Pan-India analysis of relationship between Spatial Attributes of Urban Area and Population

Urban structures in any city needs to be analyzed in conjunction to Urban Green Spaces (UGS). The relations between spatial attributes of built-up and UGS Land use/ Land Cover (LULC) can help analyzing various ecosystem services like micro-climate problems in aspects of increasing Land Surface Temperature (LST) patterns causing Urban Heat Island (UHI) inside the city. These relations between both LULC can also improve aesthetic structure of city. India, a magnanimous country comprising of 36 administrative boundaries, shows a range of diversity in population and culture inhabited by its dwellers. These large population centres have different settlement characteristics at different administrative levels (States/Union Territories, Districts, Sub-Districts, Villages/Towns and Wards/Blocks etc.) of India. These settlements can affect climate and development of country in longer duration. As such spatio-temporal analysis of urban population dynamics over different constituent land use/land cover (LU/LC) is performed in this study using open source data and software programs only. The study derives a pattern of Landscape Metrics (LSM) of built-up LULC over a period of 30 years in 7 zones of India comprising of 694 districts in total of various 28 states and 8 UTs. Landscape Metrics are one of the efficient ways to analyze the patterns of LULC in a study area. Publically available data such as Pan India Decadal LULC by ORNL DAAC for year 2005 and Copernicus Global Land service LULC for year 2015 at 100m resolution has been used as classified maps in study. These decadal LULC maps are predominantly classified using multi-temporal Landsat series data for Pan India coverage giving annual LULC classification maps consisting 19 classes with overall classification accuracy of 0.94 for all 3 year data. Built-up class present in both classified maps are used for analysis as urban patches. Landscape metric analysis is done through landscapemetrics library in RStudio® and 34 of the class level landscape metrics were calculated for urban area using multi-patch analysis for multi-year data. Significance of metrics was determined through calculation of coefficient of determination and establishment of variable importance between all 34 landscape metrics for urban and Population averaged over states and UTs containing 694 districts units of India. "Number of Patches (NP)","Total Class Area (CA)", "Total Core Area (TCA)" and "Total Edge (TE)" stood out as most viable metrics showing relation as high as R² of 0.82 between spatial attributes of urban patches and population in the Indian administrative units. Spatial relation in terms of zones of India is much more existent than temporal as yearly variation for relation between urban patches and population. North, West and North East Zone of India are showing most consistent and highest values of correlation whereas South zone and UTs lowest with Central zone being most inconsistent. Such high relations between spatial patterns of urban patches and population suggest a significant need to prioritize configuration and optimization of population in cities, which can not only affect urbanization pattern inside the city boundary but also help achieving the sustainability causes of ecosystem services in city boundary.