Machine-Learning-Assisted Characterization of Regional Heat Islands with a Spatial Extent Larger than the Urban Size

Yin Du,Zhiqing Xie,Lingling Zhang,Ning Wang,Min Wang,Jingwen Hu

DOI: https://doi.org/10.3390/rs16030599

IF: 5

2024-02-06

Remote Sensing

Abstract:Surface urban heat islands (SUHIs) can extend beyond the urban boundaries and greatly affect the thermal environment of continuous regions over an agglomeration. Traditional urban-rural dichotomy depending on the built-up and non-urban lands is challenged in characterizing regional SUHIs, such as how to accurately quantify the intensity, spatial pattern, and scales of SUHIs, which are vulnerable to SUHIs, and what the optimal scale for conducting measures to mitigate the SUHIs. We propose a machine-learning-assisted solution to address these problems based on the thermal similarity in the Yangtze River Delta of China. We first identified the regional-level SUHI zone of approximately 42,328 km2 and 38,884 km2 and the areas that have no SUHI effects from the annual cycle of land surface temperatures (LSTs) retrieved from Terra and Aqua satellites. Defining SUHI as an anomaly on background condition, random forest (RF) models were further adopted to fit the LSTs in the areas without the SUHI effects and estimate the LST background and SUHI intensity at each grid point in the SUHI zone. The RF models performed well in fitting rural LSTs with a simulation error of approximately 0.31 °C/0.44 °C for Terra/Aqua satellite data and showed a good generalization ability in estimating the urban LST background. The RF-estimated daytime Aqua/SUHI intensity peaked at approximately 6.20 °C in August, and the Terra/SUHI intensity had two peaks of approximately 3.18 and 3.81 °C in May and August, with summertime RF-estimated SUHIs being more reliable than other SUHI types owing to the smaller simulation error of less than 1.0 °C in July–September. This machine-learning-assisted solution identified an optimal SUHI scale of 30,636 km2 and a zone of approximately 23,631 km2 that is vulnerable to SUHIs, and it provided the SUHI intensity and statistical reliability for each grid point identified as being part of the SUHI. Urban planners and decision-makers can focus on the statistically reliable RF-estimated summertime intensities in SUHI zones that have an LST annual cycle similar to that of large cities in developing effective strategies for mitigating adverse SUHI effects. In addition, the selection of large cities might strongly affect the accuracy of identifying the SUHI zone, which is defined as the areas that have an LST annual cycle similar to large cities. Water bodies might reduce the RF performance in estimating the LST background over urban agglomerations.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to solve the problem of the expansion of urban heat island effects (Urban Heat Islands, UHIs) in urban agglomeration areas. In particular, when the scope of these heat islands exceeds the scale of the city itself, how to accurately quantify their intensity, spatial pattern and scale. Specifically, the research mainly focuses on the following aspects:

1. **How to accurately quantify the intensity, spatial pattern and scale of the urban heat island effect**: The traditional urban - rural division method based on the dichotomy of built - up areas and non - built - up areas is difficult to accurately describe the large - scale urban heat island effect, especially in urban agglomeration areas. Therefore, a new method is needed to solve these problems.

2. **Identify areas vulnerable to the urban heat island effect**: The research attempts to determine which areas are most vulnerable to the urban heat island effect, so as to provide targeted mitigation measures for urban planners and decision - makers.

3. **Determine the optimal scale to implement measures to mitigate the urban heat island effect**: The research hopes to find an optimal spatial scale, so that the mitigation measures implemented at this scale can most effectively reduce the impact of the urban heat island effect.

### Research background

Urban areas are usually warmer than the surrounding rural areas, and this phenomenon is called the urban heat island effect (UHI). With the rapid urbanization, industrialization and transportation development, the urban heat island effect has expanded from a single city to a larger urban agglomeration area. This regional heat island effect has a significant impact on the regional thermal environment, threatening environmental sustainability, biodiversity, and resource and energy consumption. Therefore, accurately understanding the intensity, spatial pattern and scale of the urban heat island effect is crucial for urban planning and environmental management.

### Method overview

To achieve the above goals, the research proposes a machine - learning - based method, using satellite data (such as MODIS/LST) and the Random Forest (Random Forest, RF) model to characterize the urban heat island effect in the Yangtze River Delta Urban Agglomeration (Yangtze River Delta Urban Agglomeration, YRDUA). The specific steps include:

1. **Define the intensity of the urban heat island effect**: Define the intensity of the urban heat island effect at each grid point by calculating the difference between the actual land surface temperature (LST) and the background temperature:
\[
\Delta T_i = T_i - T_{Bi}
\]
where \(\Delta T_i\) is the intensity of the urban heat island effect at the \(i\) - th grid point, \(T_i\) is the actual land surface temperature, and \(T_{Bi}\) is the background temperature.

2. **Identify urban heat island effect areas and non - urban heat island effect areas**: Use the similarity of land cover types and the long - term annual cycle of land surface temperature to distinguish urban heat island effect areas from other areas.

3. **Estimate the intensity of the urban heat island effect**: Estimate the background temperature by training the Random Forest model, and further calculate the intensity of the urban heat island effect. The model uses climate factors, geographical factors, terrain variables, biophysical parameters and urbanization parameters as input features.

### Results

The research has successfully identified the strong urban heat island effect areas and areas vulnerable to the urban heat island effect in the Yangtze River Delta urban agglomeration, and excluded heat islands unrelated to urbanization. In addition, the research has also obtained the intensity, spatial pattern and scale of the urban heat island effect at each grid point, providing a basis for further exploring the seasonal variation of the regional urban heat island effect and its relationship with vegetation change and urbanization.

### Significance

The results of this research are helpful for urban planning and environmental management, especially in shaping a reasonable urban agglomeration structure and improving human comfort. By providing accurate urban heat island effect information, it can help formulate effective strategies to deal with urban heat waves and related health risks.