Local Climate Zone Mapping Research Articles

Mapping Irregular Local Climate Zones from Sentinel-2 Images Using Deep Learning with Sequential Virtual Scenes

Recently, the local climate zone (LCZ) system has been presented to establish the connection between urban landscape and local thermal environment. However, LCZ entities are very difficult to be identified by pixel-based classifiers or object-oriented image analysis, as they are often a complicated combination of multiple ground objects (e.g., buildings, roads, grassland, etc.). Scene classifiers, especially deep learning methods can exploit the structure or contextual information of image scenes and then improve the performance of LCZ classification. However, the square and uniform-sized image patches often bring about extra challenges, as they cannot exactly match LCZ entities of diverse sizes and shapes in most cases. In this study, a sequential virtual scene method is presented to identify LCZ entities of diverse shapes and sizes, which consists of a small “core patch” for scanning diverse entities and sequential virtual scenes for providing abundant context. Specifically, the Bidirectional Long Short-Term Memory (Bi-LSTM) were used to learn the spatial relationship among virtual scenes, respectively. Importantly, a “self-attention” mechanism is designed to weigh the contribution of every virtual scene for alleviating the influences of mixed patches, according to the similarity between its hidden state and the final hidden state. Experiments prove SVS achieves better accuracies than random forest and ResNet and has the outstanding capacity of identifying irregular LCZ entities. It is a promising way to carry out LCZ mapping in cities of different types due to its flexibility and adaptability.

A study of Urban Heat Island relating "Local Climate Zones" using Landsat Images- The Case of Kathmandu Valley

Local Climate Zone (LCZ) classification has been intensively used in classification of urban and rural landscapes in the cities, which includes urban temperature studies. The urban heat island (UHI) in Kathmandu valley (Kathmandu, Bhaktapur, and Lalitpur), has been analyzed and standardized, which mainly focused on the Local Climate Zones (LCZs). The LCZs has distribute the landscape into homogeneous types on the basis of structural type, land surface cover, materials used, and into the anthropogenic activities. Such standardized classification has improved the meaning of urban research and made it easier to compare results among cities around the world. Landsat images, Google Earth, and SAGA-GIS software were used for creating LCZ map for Kathmandu for both March 2013 and March 2019 Landsat 8 TM/ETM+/OLI imagery was used to estimate land surface temperature (LST) For the estimation of LST world urban database and access portal tools (WUDAPT) algorithm was used considering emissivity. The result thus shows that the difference within the built-up scheme is around 2-4 °C whereas the difference between Building and Land cover types on the comparison is around 5-10°C. The disparity in building the land cover types shows that the UHI impact is present in the Kathmandu valley.

A Literature Survey of Local Climate Zone Classification: Status, Application, and Prospect

Rapid urban expansion and climate change have prompted further investigations into urban thermal climates and the development of local climate zone (LCZ) classification systems. LCZs, proposed 10 years ago, comprise a new and systematic classification of field sites for heat island studies to provide a reference for future LCZ research, so that scholars can understand what research has been done and identify future research trends. We analyzed LCZ studies in a database from 2012 to the present, and identified recurring themes using VOSviewer software, including LCZ mapping, measurement methods, thermal environments, and outdoor thermal comfort, among others. A systematic evaluation was performed using bibliometric analysis in the PRISMA framework—190 relevant studies were selected for subsequent analysis. Descriptive analysis showed that LCZ research has received increasing attention, particularly in China, where more than 60% of the LCZ studies were conducted. The results showed that the maximum number of articles on all themes was 57 articles on LCZ mapping, followed by studies of the thermal environment (UHI/SLT). It is hoped that this article will provide scholars in this area with an understanding of the research that has been conducted and the methods used, and provide insight into future research directions.

