Urban Heat Island Studies Research Articles

Local climate zone-based urban land cover classification from multi-seasonal Sentinel-2 images with a recurrent residual network

The local climate zone (LCZ) scheme was originally proposed to provide an interdisciplinary taxonomy for urban heat island (UHI) studies. In recent years, the scheme has also become a starting point for the development of higher-level products, as the LCZ classes can help provide a generalized understanding of urban structures and land uses. LCZ mapping can therefore theoretically aid in fostering a better understanding of spatio-temporal dynamics of cities on a global scale. However, reliable LCZ maps are not yet available globally. As a first step toward automatic LCZ mapping, this work focuses on LCZ-derived land cover classification, using multi-seasonal Sentinel-2 images. We propose a recurrent residual network (Re-ResNet) architecture that is capable of learning a joint spectral-spatial-temporal feature representation within a unitized framework. To this end, a residual convolutional neural network (ResNet) and a recurrent neural network (RNN) are combined into one end-to-end architecture. The ResNet is able to learn rich spectral-spatial feature representations from single-seasonal imagery, while the RNN can effectively analyze temporal dependencies of multi-seasonal imagery. Cross validations were carried out on a diverse dataset covering seven distinct European cities, and a quantitative analysis of the experimental results revealed that the combined use of the multi-temporal information and Re-ResNet results in an improvement of approximately 7 percent points in overall accuracy. The proposed framework has the potential to produce consistent-quality urban land cover and LCZ maps on a large scale, to support scientific progress in fields such as urban geography and urban climatology.

A bibliometric review of past trends and future prospects in urban heat island research from 1990 to 2017

The urban heat island (UHI) phenomenon is among the most evident features of human impact on the Earth’s system. This phenomenon has been widely observed and documented in many cities around the world. UHI-related publications have increased rapidly over the last three decades. However, because of a refined methodology and widening scope, a holistic understanding of research patterns and issues related to UHI research is lacking. Although others have summarized developments in UHI studies, these publications have focused on describing the current state of research rather than uncovering research trends and prospects. In the present study, we examined the evolution of UHI-related research from 1990 to 2017 and applied a scientometrics approach to identify research trends. The characteristics of publication outputs, key scientific disciplines, and cooperation between countries and institutions were determined by a citation analysis. We also discuss research trends, including future directions, approaches, and expected data. We identified two potential directions for UHI research through the results of key co-word clustering and discriminant analyses: negative impacts of UHI on public health and strategies to mitigate and adapt to UHI effects. We provide a broad review of the development of UHI research that may inspire future studies on the UHI phenomenon by new researchers in this field.

A study of urban heat island of Banda Aceh City, Indonesia based on land use/cover changes and land surface temperature

This article described the spatial and temporal of land surface temperature (LST) patterns in Banda Aceh City, Indonesia, in the context of urban heat island (UHI) phenomenon. Landsat imaginary in 1998 and 2018 was used in this study, which represents the conditions before and after the tsunami. Geographic Information System (GIS) and Remote Sensing (RS) technique were used for data analysis. The 1998 and 2018 LUC maps were derived from remote sensing satellite images using a supervised classification method (maximum likelihood). Both LUC maps contained five categories, namely built-up area, vegetation, water body, vacant land, and wet land. The 1998 LUC map had a kappa coefficient 0.91, while the 2018 LUC map had 0.84. It was found that the built-up area increased by 100%, while the vegetation category fell 50%. The overall mean LST in the study area increased 5.9 0 C between 1998 and 2018, with the highest mean increase in the built-up area category. The study recommends that LST should be taken into consideration in urban planning process to realize sustainable urban development. It also emphasizes the importance of optimizing the availability of green open space to reduce UHI effects and helps in improving the quality of the urban environment.

A study of urban heat island of Banda Aceh City, Indonesia based on land use/cover changes and land surface temperature

This article described the spatial and temporal of land surface temperature (LST) patterns in Banda Aceh City, Indonesia, in the context of urban heat island (UHI) phenomenon. Landsat imaginary in 1998 and 2018 was used in this study, which represents the conditions before and after the tsunami. Geographic Information System (GIS) and Remote Sensing (RS) technique were used for data analysis. The 1998 and 2018 LUC maps were derived from remote sensing satellite images using a supervised classification method (maximum likelihood). Both LUC maps contained five categories, namely built-up area, vegetation, water body, vacant land, and wet land. The 1998 LUC map had a kappa coefficient 0.91, while the 2018 LUC map had 0.84. It was found that the built-up area increased by 100%, while the vegetation category fell 50%. The overall mean LST in the study area increased 5.90C between 1998 and 2018, with the highest mean increase in the built-up area category. The study recommends that LST should be taken into consideration in urban planning process to realize sustainable urban development. It also emphasizes the importance of optimizing the availability of green open space to reduce UHI effects and helps in improving the quality of the urban environment.

