Long-term monitoring, predicting and connection between built-up land and urban heat island patterns based on remote sensing data

Abstract

The alterations observed in urbanized areas have given rise to urban climate change, contributing to the emergence of urban heat islands (UHIs). This study investigates changes and predicts the built-up land/UHIs in Rasht city from 1991 to 2031. Built-up lands were classified using the normalized built-up composite index (NBCI) and their prediction for 2031 was performed. Surface biophysical parameters were then derived for the prediction of land surface temperature (LST) for 2031 using multiple linear regression (MLR) and Markov chain-cellular automata (CA-Markov) modeling. Finally, alterations in both built-up land and UHI within the city were scrutinized across various geographical directions and temporal periods. The study's findings reveal commendable overall classification accuracy for NBCI (ranging from 87% to 91% across different years) and CA-Markov (89%) in 2021. The MLR analysis produced favorable results with a root mean square error of 1.33 K in predicting LST for 2021. The significant correlation (R = 0.89) between changes in built-up lands and UHI indicatesthat built-up land/UHI exhibit a notable degree of freedom and sprawl, resulting in a negative urban degree-of-goodness.These results demonstrate the direct effects of built-up lands on UHI changes. Therefore, by determining the appropriate pattern in the built-up lands, it is possible to control the pattern of UHI. These findings hold practical significance for urban planners, offering valuable insights to mitigate adverse impacts on the urban environment.

Keywords

• Remote sensing data
• Built-up lands
• Urban heat island (UHI)
• Prediction
• Degree-of-goodness