Monitoring and Modeling Urban Temperature Patterns in the State of Iowa, USA, Utilizing Mobile Sensors and Geospatial Data

• Published in: Applied sciences 2024-11, Vol.14 (22), p.10576
• Main Authors: Abbeg Coproski, Clemir, Liang, Bingqing, Dietrich, James T., DeGroote, John
• Format: Article
• Language: English
• Subjects: Accuracy; Algorithms; Cities; Data collection; Data mining; Datasets; Geospatial data; Heat; LiDAR; Machine learning; mobile sensors; Morphology; Optical radar; random forest; Regression analysis; Remote sensing; Rural areas; Sensors; Spatial data; Temperature; Urban areas; Urban climatology; Urban heat islands; urban morphology; urban temperature patterns; Vegetation
• ISSN: 2076-3417; 2076-3417
• DOI: 10.3390/app142210576

Thorough investigations into air temperature variation across urban environments are essential to address concerns about city livability. With limited research on smaller cities, especially in the American Midwest, the goal of this research was to examine the spatial patterns of air temperature across multiple small to medium-sized cities in Iowa, a relatively rural US state. Extensive fieldwork was conducted utilizing manually built mobile temperature sensors to collect air temperature data at a high temporal and spatial resolution in ten Iowa urban areas during the afternoon, evening, and night on days exceeding 32 °C from June to September 2022. Using the random forest machine-learning algorithm and estimated urban morphological variables at varying neighborhood distances derived from 1 m2 aerial imagery and derived products from LiDAR data, we created 24 predicted surface temperature models that demonstrated R2 coefficients ranging from 0.879 to 0.997 with the majority exceeding an R2 of 0.95, all with p-values < 0.001. The normalized vegetation index and 800 m neighbor distance were found to be the most significant in explaining the collected air temperature values. This study expanded upon previous research by examining different sized cities to provide a broader understanding of the impact of urban morphology on air temperature distribution while also demonstrating utility of the random forest algorithm across cities ranging from approximately 10,000 to 200,000 inhabitants. These findings can inform policies addressing urban heat island effects and climate resilience.