Overcoming built data-scarcity in developing cities: Hidden Markov methods to construct reliable building footprint data across urban climate risk zones

Although many studies have examined the impact of urban form on surface urban heat island intensity (SUHII) based on local climate zones (LCZ), the spatial-temporal heterogeneity within and between LCZs remains underexplored. Additionally, there is a need to identify key urban form indicators influencing the thermal environment across different LCZs. To address these gaps, this study investigates the influence of urban form on the spatial heterogeneity of SUHII at a 150 m x 150 m grid scale, focusing on the relationship between the two from the LCZ perspective. Using Macau, a subtropical high-density area, as a case study, this paper calculated urban form indicators, SUHII, and LCZ based on multi-source data. It applied geographically weighted regression (GWR) and geographically and temporally weighted regression (GTWR) models to analyze the spatial-temporal non-stationarity between urban form and the urban thermal environment. Additionally, Geodetector was used to rank the interpretive strength of driving factors within each LCZ class. Results show that SUHII of LCZ 8 is the highest. Comparing three regression models (OLS, GWR, and GTWR), the GWR model demonstrates the best fit for accounting for thermal variations resulting from changes in urban form. The regression performance varies significantly among LCZ classes, particularly between compact and open types. Furthermore, the results of Geodector can serve as references for urban form optimization for cooling communities, according to different contributions of driving factors in each LCZ class. For instance, reducing building density is generally relatively effective in all types of LCZ, and increasing porosity has the most significant and typical cooling effect in LCZ 2. For open LCZs, NDVI is the key factor to improve their thermal environments. This study sheds new light on the spatio-temporal variations of the “urban form-SUHII-LCZ” linkage and provides specific planning recommendations for urban heat mitigation, especially in subtropical high-density cities.

The high intensity of urban development in recent years has replaced many natural surfaces with artificial objects. Buildings and roads have increased heat storage, which has become one of the leading causes of urban hyperthermia. In the past, studies on urban form and the thermal environment mainly used surface coverage analysis. However, surface classification cannot effectively consider three-dimensional urban information such as building height and density. Therefore, this study will use Local Climate Zone (LCZ) as a classification method to investigate urban form and temperature. LCZ is a classification method that explores the association between urban form and temperature. Since the classification process of LCZ is a supervised classification using satellite images, the results are subjectively influenced. Therefore, this project selects several representative zones, combines the LCZ classification with the types defined by each category, and conducts a questionnaire survey on 50 students with architecture and landscape planning backgrounds to compile the classification results of most people in each representative zone. The survey results were used for training, and the types of LCZs in the study site were summarized. The results showed that the accuracy of the objective classification results was 5.5% higher than that of the subjective classification. Increasing the number of training areas from 3 to 8 for each type also improved the accuracy by about 4.5%. In addition, this study analyzed the LCZ classification of buildings and landscapes in each region and the influence of temperature through the temperature data of the Meteorological Bureau in 2022. It was found that the average LCZ temperature in built-up areas could reach 32°C, while the average LCZ temperature in non-built-up areas was about 30.5°C in summer daytime. This study will serve as a reference for urban design to reduce the temperature for the potential of sustainable urban development.

Investigate the urban growth and urban-rural gradients based on local climate zones (1999–2019) in the Greater Bay Area, China

An investigation into heat storage by adopting local climate zones and nocturnal-diurnal urban heat island differences in the Tokyo Prefecture