The characteristic and transformation of 3D urban morphology in three Chinese mega-cities

Urban morphology exerts an important influence on human settlements and serves to promote sustainable development. However, few studies have focused on the three dimensions (3D) of urban morphology. Therefore, to explore 3D urban morphological characteristics and transformations, we adopted a standard urban morphology classification scheme, the local climate zone (LCZ), and selected three Chinese megacities as case studies. Remote sensing images and machine-learning technology were used to create spatiotemporal LCZ maps. Our results reveal that cities have diverse forms, with the dominant urban morphology varying across different cities and periods. However, cities still share several common patterns. The dynamic spatial distribution of LCZ density can be divided into three patterns: spreading, shrinking and relocation. In recent years, Chinese megacities have mainly expanded with open high-rise (OHR) and large low-rise (LLR) classes. Combined with the increase in building height, urban landscapes have been homogeneous. Additionally, 3D urban morphology is driven by several factors, such as economic activity and the dissociation of population growth from urban expansion. Detailed spatiotemporal LCZ maps can help us better understand urban morphology transformations and explore the corresponding driving forces, providing a reference for urban planners and policymakers seeking to make cities healthier and more livable.

Modeling intra-urban differences in thermal environments and heat stress based on local climate zones in central Wuhan

Urban morphology exerts a strong influence on urban thermal environments, and local climate zones (LCZs) explicitly link urban morphology with microclimate. However, how temperature, humidity, and thermal comfort vary across different LCZs from a modeling perspective remains understudied. Here, we first developed a high-precision LCZs map of central Wuhan using a combination of GIS- and RS methods, and then simulated intra-urban differences in 2-m air temperature (T2), water vapor mixing ration (Q2), and heat index (HI) through incorporation of the LCZs map into a mesoscale model coupled with an urban canopy model. Results showed that the urban core area was dominated by the compact forms (i.e., LCZs 1–3), surrounded by the open forms (i.e., LCZs 4–6) and the natural/semi-natural types on the periphery (i.e., LCZs D and G). The intra-urban variability of nighttime T2 was up to 2.5 °C and daytime Q2 up to 0.9 g/kg across LCZs of the artificial built types. The compact forms were hotter but drier than the corresponding open forms, and LCZ 9 had the lowest T2 but highest Q2. HI exceeded the danger line (i.e., 41 °C) during 12:00–21:00 LST across the LCZs, with the intra-urban variability of HI in the daytime smaller than in the nighttime. That is, HI was largely influenced by Q2 in the daytime but mainly determined by T2 in the nighttime. Findings of this study highlighted that mitigation measures should be taken to alleviate the pressure of high temperature and humidity on human thermal comfort.

Comparing Local Climate Zone mapping results of Tirana through different approaches

Comparing Local Climate Zone mapping results of Tirana through different approaches

The urban morphology classification under local climate zone scheme based on the improved method - A case study of Changsha, China

Generating local climate zone (LCZ) maps has been a new way for understanding urban morphology recently. So far, LCZ mapping has two major methods. WUDAPT workflow is widely used because of time-saving mapping process, while GIS stream is more complex and has obstacles such as lack of localized LCZ parameters, difficulty of parameter collection and unclear mapping specifications. Based on the original LCZ scheme, the study aims to fill the research gap by adopting GIS-based rules to identify urban morphology using multi-source dataset. Firstly, the study adjusts the LCZ types and LCZ parameters according to the current situation of Changsha. Secondly, a comprehensive set of LCZ parameters are calculated and evaluated, including Building Height (BH), Building Surface Fraction (BSF), Floor Area Ratio (FAR), Sky View Factor (SVF), Street Canyon Ratio (SCR), Impervious Surface Fraction (ISF), Pervious Surface Fraction (PSF), Tree Information (TH), Terrain Roughness (TR) and Albedo (AL). Thirdly, each 250 m × 250 m cell is classified according to the LCZ parameter threshold. The result indicates a higher OA (84.40%) of GIS stream comparing WUDAPT workflow. This study provides a clear depiction of improved GIS stream, which contributes to the development of LCZ mapping and also provides support for climate-sensitive urban development.