The coupling relationship between construction land expansion and high-temperature area expansion in China’s three major urban agglomerations

ABSTRACT With the development of urbanization in China, construction land (CL) expansion is obvious, and urban heat island (UHI) is becoming more and more serious. Many scholars have shown a strong interest in the study of UHI caused by land expansion, but less attention has been paid to the coupling relationship between CL expansion and high-temperature area (HTA) expansion during the process of urbanization. This study obtained the land surface temperature (LST) data of China’s three major urban agglomerations (CTM-UAs) in 2000, 2005, 2010 and 2015. Then this paper used overall coupling model and spatial coupling model to reveal the coupling relationship between the CL expansion and HTA expansion of CTM-UAs in different periods. The results showed that: (1) From 2000 to 2015, the trend of CL expansion and HTA expansion in CTM-UAs was synchronous; (2) The distance between the CL barycentre and HTA barycentre in CTM-UAs was shrinking, the spatial overlap was constantly strengthening, and the overall coupling was constantly strengthening. And the highest value of the overall coupling was reached in 2015; (3) The number of grid units of very strong coupling and strong coupling increased with fluctuation in Beijing-Tianjin-Tangshan Urban Agglomeration (BTT), that in Yangtze River Delta Urban Agglomeration (YRD) was increasing, and that of Guangdong-Hong Kong-Macao Urban Agglomeration (GHM) has gone through the process of increasing to decrease.

Urban Geocryology: Mapping Urban–Rural Contrasts in Active-Layer Thickness, Barrow Peninsula, Northern Alaska

The maximum depth of seasonal thaw is a critical design factor for civil infrastructure in permafrost regions. Although maps of active-layer thickness (ALT) have been created for localized areas in undisturbed terrain, this has rarely been done within urbanized areas. The modified Berggren solution was used to map ALT at a resolution of 30 × 30 m over the 150-km2 Barrow Peninsula in northern Alaska. Emphasis was placed on analyzing differences in accuracy obtained in urbanized and relatively undisturbed tundra. Although the modified Berggren solution is known to provide more accurate estimates of frost and thaw depth than the Stefan solution, it has not been used previously in mapping applications. As part of the Barrow Urban Heat Island Study, seventy-one miniature data loggers were installed in and surrounding the City of Utqiaġvik (formerly Barrow) to measure air and soil temperature. The resulting data were used to calculate air and soil surface temperature fields, as well as summer n-factors, based on nine urban and rural land-cover classes. Regional soil and land-cover maps were used to obtain additional input data. Validation was performed by comparing probed ALT measurements with predicted pixel values. Model results confirm that the presence of urban infrastructure increases both the magnitude and the geographic variability of ALT relative to surrounding undisturbed tundra. The Berggren solution performed well for estimating mean values for land-cover classes in both rural and urban areas and has considerable potential as a tool for mapping ALT in other applications. Key Words: active layer, Alaska, Barrow, frozen ground, geocryology, mapping, permafrost, urban, Utqiaġvik.

Urban heat island, urban climate maps and urban development policies and action plans

Abstract This paper aims at drawing attention to the deficiencies of urban heat island (UHI) mitigation strategies and urban climate maps (UCMs) towards practical implementations and efficient interventions in urban development policies and action plans (UDP/AP). The paper critically reviews the knowledge of the UHI and UCMs to identify the major shortcomings in the production of action plans. The shortcomings are identified in three aspects including (1) active involvement in UDP/AP (2) urban managerial and governmental actions, and (3) public engagement and participation. More specifically, it is first discussed that the current synthesizing paradigm of UCMs and UHI studies only considers the physical aspects of the problem/solution. Second, climate recommendations have paid a little attention to the imperative role, capacities and motivations of urban governance and managerial organizations. Third, mitigation strategies have been proposed without considering public needs and willingness. To overcome these challenges, a collaborative platform of UCM systems should be developed through a holistic approach. Such platform should provide a ground for efficient interventions in UDP/AP as well as interactions of all actors and indicators. To this end, the technical challenges of the UHI studies and UCM systems to link with UDP/AP are discussed. Second, UCMs should explicitly define the role of urban climate governance, prioritize the actions and state the policy agendas with regard to local government capacities and motivational elements. Finally, local communities’ needs’ and stockholders’ opinions should be involved in the synthesizing process. As a potential solution, need-base approaches and weighing systems should be further developed to consider all urban actors’ perspectives.