The current research on urban heat island (UHI) effect mostly focuses on the analysis of land use type changes and the surface UHI intensity. Few studies on the urban canopy heat island effect at a block scale of local climate zone (LCZ), although the canopy heat island effect is a key factor affecting human thermal comfort. Therefore, this study will combine the LCZ classification system and the urban weather generator (UWG) model to simulate and quantitatively analyse the urban canopy heat island effect in Beijing at the block scale. First, based on Sentinel-2 Multispectral remote sensing images, the residual neural network (ResNet) method was used to obtain the LCZ of Beijing, and the results of LCZs was validated based on the google earth engine (GEE). Then, according to the classification results of local climate zones, the input parameters of the UWG and their corresponding value ranges are calculated. Finally, the UWG model is used to simulate the canopy temperature in different local climate zones, and the urban canopy temperature is validated using the meteorological station dataset. We quantitatively analyse the temperature differences between different types of LCZs. The results shows that the canopy heat island effect in Beijing gradually weakened outward from the city centre. This is mainly due to the relatively dense distribution of compact local climate zones in the centre of Beijing, while the surrounding areas of Beijing have lower building density and better natural coverage. The canopy urban heat island intensity of built-up LCZs is significantly stronger than that of natural-covered LCZs. The heat island intensity of the compact LCZ is higher than that of the open LCZ with the same building height. However, for LCZs with comparable compactness, the heat island intensity of high-level LCZs is higher than that of low-level LCZs.

Cities are frontlines to tackle climate change challenges including the urban heat island (UHI) effect. The classification and mapping of local climate zones (LCZs) can effectively and consistently describe the urban surface structure across urban regions. This study pays attention to two mainstream methods in classifying LCZs, namely, by using geographic information system (GIS) data such as building footprints or remote sensing (RS) satellite images. Little has been done to compare the divergence and coherence of the abovementioned two methods in modeling UHI. Thus, by comparing pairwise LCZ classes of different urban form characteristics in Guangzhou, this study investigated how GIS- and RS-based approaches complement or conflict with each other in explaining the variance of UHI measured by land surface temperature (LST). First, while both GIS-based (R2 0.724) and RS-based (R2 0.729) approaches can effectively explain heat risks measured by LST, the RS-based method slightly outperforms the GIS counterpart. Second, the sizes of LCZs classified by two methods in urban core districts tend to converge but diverge in urban outskirts with disparities in low-rise urban forms. Both approaches found that LCZs with higher heights are all cooler among compact forms. LCZ E is always related to the highest average LST, and LCZ 7, 8, and 10 contribute significantly to heat islands from both GIS and RS results. This study has developed a comparable framework that is evident based for city planners, architects, and urban policy makers to evaluate which approaches can more accurately reveal relations between UHI and urban geometry with land cover.

Spatially-heterogeneous impacts of surface characteristics on urban thermal environment, a case of the Guangdong-Hong Kong-Macau Greater Bay Area

The Local Climate Zone (LCZ) classification system is used in this study to analyze the impacts of urban morphology on a surface urban heat island (SUHI). Our study involved a comparative analysis of SUHI effects in two Japanese cities, Sapporo and Hiroshima, between 2000 to 2022. We used geographical-information-system (GIS) mapping techniques to measure temporal LST changes using Landsat 7 and 8 images during the summer’s hottest month (August) and classified the study area into LCZ classes using The World Urban Database and Access Portal Tools (WUDAPT) method with Google Earth Pro. The urban thermal field variance index (UTFVI) is used to examine each LCZ’s thermal comfort level, and the SUHI heat spots (HS) in each LCZ classes are identified. The research findings indicate that the mean LST in Sapporo only experienced a 0.5 °C increase over the time, while the mean LST increased by 1.8 °C in Hiroshima City between 2000 and 2022. In 2000, open low-rise (LCZ 6) areas in Sapporo were the hottest, but by 2022, heavy industry (LCZ 10) became the hottest. In Hiroshima, compact mid-rise (LCZ 2) areas were the hottest in 2000, but by 2022, heavy-industry areas took the lead. The study found that LCZ 10, LCZ 8, LCZ E, and LCZ 3 areas in both Dfa and Cfa climate classifications had unfavorable UTFVI conditions. This was attributed to factors such as a high concentration of heat-absorbing materials, impervious surfaces, and limited green spaces. The majority of the SUHI HS and areas with the highest surface temperatures were situated near industrial zones and large low-rise urban forms in both cities. The study offers valuable insights into the potential long-term effects of various urban forms on the SUHI phenomenon.