A global map of local climate zones to support earth system modelling and urban-scale environmental science

Abstract. There is a scientific consensus on the need for spatially detailed information on urban landscapes at a global scale. These data can support a range of environmental services, since cities are places of intense resource consumption and waste generation and of concentrated infrastructure and human settlement exposed to multiple hazards of natural and anthropogenic origin. In the face of climate change, urban data are also required to explore future urbanization pathways and urban design strategies in order to lock in long-term resilience and sustainability, protecting cities from future decisions that could undermine their adaptability and mitigation role. To serve this purpose, we present a 100 m-resolution global map of local climate zones (LCZs), a universal urban typology that can distinguish urban areas on a holistic basis, accounting for the typical combination of micro-scale land covers and associated physical properties. The global LCZ map, composed of 10 built and 7 natural land cover types, is generated by feeding an unprecedented number of labelled training areas and earth observation images into lightweight random forest models. Its quality is assessed using a bootstrap cross-validation alongside a thematic benchmark for 150 selected functional urban areas using independent global and open-source data on surface cover, surface imperviousness, building height, and anthropogenic heat. As each LCZ type is associated with generic numerical descriptions of key urban canopy parameters that regulate atmospheric responses to urbanization, the availability of this globally consistent and climate-relevant urban description is an important prerequisite for supporting model development and creating evidence-based climate-sensitive urban planning policies. This dataset can be downloaded from https://doi.org/10.5281/zenodo.6364594 (Demuzere et al., 2022a).

Mapping Local Climate Zones in the Urban Environment: The Optimal Combination of Data Source and Classifier.

The novel concept of local climate zones (LCZs) provides a consistent classification framework for studies of the urban thermal environment. However, the development of urban climate science is severely hampered by the lack of high-resolution data to map LCZs. Using Gaofen-6 and Sentinel-1/2 as data sources, this study designed four schemes using convolutional neural network (CNN) and random forest (RF) classifiers, respectively, to demonstrate the potential of high-resolution images in LCZ mapping and evaluate the optimal combination of different data sources and classifiers. The results showed that the combination of GF-6 and CNN (S3) was considered the best LCZ classification scheme for urban areas, with OA and kappa coefficients of 85.9% and 0.842, respectively. The accuracy of urban building categories is above 80%, and the F1 score for each category is the highest, except for LCZ1 and LCZ5, where there is a small amount of confusion. The Sentinel-1/2-based RF classifier (S2) was second only to S3 and superior to the combination of GF-6 and random forest (S1), with OA and kappa coefficients of 64.4% and 0.612, respectively. The Sentinel-1/2 and CNN (S4) combination has the worst classification result, with an OA of only 39.9%. The LCZ classification map based on S3 shows that the urban building categories in Xi’an are mainly distributed within the second ring, while heavy industrial buildings have started to appear in the third ring. The urban periphery is mainly vegetated and bare land. In conclusion, CNN has the best application effect in the LCZ mapping task of high-resolution remote sensing images. In contrast, the random forest algorithm has better robustness in the band-abundant Sentinel data.

Comparing Object-Based and Pixel-Based Methods for Local Climate Zones Mapping with Multi-Source Data

The local climate zones (LCZs) system, a standard framework characterizing urban form and environment, effectively promotes urban remote sensing research, especially urban heat island (UHI) research. However, whether mapping with objects is more advantageous than with pixels in LCZ mapping remains uncertain. This study aims to compare object-based and pixel-based LCZ mapping with multi-source data in detail. By comparing the object-based method with the pixel-based method in 50 and 100 m, respectively, we found that the object-based method performed better with overall accuracy (OA) higher at approximately 2% and 5%, respectively. In per-class analysis, the object-based method showed a clear advantage in the land cover types and competitive performance in built types while LCZ2, LCZ5, and LCZ6 performed better with the pixel-based method in 50 m. We further employed correlation-based feature selection (CFS) to evaluate feature importance in the object-based paradigm, finding that building height (BH), sky view factor (SVF), building surface fraction (BSF), permeable surface fraction (PSF), and land use exhibited high selection frequency while image bands were scarcely selected. In summary, we concluded that the object-based method is capable of LCZ mapping and performs better than the pixel-based method under the same training condition unless in under-segmentation cases.