Effects of source and sink landscape pattern on land surface temperature: An urban heat island study in Wuhan City

Effects of source and sink landscape pattern on land surface temperature: An urban heat island study in Wuhan City

How relevant are local climate zones and urban climate zones for urban climate research? Dijon (France) as a case study

Several typologies of urban surface properties have been proposed, in recent years, for urban heat island studies and climate modeling. Some were specifically developed for cities and urban climate issues, like the Urban Climate Zones, and the more recent Local Climate Zones. The initial objective of this paper is to evaluate the capacity of these two typologies to identify thermal environments in and around cities, and to determine which typology best captures the daily spatio-temporal patterns of surface and urban canopy heat islands. To simulate urban climate with a model, LULC data based on a given typology are required. To avoid circularity, we combined the Corine Land Cover database, with data for the whole of Europe, and the BD TOPO database, for France, to form a new tool, CLC_USGS, which we used as input for the WRF limited-area model, with a 150-m grid resolution. The capacity of each typology to identify coherent thermal zones was estimated for Dijon, a medium-sized French city, during a three-week heat wave, over a 24-hour period, in conditions favorable to urban heat island development. The comparison was based on hourly air temperatures directly output from the WRF model, those obtained from the purpose-built MUSTARDijon 47-sensor meteorological network, and NDVI values and land surface temperatures estimated from Landsat images for 11 July 2015 at 1030 UTC. Typical diurnal variations and spatial contrasts of surface and air temperatures were identified in both simulations and observations. As both typologies show significant capacity for identifying thermally coherent intra-urban areas, this study suggests that they could both be useful for urban climate applications. The typology that is most generally applicable in worldwide contexts is Local Climate Zones.

Feature Importance Analysis for Local Climate Zone Classification Using a Residual Convolutional Neural Network with Multi-Source Datasets

Global Local Climate Zone (LCZ) maps, indicating urban structures and land use, are crucial for Urban Heat Island (UHI) studies and also as starting points to better understand the spatio-temporal dynamics of cities worldwide. However, reliable LCZ maps are not available on a global scale, hindering scientific progress across a range of disciplines that study the functionality of sustainable cities. As a first step towards large-scale LCZ mapping, this paper tries to provide guidance about data/feature choice. To this end, we evaluate the spectral reflectance and spectral indices of the globally available Sentinel-2 and Landsat-8 imagery, as well as the Global Urban Footprint (GUF) dataset, the OpenStreetMap layers buildings and land use and the Visible Infrared Imager Radiometer Suite (VIIRS)-based Nighttime Light (NTL) data, regarding their relevance for discriminating different Local Climate Zones (LCZs). Using a Residual convolutional neural Network (ResNet), a systematic analysis of feature importance is performed with a manually-labeled dataset containing nine cities located in Europe. Based on the investigation of the data and feature choice, we propose a framework to fully exploit the available datasets. The results show that GUF, OSM and NTL can contribute to the classification accuracy of some LCZs with relatively few samples, and it is suggested that Landsat-8 and Sentinel-2 spectral reflectances should be jointly used, for example in a majority voting manner, as proven by the improvement from the proposed framework, for large-scale LCZ mapping.

The role of structure cooling to reduce the effect of urban heat island in Mumbai