To better investigate the Urban Heat Island (UHI) effect, a standardized framework known as Local Climate Zones (LCZ) has been developed and widely applied to numerous cities. However, cities from least-developed countries with heterogeneous typologies are underrepresented in the LCZ literature. This study assesses the applicability of the LCZ framework in the slum-dominant built-up environment of Kabul. Using a combined method involving GIS and remote sensing, we classified natural and built-type LCZs and analyzed LCZ-LST fluctuations. The analysis revealed that four new subclasses cover 23 % of the built type LCZs: LCZ 35 (mid/high-rise buildings within compact lowrise layouts) and LCZ 65 (midrise buildings among open lowrise areas) have the lowest LSTs at 34.36 °C and 34.42 °C in July, respectively. In contrast, LCZ 73 (two/three-story buildings in lightweight configurations), and LCZ 9F (sparse buildings on bare soil or sand) have higher LSTs at 37.2 °C and 38.6 °C in July, respectively. These subclasses showed distinct zone parameter thresholds compared to standard LCZs. In most built-type LCZs, Average Building Height (ABH) and Pervious Surface Fraction (PSF) negatively influenced LST, while impervious surfaces and Sky View Factor contributed to higher LST. Based on the findings, LCZ-specified strategies (Vegetation, urban form, and using high-albedo materials) for LST mitigation are proposed. Furthermore, we provide planning, design, and policy recommendations aimed at mitigating urban heat, with potential applicability to other cities facing rapid urbanization and growth of informal settlements. The findings can inform action toward urban climate change adaptation.

Comparison between three convolutional neural networks for local climate zone classification using Google Earth Images: A case study of the Fujian Delta in China

Urban heat islands are representative problems in urban environments. The impact of spectral indexes on land-surface temperature (LST) under different urban forms, climates, and functions is not fully understood. Local climate zones (LCZs) are used to characterize heterogeneous cities. In this study, we quantified the contribution of three cities to high-temperature zones and surface urban heat island intensity (SUHII) across LCZs and seasons, used Welch and Games–Howell tests to analyze the difference in LST, then described the spatial pattern characteristics of LST, and used a geographically weighted regression model to analyze the relationship between spectral indexes and LST. The results showed that compact midrise, compact low-rise (LCZ 3), large low-rise (LCZ 8), heavy industry (LCZ 10), and bare rock or paved (LCZ E) contributed greatly to high-temperature zones and had strong SUHII. There were 92–98% significant differences between different LCZs. The spatial aggregation of LST gradually weakened with a decrease in temperature. The modified normalized difference water index (MNDWI) in most LCZs of all seasons for Wuhan could reduce LST well, while MNDWI only had cooling effects in winter for Nanjing and Shanghai. Normalized difference vegetation index (NDVI) in most LCZs performed a cooling role during summer and transition seasons (spring and autumn), while it showed a warming effect in winter. The cooling effect of NDVI in open building types was stronger than that of compact building types, while the cooling effect of MNDWI was better in compact building types than in open building types. With the increase of normalized difference built-up index (NDBI), all LCZs showed warming effects, and the magnitude of LST increase varied in different cities and seasons. These results contribute further insight into thermal environment in heterogeneous urban areas.

Studying air Urban Heat Islands (AUHI) in African cities is limited by building height data scarcity and sparse air temperature (Tair) networks, leading to classification confusion and gaps in Tair data. Satellite imagery used in surface UHI (SUHI) applications overcomes the gaps which befall AUHI, thus making it the primary focus of UHI studies in areas with limited Tair stations. Consequently, we used Landsat 30 m imagery to analyse SUHI patterns using Land Surface Temperature (LST) data. Local climate zones (LCZ) as a UHI study tool have been documented to not result in distinct thermal environments at the surface level per LCZ class. The goal in this study was thus to explore relationships between LCZs and LST patterns, aiming to create a building height (BH)-independent LCZ framework capable of creating distinct thermal environments to study SUHI in African cities where LiDAR data are scarce. Random forests (RF) classified LCZ in R, and the Single Channel Algorithm (SCA) extracted LST via the Google Earth Engine. Statistical analyses, including ANOVA and Tukey’s HSD, assessed thermal distinctiveness, using a 95% confidence interval and 1 °C threshold for practical significance. Semi-Automated Agglomerative Clustering (SAAC) and Automated Divisive Clustering (ADC) grouped LCZs into thermally distinct clusters based on physical characteristics and LST data internal patterns. Built LCZs (1–9) had higher mean LSTs; LCZ 8 reached 37.6 °C in Spring, with a smaller interquartile range (IQR) (34–36 °C) and standard deviation (SD) (1.85 °C), compared to natural classes (A–G) with LCZ 11 (A–B) at 14.9 °C/LST, 17–25 °C/IQR, and 4.2 °C SD. Compact LCZs (2, 3) and open LCZs (5, 6), as well as similar LCZs in composition and density, did not show distinct thermal environments even with building height included. The SAAC and ADC clustered the 14 LCZs into six thermally distinct clusters, with the smallest LST difference being 1.19 °C, above the 1 °C threshold. This clustering approach provides an optimal LCZ framework for SUHI studies, transferable to different urban areas without relying on BH, making it more suitable than the full LCZ typology, particularly for the African context. This clustered framework ensures a thermal distinction between clusters large enough to have practical significance, which is more useful in urban planning than statistical significance.

