Development of an integrated index to quantify thermal comfort and walkability in urban areas

. Although outdoor thermal comfort is extensively investigated in urban areas, the measures are barely focused to determine the walkability through these spaces. Therefore, a space with a high level of thermal comfort can experience a low level of pedestrian agglomeration while a space with a low level of thermal comfort can be massively used by inhabitants. Therefore, the solution to urban design and planning can be significantly altered if both dimensions are simultaneously taken into the account. This study investigates the relationship between spatial configuration and thermal comfort potential to evaluate the effect of spatial configuration on outdoor environmental quality. For this purpose, a framework is developed to understand the impact of built urban areas on thermal comfort and space syntax performance using a high-resolution spatial model to simulate the correlation of thermal comfort and betweenness centrality of a case study neighbourhood in the hot and humid climate area of Al-Khobar in Saudi Arabia. The mixed-use neighbourhood is analysed by the universal thermal comfort index and several space syntax metrics. The presented study uses Grasshopper environment and Ladybug and Decoding Spaces tools. The simulation study expressed that significant changes in orientation and buildings heights have a remarkable effect on improving OTC and space syntax in the urban neighbourhood.


Outdoor thermal comfort
Comfortable thermal conditions in outdoor spaces can enhance urban life and activities, while heat stress can disrupt them. Higher temperatures in urban areas lead to reduced outdoor thermal comfort, increased concentration of pollutants, and higher energy demand for buildings, resulting in health problems and changes to outdoor activities. Heat stress can cause heat-related illnesses and even death, particularly in hot climates (Aghamolaei et al., 2020;Mirzaei, 2015). The significance of pedestrian health and thermal comfort is increasingly being recognized in the design and planning of metropolitan areas. Inappropriate urban design can worsen pedestrian outdoor comfort (Mirzaei, 2010) The measure for influencing how individuals perceive their surroundings is thermal comfort, which is quantified by air temperature, radiant temperature, air velocity, and humidity. In urban planning, the quantitative assessment of thermal comfort and walkability are considered because they affect the physical and mental health of dwellers . Urban open spaces can improve the quality of living while walkability advantages are related to spatial quality (Gehl, J., 2010). A pedestrian neighbourhood can encourage healthy walking habits and create a lively vibrant street life (Saelens et al., 2008;Speck, 2013). Outdoor thermal discomfort is a major significant problem that affects pedestrian in urban canyons and open spaces, affecting pedestrians in urban canyons and open spaces surrounded by built buildings. * Corresponding author: Mona.Alnimer@nottingham.ac.uk Collecting pedestrian data for every urban area and street network is challenging. Therefore, pedestrian volume models are used for estimating pedestrian movement across a network (Lerman, Rofe, et al., 2014). Nonetheless, limited studies are reported on the interconnection of people's behaviour and their chosen pathways in cities (Bontje, M., & Hulsbergen, E., 2014). Other researchers have explored the complexity of physical, physiological, or psychological processes triggered to adapt to outdoor conditions (Baker et al., 2001). The complete index used to describe outdoor comfort is the Universal Thermoclimate Index (UTCI), which assesses the outdoor thermal environment based on dynamic physiological responses (Brode, Fiala et al., 2012). Encouraging walking as an integral aspect of creating healthy cities is essential. Being a crucial means of transportation, walking provides significant advantages for individuals and the community (He and He 2023). While outdoor thermal comfort is extensively studied in urban areas, the extent to which a street, either with a high or low level of comfort, is used by the pedestrian has barely been considered. This implies that there is a lack of studies on the interconnection of people's behaviour and their chosen pathways in cities. The aim of this study is to propose a measurement index to assess the simultaneous effect of thermal conditions and centrality. This index, hence, quantifies the design outputs at various stages to support the developing of the design solutions and reducing the risk of failure during the design process. The presented study describes a novel simulation workflow that is developed which connects thermal comfort with walkability using Grasshopper. The developed model measures the relationship of the thermal comfort and space syntax. The Ladybug and Decoding Spaces tools are used to model the