ABSTRACTIn India, for centuries, structure cooling principles have been part of strong architectural heritage. Buildings like Taj Mahal, Gol Gumbaz and temples of Hampi are a few notable examples of naturally cooled buildings. Some have used thermal mass, some have used water, while others use the combination of both to provide thermal comfort in the building as well as outside the building. Anthropogenic activities in urban areas increase usage of devices like air conditioning for human comfort inside a building. Typically, these principles discourage any external and/or induced forms of cooling. Less use of artificial systems means less heat and emitting less greenhouse gases in the environment. In an illustrative experiment, structural cooling system is introduced in a commercial building in Mumbai (hot and humid climate) to reduce the cooling load of the conventional air-conditioning system reducing overall heat load. An air-conditioning system typically adds ∼30% more heat from the compressors. This increases the ambient temperature of the area. In dense urban environments, this process escalates till a heat island is formed [Oke, T. (1982). The energetic basis of the urban heat island. Journal of the Royal Meteorological Society, 108(455), 1–24; Weng, Q., Dengsheng, L., & Schubring, J. (2004). Estimation of land surface temperature – vegetation abundance relationship for urban heat island studies. Remote Sensing of Environment, 89, 467–483]. A combination of structure cooling and an air conditioner can be used to significantly reduce load on the air conditioners. It will drastically reduce the energy consumption and resultantly the pollution in the environment. In Mumbai, India’s largest city, the ambient temperatures are sometimes 40–45°C in peak summers. To reduce the cooling load of the air-conditioning system, the structural cooling system was introduced as an innovative technique to reduce heat inside the structure and also in the environment. Application of such technique will be very useful on the structures like airport, station areas and commercial complexes to reduce the heat and pollution in the environment.

URBAN HEAT ISLAND MICRO-MAPPING VIA 3D CITY MODEL

Abstract. Urban Heat Island (UHI) phenomenon has been a topic of intense study over the past several years. However, to visualise UHI model is still an issue. Common visualisation of UHI by using digital thematic maps shows that it is hard to perceive its impacts especially in a sophisticated micro-area such as in urbanized cities. Moreover, different building façade’s material gives different UHI value. Therefore, there is a need in computing and visualising this phenomenon in three-dimensional (3D) perspectives. Recently, the development of 3D city modelling shows the potential of solving these gaps. This can be seen from the characteristics of 3D city models that are suitable in representing micro-areas (complex cities) for UHI studies. Based on this issue, this research aims to produce a 3D UHI model by using 3D city models as a tool for efficient and sustainable building design. The main objective is to produce a new approach in visualising UHI in 3D perspectives by instigating 3D city models. Thus, the UHI effect could be predicted precisely by calculating the building façades value. This research explores the 3D shadow analysis, 3D solar radiation and 3D orientation analysis in UHI modelling via 3D city models. From the analyses, the results show that the 3D city models are capable in presenting the solar radiation value for each building façade. Furthermore, this approach can be used to simulate future UHI analysis-prediction and advantageous for pre-development planning.

Understanding Land Surface Temperature Differences of Local Climate Zones Based on Airborne Remote Sensing Data

The local climate zones (LCZ) scheme has attracted the interest of climate researchers as it enables the standardized study of urban heat islands by combining thermal and physical parameters of built and natural structures. Most recent work on LCZ has concentrated on understanding air temperature differences, adapting the scheme to different contexts and improving satellite-based classification methods. However, studies using very high-resolution imagery, including 3-D descriptors and analyzing their land surface temperature (LST) variability are scarcer. Since a correct delineation of LCZ implies significant temperature differences among classes, the aims of this study are 1) to test a GIS-based method for the classification of LCZ based on cut-off values valid to the Australian context; and 2) to examine the quality of classifications by analyzing the LST variability among LCZ using high-resolution airborne remote sensing data. Results show that diurnal and nocturnal mean LSTs significantly differ among most LCZs. Welch's ANOVA and subsequent post hoc tests for pairwise comparisons also demonstrate that these differences prevail for 71.8% of zones during the daytime and 73.6% of zones at night. Overall, LCZs A, 8, and 3 are the most distinguishable zones during the daytime and LCZs D, G, 1, 4, and 5 are well differentiated at night. In contrast, LCZs 1 and 4 are the least distinguishable at daytime and LCZs 10, A, and E are not well differentiated at night. The present study has successfully validated the present airborne-based classification method which is contingent on further accuracy assessment and improvements.

Landsat-based vegetation abundance and surface temperature for surface urban heat island studies: the tale of Greater Amman Municipality

ABSTRACTThis study aimed at characterizing the spatial and temporal characteristics of vegetation abundance (represented by At-Sensor Normalized Difference Vegetation Index (ASNDVI) and/or Land Surface Normalized Difference Vegetation Index (LSNDVI)) and surface temperature (represented by At-Sensor Brightness Temperature (ASBT) and/or Land Surface Temperature (LST)) and investigating the types of the relationships between these variables in different seasons in different years using Landsat data in Greater Amman Municipality in Jordan. Correlation analysis using Pearson’s correlation coefficient showed that using either ASNDVI or LSNDVI and either ASBT or LST should lead to the same results. Change analysis from 1987–2016 using map algebra showed that the majority of the municipality was subjected to both no change and small increase in vegetation abundance in summer and winter and both no change and small increase in surface temperature in summer and both no change and small decrease in surface temperature in winter. Vegetation abundance variability reflected partly the land cover characteristics of the municipality while surface temperature variability indicated the potential of the presence of an inverted surface urban heat island pattern. Correlation analysis using contingency tables showed the presence of negative relationship in summer and positive relationship in winter between vegetation abundance and surface temperature.

