Bhartendu Pandey

The present study examines (1) the capability of VIIRS NTLs to explain urban infrastructure access variations as observed from census data and (2) urban infrastructure inequalities in India. In the absence of detailed household-level infrastructure access and amenities data, we compiled a spatial database of ward-level household infrastructure access and amenities data that could be used to study multi-scalar infrastructure inequalities in association with other spatially explicit database. In addition, we applied a time series analysis algorithm to VIIRS time series data that summarizes the variation into four components: (1) intercept, (2) long-term trend, (3) phase, and (4) amplitude, which could be used to study the time series signal. In the present study, we analyzed the intercept term to understand how well VIIRS NTLs capture urban infrastructure access variations. We find that NTLs are limited in explaining infrastructure access variations across scales and that there is a nonlinear relationship between the two. Our results also show that infrastructure access/amenity levels follow a Kuznets-type relationship, which is consistent with a Bernoulli’s process. In addition, our results show highest level of inequalities in the case of access to sewer system and lowest in the case of access to electricity. We show that these patterns in inequality could be linked to urbanization but there are significant inter-urban inequalities in case of fully urbanized regions, which are related to urban size. These patterns are indicative of a combined spatial and hierarchical diffusion process that will be examined in future research.

Find the published paper here: https://www.pnas.org/doi/10.1073/pnas.2119890119