Effect of Landscape Metrics on Varied Spatial Extents of Bangalore, India

H S Sudhira, Priyadarshini J Shetty, Shashikala Gowda, K V Gururaja


Landscape fragmentation and dispersed urban growth can be viewed as both cause and consequence of land-use change
especially in the context of urban growth. Several metrics from landscape ecology has been already applied to quantify the
urban landscapes. However, the use of these metrics in land-use planning and policy making is still lacking. Furthermore, what
is critical is to understand the effect and applicability of these metrics at different scales and extent. Typically these metrics
are applied at given city’s landscape level. However, these may not capture the variations in certain parts of the city since the
estimation of the metrics would get aggregated at the city’s landscape level. In order to examine the effects of some of these
metrics at different spatial extents, a study has been carried out by applying some of the popular landscape metrics for the city
of Bangalore. We have created two subset images of varying extent within the city’s landscape called: north-east and south-west
subsets. Satellite remote sensing data for two time periods 2000 and 2009 were collected and analyzed. Through a multi-stage
classification process, post-classification change detection was performed. A highlight of the research is that it utilizes the Landsat
ETM+ data with Scan Line Corrector (SLC)-off scenes by employing methods to rectify the anomalies. Landscape metrics viz.,
Total Class Area (CA), Percentage of Landscape (PLAND), Shannon’s Diversity Index (Entropy), Largest Shape Index (LSI),
Largest Patch Index (LPI), Clumpiness Index (CLUMPY), Normalized Landscape Shape Index (nLSI) and Contagion Index
(CONTAG) were estimated for the entire extent and the two subsets. The analyses revealed that most metrics at landscape-level
and the class-level suggest similar trends over the two time periods. However, metrics like LPI and CONTAG did capture the
variations across the different extents and time. Some metrics like the CA and PLAND were useful to depict the extent of the
land cover changes. Given the radial pattern of outgrowth for a city like Bangalore, most metrics seem to be conveying similar
responses to the land cover changes despite the variation of extents. On the one hand, this study ascertained the rapidly changing
land cover and its effect on the landscape elements and on the other hand, it could study the effect of landscape metrics on
varied spatial extent. Thus, suggesting that perhaps these metrics could even be applied at varied extents while not affecting the
overall inferences drastically. Finally, the paper concludes stressing the utility of landscape metrics as potential tools that can
be employed for devising land-use policies for future urban expansion.

Full Text: