Construction of continuous local climate zone time series from all available Landsat data for assessing the variability of urban climate

Abstract

Comprehensive change information of urban surface heterogeneity with reference to diversified clusters of climate-relevant elements is essential for evaluation and management of the evolving urban thermal environment. Previous temporal assessments in this regard are primarily based on analysis of discrete images at multiple time points which failed to reveal the long-term time-consistent and continuous changes. Construction of highly frequent time-series of surface microclimatic units by leveraging the temporal dimensions of satellite observations is efficient for understanding the urban climate dynamics. This study implements a robust time-series analysis of all available Landsat (5, 7 & 8) observations (1992–2022) to construct a continuous time-series of Local Climate Zones (LCZ), a climate-based classification system of urban surface, for the case of Kolkata. The time-dependent parameters of land surface temperature were engaged to explore urban thermal zones (UTZs), a more climate dependent units of urban thermal environment, for assessment of current and expected future thermal behaviour of LCZs. Results bring out that LCZs exhibited systematic transformation trajectories (non-urban to open built-up to compact built-up), and diversified growth patterns in terms of landscape metrics and growth modes. The compact built-up LCZs were increasingly agglomerated while non-urban LCZs were increasingly fragmented over time to intensify the thermal environment. The LCZ-UTZ association analysis identifies the strongest thermal sensitivity for heavy industry LCZ with 80.37 % and 55.02 % of its area to be highly intensive and highly potential for warming in future, respectively. The dense built up LCZs showed higher thermal intensity and lower potential compared to open built-up LCZs.