Patch-Based Local Climate Zones Mapping and Population Distribution Pattern in Provincial Capital Cities of China

Accurate urban morphology provided by Local Climate Zones (LCZ), a universal surface classification scheme, offers opportunities for studies of urban heat risk, urban ventilation, and transport planning. In recent years, researchers have attempted to generate LCZ maps worldwide with the World Urban Database and Access Portal Tools (WUDAPT). However, the accuracy of LCZ mapping is not satisfactory and cannot fulfill the quality demands of practical usage. Here, we constructed a high-quality sample dataset from Chinese cities and presented a patch-based classification framework that employs chessboard segmentation and multi-seasonal images for LCZ mapping. Compared with the latest WUDAPT method, the overall accuracy for all LCZ types (OA) and urban LCZ types (OAu) of our framework increased by about 10% and 9%, respectively. Furthermore, based on the analysis of population distribution, we first gave the population density of different built-up LCZs of Chinese cities and found a hierarchical effect of population density among built-up LCZs in different size cities. In summary, this study could serve as a valuable reference for producing high-quality LCZ maps and understanding population distribution patterns in built-up LCZ types.

Spatial-temporal variations of surface urban heat island: An application of local climate zone into large Chinese cities

The land surface temperature (LST)—based surface urban heat island (SUHI) is suitable for exploring thermal environments at large scales. Despite a few studies focused on the intra-urban SUHI intensity (SUHII) of a single city or cities with similar climates by adopting the local climate zone (LCZ) scheme, exploration of the spatial-temporal characteristics of the intra-urban SUHII and SUHI extent (SUHIE) is still limited across diverse climate zones. This study aims to (1) propose an improved LCZ mapping method to refine historical LCZ data for highly urbanized Chinese cities; (2) put forward a buffer-based approach to quantify the actual area affected by the SUHI; (3) investigate the spatial-temporal pattern of SUHI in large Chinese cities. The results show that the intra-urban SUHII varies across seasons, times and background climate—highest during summer days and lowest during winter days, decreasing with background climate becoming warm and wet. SUHIE displays an apparent temporal pattern similar to SUHII and has an increasing trend, but it relates more to urban areas (particularly the area of LCZ 1) than background climate. Due to the positive impact of LCZ 1 on both SUHII and SUHIE, its area should be strictly restricted in design practice. Although LCZ 4 has the advantage over LCZ 1 in mitigating SUHII, it correlates positively to SUHI expansion. From a planning perspective, a balance between the reduction of SUHII and SUHIE should be considered. This study provides evidence-based planning recommendations to mitigate urban heat and create a comfortable built environment.

Improving the local climate zone classification with building height, imperviousness, and machine learning for urban models

The Local Climate Zone (LCZ) classification is already widely used in urban heat island and other climate studies. The current classification method does not incorporate crucial urban auxiliary GIS data on building height and imperviousness that could significantly improve urban-type LCZ classification utility as well as accuracy. This study utilized a hybrid GIS- and remote sensing imagery-based framework to systematically compare and evaluate different machine and deep learning methods. The Convolution Neural Network (CNN) classifier outperforms in terms of accuracy, but it requires multi-pixel input, which reduces the output’s spatial resolution and creates a tradeoff between accuracy and spatial resolution. The Random Forest (RF) classifier performs best among the single-pixel classifiers. This study also shows that incorporating building height dataset improves the accuracy of the high- and mid-rise classes in the RF classifiers, whereas an imperviousness dataset improves the low-rise classes. The single-pass forward permutation test reveals that both auxiliary datasets dominate the classification accuracy in the RF classifier, while near-infrared and thermal infrared are the dominating features in the CNN classifier. These findings show that the conventional LCZ classification framework used in the World Urban Database and Access Portal Tools (WUDAPT) can be improved by adopting building height and imperviousness information. This framework can be easily applied to different cities to generate LCZ maps for urban models.