Abstract. Geographical features may have a considerable effect on local climate. The local climate zone (LCZ) system proposed by Stewart and Oke (2012) is nowadays seen as a standard approach for classifying any zone according to a set of urban canopy parameters. While many methods already exist to map the LCZ, only few tools are openly and freely available. This paper presents the algorithm implemented in the GeoClimate software to identify the LCZ of any place in the world based on vector data. Six types of information are needed as input: the building footprint, road and rail networks, water, vegetation, and impervious surfaces. First, the territory is partitioned into reference spatial units (RSUs) using the road and rail network, as well as the boundaries of large vegetation and water patches. Then 14 urban canopy parameters are calculated for each RSU. Their values are used to classify each unit to a given LCZ type according to a set of rules. GeoClimate can automatically prepare the inputs and calculate the LCZ for two datasets, namely OpenStreetMap (OSM, available worldwide) and the BD TOPO® v2.2 (BDT, a French dataset produced by the national mapping agency). The LCZ are calculated for 22 French communes using these two datasets in order to evaluate the effect of the dataset on the results. About 55 % of all areas have obtained the same LCZ type, with large differences when differentiating this result by city (from 30 % to 82 %). The agreement is good for large patches of forest and water, as well as for compact mid-rise and open low-rise LCZ types. It is lower for open mid-rise and open high-rise, mainly due to the height underestimation of OSM buildings located in open areas. Through its simplicity of use, GeoClimate has great potential for new collaboration in the LCZ field. The software (and its source code) used to produce the LCZ data is freely available at https://doi.org/10.5281/zenodo.6372337 (Bocher et al., 2022); the scripts and data used for the purpose of this article can be freely accessed at https://doi.org/10.5281/zenodo.7687911 (Bernard et al., 2023) and are based on the R package available at https://doi.org/10.5281/zenodo.7646866 (Gousseff, 2023).

Exploring the relationship between urban form, land surface temperature and vegetation indices in a subtropical megacity

The importance of easy wayfinding in complex urban settings has been recognized in spatial planning. Empirical measurement and explicit representation of wayfinding, however, have been limited in deciding spatial configurations. Our study proposed and tested an approach to improving wayfinding by incorporating spatial analysis of urban forms in the Guangdong-Hong Kong-Macau Great Bay Area in China. Wayfinding was measured by an indicator of intelligibility using spatial design network analysis. Urban spatial configurations were quantified using landscape metrics to describe the spatial layouts of local climate zones (LCZs) as standardized urban forms. The statistical analysis demonstrated the significant associations between urban spatial configurations and wayfinding. These findings suggested, to improve wayfinding, 1) dispersing LCZ 1 (compact high-rise) and LCZ 2 (compact mid-rise) and 2) agglomerating LCZ 3 (compact low-rise), LCZ 5 (open mid-rise), LCZ 6 (open low-rise), and LCZ 9 (sparsely built). To our knowledge, this study is the first to incorporate the LCZ classification system into the wayfinding field, clearly providing empirically-supported solutions for dispersing and agglomerating spatial configurations. Our findings also provide insights for human-centered spatial planning by spatial co-development at local, urban, and regional levels.