# The influence of city size versus urban form on land surface temperature variation and the surface urban heat island effect: A cross-city analysis of German cities

## TL;DR

This study examines how city size and urban form affect land surface temperatures in German cities, finding that urban form patterns are linked to higher temperatures.

## Contribution

The paper introduces a novel method using high-resolution urbanization-level data and landscape metrics to analyze the surface urban heat island effect.

## Key findings

- The surface urban heat island effect is present in all studied German cities.
- City size does not positively correlate with increased heat stress.
- Urban form patterns related to density and shape are associated with higher land surface temperatures.

## Abstract

This paper seeks to: 1) determine the extent of the intra-urban surface urban heat island (SUHI) effect within German cities using novel urbanization-level data with high spatial resolution, and 2) to assess the influence of city size and urban form using this data together with landscape metrics. The study uses aggregated Landsat LST data to measure the intra-urban SUHI effect across German cities of various population sizes. The first stage of the research utilizes urbanization-level data based on differing thresholds of population and building density, providing a novel method of measuring the correlation between the intra-urban SUHI phenomenon and density. The second stage assesses the impact of urban form by calculating specific landscape metrics using high-resolution land cover data, which is then clustered to identify homogeneous patterns among cities. We show that the SUHI effect can be found within all cities included in the study; however, a positive correlation between increase in city size and increase in heat stress could not be confirmed. The study shows that German urban areas are characterized by specific patterns of urban form that correspond with size, and specific urban forms particularly related to density and shape correspond with higher overall LST.