urbanisation has resulted in higher population densities in certain wards, which incidentally have higher LST due to higher level of anthropogenic activities.

The growth poles are towards N, NE, S and SE of the city indicating the intense urbanisation process due to growth agents like setting up of IT corridors, Peenya industrial units, etc. The growth in northern direction can be attributed to the new International Airport, encouraging other commercial and residential hubs. The southern part of the city is experiencing new residential and commercial layouts and the north-western part of the city outgrowth corresponds to the Peenya industrial belt along the Bangalore-Pune National Highway 4. The forecast of the growth of built up pixels in various directions during the next 15 years would be 48% in N, 51% in NE, 41% S and 38% in SE directions.