Diurnally continuous dynamics of surface urban heat island intensities of local climate zones with spatiotemporally enhanced satellite-derived land surface temperatures

Interpretation of diurnally continuous dynamics of surface urban heat island intensities (SUHIIs) of various local climate zones (LCZs) is essential for deepening the understanding of the urban climate. However, an easily applicable method for quantifying diurnally continuous dynamics of LCZ-based SUHIIs is still lacking, and the understanding of the diurnally continuous patterns of LCZ-based SUHIIs remains largely incomplete. To alleviate these gaps, here we developed a simple yet effective method for producing diurnally continuous land surface temperatures (LSTs) of various LCZ types based on an LCZ map and MODIS LSTs. With the proposed method, we further investigated the diurnally continuous dynamics of LCZ-based SUHIIs in Nanjing, China. We obtained the following findings: First, in terms of the annual mean, the diurnal SUHII patterns among various built LCZ types were similar and had a small valley within 0–2 h after sunrise and a peak within 8–12 h after sunrise. Second, the midrise/high-rise built and heavy industry types (i.e., LCZ 4, 5, and 10) exhibited relatively high SUHIIs during a diurnal cycle and the associated daily duration with a significant SUHII (>0.5 K) was greater than 20 h. Third, the diurnal SUHII patterns in summer were analogous to the annual mean patterns for all built LCZ types, yet the patterns of several built LCZ types in winter contrasted significantly the annual mean ones. Our findings are valuable for deepening the understanding of the diurnally continuous dynamics of LCZ-based SUHIIs.

Geographical Detection of Urban Thermal Environment Based on the Local Climate Zones: A Case Study in Wuhan, China

The urban morphology has impacts on the urban thermal environment, which has drawn extensive attention, especially in metropolitan regions with intensive populations and high building densities. This study explored the relationship between the urban morphology and spatial variation of land surface temperature (LST) in Wuhan by using the local climate zone (LCZ) and seven natural and social factors. A deep learning model (light LCZ model) was used to generate LCZ map in Wuhan, and a geographic detector model was utilized to explore the driving mechanism of LST spatial differentiation. The results show that the LST difference between LCZ classes in summer is greater than that in winter, and the LST of the built-up classes (LCZ 1–10) are significantly higher than that of the vegetation classes in summer. Among the six residential building classes (i.e., LCZ 1–6), LCZ 1 is characterized by compact and high buildings and has the largest average LST. Building density and height have a warming effect, and the building density has a stronger effect than the height. Compared with other natural and social factors, LCZ has the largest explanatory power for LST spatial differentiation in the main urban area and surrounding areas of Wuhan, with explanatory power (q) values reaching 0.660 (summer) and 0.316 (winter). The types of interaction for all pairwise cases are mutual and nonlinear. The strongest interaction is MNDWI-NDBI combination (0.780) in summer and LCZ-NDBI combination (0.460) in winter.

Coupling Dual Graph Convolution Network and Residual Network for Local Climate Zone Mapping

Local climate zone (LCZ) has become a new standard classification scheme in urban landscapes and showed great potential in urban climate research. Traditional classifiers and ordinary neural networks only consider the spectral or local spatial features of the pixel, ignoring the effect of nonlocal information on the LCZ classification. The graph convolutional network (GCN) has been used to exploit the relationship between adjacent and global land covers owing to the ability to conduct flexible convolution over graphs. In this work, we integrated a convolutional neural network and two GCNs into an end-to-end hybrid framework and generated LCZs directly from the original images. Local-, regional-, and global-level features were extracted and grouped complementarily to foster better performance. Experiments were conducted in six cities around the world to verify the effectiveness of our method. Results showed that the average classification accuracy of the six cities reached 0.956 and performed better than any other comparable model. Ablation experiments also demonstrated the mutual promotion of the different modules. Finally, the small sample experiment provided a practical reference for the LCZ classification in the absence of samples in future.