Assessing local climate zones in arid cities: The case of Phoenix, Arizona and Las Vegas, Nevada

The local climate zone (LCZ) classification scheme is a standardization framework to describe the form and function of cities for urban heat island (UHI) studies. This study classifies and evaluates LCZs for two arid desert cities in the Southwestern United States – Phoenix and Las Vegas – following the World Urban Database and Access Portal Tools (WUDAPT) method. Both cities are classified into seven built type LCZs and seven land-cover type LCZs at 100-m resolution using Google Earth, Saga GIS, and Landsat 8 scenes. Average surface cover properties (building fraction, impervious fraction, pervious fraction) and sky view factors of classified LCZs are then evaluated and compared to pre-defined LCZ representative ranges from the literature, and their implications on the surface UHI (SUHI) effect are explained. Results suggest that observed LCZ properties in arid desert environments do not always match the proposed value ranges from the literature, especially with regard to sky view factor (SVF) upper boundaries. Although the LCZ classification scheme was originally designed to describe local climates with respect to air temperature, our analysis shows that much can be learned from investigating land surface temperature (LST) in these zones. This study serves as a substantial new resource laying a foundation for assessing the SUHI in cities using the LCZ scheme, which could inform climate simulations at local and regional scales.

Satellite Images and Gaussian Parameterization for an Extensive Analysis of Urban Heat Islands in Thailand

For the first time, an extensive study of the surface urban heat island (SUHI) in Thailand’s six major cities is reported, using 728 MODIS (MODerate Resolution Imaging Spectroradiometer) images for each city. The SUHI analysis was performed at three timescales—diurnal, seasonal, and multiyear. The diurnal variation is represented by the four MODIS passages (10:00, 14:00, 22:00, and 02:00 local time) and the seasonal variation by summer and winter maps, with images covering a 14-year interval (2003–2016). Also, 126 Landsat scenes were processed to classify and map land cover changes for each city. To analyze and compare the SUHI patterns, a least-square Gaussian fitting method has been applied and the corresponding empirical metrics quantified. Such an approach represents, when applicable, an efficient quantitative tool to perform comparisons that a visual inspection of a great number of maps would not allow. Results point out that SUHI does not show significant seasonality differences, while SUHI in the daytime is a more evident phenomenon with respect to nighttime, mainly due to solar forcing and intense human activities and traffic. Across the 14 years, the biggest city, Bangkok, shows the highest SUHI maximum intensities during daytime, with values ranging between 4 °C and 6 °C; during nighttime, the intensities are rather similar for all the six cities, between 1 °C and 2 °C. However, these maximum intensities are not correlated with the urban growth over the years. For each city, the SUHI spatial extension represented by the Gaussian footprint is generally not affected by the urban area sprawl across the years, except for Bangkok and Chiang Mai, whose daytime SUHI footprints show a slight increase over the years. Orientation angle and central location of the fitted surface also provide information on the SUHI layout in relation to the land use of the urban texture.

Analyzing the diurnal variations of land surface temperatures for surface urban heat island studies: Is time of observation of remote sensing data important?

Land surface temperature (LST) images show the nonexistence of surface urban heat island (SUHI) effect over Jaipur and Ahmedabad cities during the daytime, whereas a clear picture of SUHI effect during nighttime. But in Ludhiana, diurnal SUHI effect has been observed. Hence the contrast diurnal LST pattern of three cities in India has been analyzed in the study. The analysis indicates that nighttime images can be used for the analysis of SUHI effect over any area. The use of daytime images for this analysis depends on the surface cover and time of observation which may show a false temporary SUHI effect. Time of observation is very important for analyzing the efficacy of remote sensing data for diurnal LST variations for different land covers and SUHI studies. SUHI varies during the day, and it is more significant during the evening/night. Diurnal LST relationship at different time intervals with various land surface parameters like vegetation, built-up, soil, etc. has also been investigated in the present research. Various land surfaces play a significant role in the diurnal variations of LSTs. The study discusses the adequacy of the observation time, from the viewpoint of heat island measures strategies.