A Vector-Based Mapping in GIS Environment to Classify Local Climate Zone

Local climate zones (LCZ), which are widely used in urban heat island studies, include climate-based classification of the city to determine the climatic differences in the metropolitan area. Therefore, mapping methods have been developed to visualise LCZ classes. Compared to the raster-based mapping method, the vector-based mapping method, which makes a more precise and accurate evaluation, is not widely used due to the difficulty in creating and obtaining a dataset. This study aims to implement a GIS-based approach in creating a vector-based LCZ map in the example of Adana City, Turkey. The method of the study consists of five steps: a) determination of classification criteria; b) mapping of each classification criteria via Arc-GIS; c) creation of the decision tree for the creation of the LCZ map; d) overlapping of the classification criteria in line with the decision tree; e) development of suggestions in line with the findings. Vector-based LCZ classification has minimised the classification mistakes that may arise from resolution and has enabled the creation of a highly accurate LCZ map. The LCZ map obtained from the study is expected to be a base map in studies on urban climate studies and to guide decision-makers.

Local climate zone mapping of Tirana, Albania

Local climate zone mapping of Tirana, Albania

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

City development experiences scale expansion from urban to rural regions and built-up intensification across urban domains. This study adopts the local climate zone (LCZ) scheme to investigate urban-rural urbanization gradients from a three-dimensional perspective. A time series of LCZ maps were developed over the last two decades, i.e., 1999, 2009, and 2019, using Earth observation data and machine-learning classification based on Google Earth Engine. Spatiotemporal changes of urbanization gradients and urban forms were captured by measuring the Euclidean distance of built-up LCZ types from compact high-rise patterns of 2019. Results showed that cities in the study region experienced pronounced spatial coverage expansion and compact and high-rise intensification between 1999 and 2019. This phenomenon was especially noticeable in zones close to compact high-rise patterns and periods from 1999 to 2009. In 2019, the GBA region showed central and sprawl urban forms; however, urban forms differed among the eleven cities. A city with more build-up LCZ areas or closer to the Pearl River estuary tended to present central and compact urban forms. In contrast, a city with less build-up LCZ areas or far from the estuary trended to show decentralized and sprawl urban forms. Our findings indicate that the urbanization-gradient derived from LCZ in mega-urban and urban areas is an effective indicator for characterizing urbanization development given the unique role of LCZ presented to differentiate urban expansion and intensification, as well as urban forms. LCZ scheme shows the potentiality in identifying urban central zones and measuring the urbanization gradient in mega-urban urban areas. Besides, the LCZ scheme provides a unique perspective to investigate the urban form of cities which exhibited polycentric structures.

Local climate zones mapping using object-based image analysis and validation of its effectiveness through urban surface temperature analysis in China

The concept of local climate zones (LCZ) received wide acceptance and it is now a global standard for urban structure classification. At present, remote sensing-based LCZ classification studies focus on the pixel level, and object-level-based investigations are scant. In the present study, an object-based remote sensing image analysis was utilized for LCZ mapping of three cities in the Yangtze River Delta including Shanghai, Nanjing, and Hangzhou. We also analyzed the spatial and temporal distributions of LCZ and established relationships between these zones and the surface temperatures in these megacities. According to the spatial and temporal patterns based on the spatial aggregation index, the LCZ in the three cities are mostly aggregated, and these are characterized by intense aggregation around low-rise buildings and weak aggregation near high-rise buildings. Analysis of seasonal characteristics of the surface temperatures of the LCZ types reveals that in the study area, the heat island effect is substantially higher during the summer than in the winter. Based on results from the single-factor and multiple comparison analyses, significant differences were confirmed between the surface temperatures of various object-based LCZs. Hence, we concluded that object-based LCZ classification is suitable for characterizing the urban heat island effect. These results also validate the applicability of the object-based image analysis (OBIA) in LCZ remote sensing mapping. These findings will advance the development and application of OBIA in LCZ mapping.