Responses of Urban Land Surface Temperature on Land Cover: A Comparative Study of Vienna and Madrid

The relationship between the land cover (LC) characteristics and the land surface temperature (LST) is significant for surface urban heat island (SUHI) study and for sustainability research. To better understand how the land surface temperature (LST) responds to LC, two urban areas, Vienna and Madrid, with different climatic conditions are selected and compared, using Landsat-8 OLI data and urban atlas data. To determine a suitable scale for analyzing the relationship between LC and LST, a correlation analysis at different sizes of spatial analytical scales is applied. To demonstrate the LC composition effects on LST, a regression analysis of the whole study area and in the specific circumstance is undertaken. The results show that: (1) In the summer, Vienna presents high temperature in the urban areas and low temperature in the surrounding rural areas, while Madrid displays the opposite appearance, being relatively cooler in the urban areas as compared to the rural areas, with the main different factors affecting elevated urban LST; (2) Suitable analytical scales are suggested in studying the LC–LST relationship between different LC characteristics in the two study areas; (3) Negative effects on the LST appear when the area of cooling sources, such as water or urban greenery, reaches 10% at a 990 × 990 m2 scale in Vienna. Built-up area is the main factor affecting elevated urban LST where such areas cover the majority at a 990 × 990 m2 scale in Madrid. These findings provide a valuable view regarding how to balance the urban surface thermal environment through urban planning.

Pixel-Wise vs. Object-Based Impervious Surface Analysis from Remote Sensing: Correlations with Land Surface Temperature and Population Density

Impervious surface areas (ISA) are heavily influenced by urban structure and related structural features. We examined the effects of object-based impervious surface spatial pattern analysis on land surface temperature and population density in Guangzhou, China, in comparison to classic per-pixel analyses. An object-based support vector machine (SVM) and a linear spectral mixture analysis (LSMA) were integrated to estimate ISA fraction using images from the Chinese HJ-1B satellite for 2009 to 2011. The results revealed that the integrated object-based SVM-LSMA algorithm outperformed the traditional pixel-wise LSMA algorithm in classifying ISA fraction. More specifically, the object-based ISA spatial patterns extracted were more suitable than pixel-wise patterns for urban heat island (UHI) studies, in which the UHI areas (landscape surface temperature >37 °C) generally feature high ISA fraction values (ISA fraction >50%). In addition, the object-based spatial patterns enable us to quantify the relationship of ISA with population density (correlation coefficient >0.2 in general), with global human settlement density (correlation coefficient >0.2), and with night-time light map (correlation coefficient >0.4), and, whereas pixel-wise ISA did not yield significant correlations. These results indicate that object-based spatial patterns have a high potential for UHI detection and urbanization monitoring. Planning measures that aim to reduce the urbanization impacts and UHI intensities can be better supported.

The effect of multi-dimensional indicators on urban thermal conditions

Urban heat island (UHI) studies have recognized ten factors as increasing the inner-city temperature compared with that of the surrounding suburbs. The UHI effect is a leading cause of heat-related diseases and mortality in many nations. However, there are still two main shortcomings. First, the effect of UHI is not well recognized in arid and semi-arid regions. Second, the association of multi-dimensional information with surface temperature in urban areas must be examined. This study focuses on the height-related aspects of urban geometry in an arid region. A range of multispectral and spatial vector data were used to derive the surface temperature and two-dimensional (2D) and three-dimensional (3D) information of the study area. All information was aggregated into a grid with common spatial resolution to create a homogeneous dataset. The machine learning statistical model of a boosted regression tree (BRT) was used to reflect the relative influence of 2D and 3D indicators with land surface temperature. Our results showed a cooler surface temperature in the city than in the surrounding area, leading to the question of whether the established UHI definition encompasses all types of cities. In addition, the thermal band was able to distinguish different spatial structures in the study area. The BRT analysis demonstrated that both multi-dimensional 2D and 3D indicators affect the surface temperature. In particular, the 3D indicators play a more important role than 2D indicators in shaping the surface temperature at different urban geometries of the study area. This new method can help urban planners identify the most influential 2D and 3D indicators that affect the surface temperature in different districts of a city.