Metrics of closeness and betweenness centrality in Space Syntax
Spaces are voids between walls, fences, and other obstructions that restrict pedestrian circulation or the perceived environment, which include squares, fields, streets, rooms, etc. (Klarqvist and Jiang, 1993). Buildings are made of a sequence of interconnected spaces, each linked to at least one other area (Al-Sayed 2014). The structural patterns that define and connect these areas may have an underlying social significance that influences the general behaviour of the human environment.
A significant measure in space syntax theory is angular segment choice or angular shortest path betweenness, which measures the similarity in explaining aggregate pedestrian and vehicular movement (Pennet et al., 1998;Hillier and Iida, 2005). A vital space syntax measure is the choice (or betweenness centrality), representing the extent to which a street functions as an intermediate location within the network. This choice is considered a through movement potential (Omer and Kaplan, 2017) because it counts the number of times a street segment lies between all the origin-destination pairs in the network. Several studies have shown that the integration measure (closeness) and the choice measure (betweenness) can predict human movement (Hillier & Hanson, 1984;Kim & Piao, 2017;Turner, 2002). The Space Syntax analysis results utilize the metric of "Betweenness Centrality" to display the significance of each pathway segment within the connectivity network. Pathway segments with high values of betweenness centrality may indicate key streets and public spaces and are representative of areas that require attention for investment or improvement. These high betweenness centrality pathways may also be prone to congestion as they provide the shortest path to attractive destinations. Fig 1 presents an initial Space Syntax result, highlighting pathways in blue for low betweenness centrality and in red for high betweenness centrality. An axial map's objective is to record the most significant number of connections within the spatial layout. The street network geometry is constructed before preparing the special graph, defining distance, and calculating betweenness centrality (BA) for each street segment. The red line shows the shortest distance a person would take from one-line segment to the next, showing how congested the street with the highest betweenness centrality would be, and provides the analysis of the pedestrian movement from two points of interest. The analysis indicates that there is a single path with high betweenness centrality, spanning the entire network, when considering an infinite shortest path radius. This path would be the most efficient route to visit all the desired destinations rapidly.

Universal thermal climate index
The Fiala model, which depicts the complexity of the physiological behaviour of the human body, particularly its thermal response to changing environmental circumstances, is the foundation for UTCI computations (Höppe, 1999). Table 1 shows the UTCI range for different stress categories as being used in this study.

Proposed modelling framework
The step-by-step process adopted to develop the framework is explained in Fig. 2. The first phase is the assessment of the effect of spatial configuration on outdoor

Fig. 2.
Work Framework environmental quality. A 3D geometrical model of a lowrise neighbourhood that considers building typology, height, and orientation is utilized as the cases study to demonstrate the performance of the proposed framework. It is combined with UTCI to read current and future climatic information from EnergyPlus weather (EPW) data. The study considers the number of solar hours, radiation, shadowing neighbourhood, and rose wind to anticipate the perspective of OTC. The Ladybug plugin is then used as it connects EnergyPlus weather files into Grasshopper and performs an environmental study of urban 3D models. These plugins' simulation tools use different inputs, such as geometry, materials, zones, etc., which can be further modified within Grasshopper. Ladybug can be interfaced with EPW files, and a scope of visualizations where basic calculations can be performed based on weather data. The second phase involves the use of space syntax analysis to present the betweenness centrality metric by using the Decoding Space plugin, which depicts the outcomes for each pathway segment. The betweenness centrality metric measures the importance of a pathway in the connectivity network, and high values of this metric for pathway segments may indicate crucial streets and public spaces. These areas should receive investment or improvement attention, as they also represent potential congestion points. Also, space syntax analysis examines the spatial interrelations within a space and analyses human behaviour, and pedestrian movement. The third phase correlates the two results that are found in high-resolution spatial models by coupling a thermal comfort plugin (Ladybug) and a space syntax plugin (Decoding Space). The process explains that the disparities in comfort performances are associated with the impacts of people's movement. The results found by high-resolution spatial models using the thermal comfort plugin (Ladybug) and space syntax plugin (Decoding Spaces). The faces of each grid and the vertices of each face are created and developed after taking the geometry grids from the thermal analysis and breaking it down into their parts. A numerical model is developed to examine the variations in comfort performance in response to the concurrent impacts of people's mobility.

Interlink calculation.
To compare thermal performance with the base case configuration, parametric and nonparametric correlation coefficients are calculated for the surfaces. (1) where X represents the data of syntax and Y is the data of comfort. ‫,ܺ(ݒܿ‬ ܻ) is the covariance between these values. ߪܺ is the standard deviation of X and ߪܻ is the standard deviation of where ‫,)ܺ(ܴ(ݒܿ‬ ܴ(ܻ)) is the covariance between the ordered (ranked) values of X, ܴ(ܺ), and the ordered (ranked) values of Y, ܴ(ܻ). ߪܴ(ܺ) is the standard deviation of ranked values of X and ߪܴ(ܻ) is the standard deviation of the ranked values of Y.
It is necessary to acquire both values of UTCI and betweenness centrality for specific locations to compare the measures. UTCI is an equivalent temperature (°C), and in space syntax analysis, betweenness centrality units are often represented as a value between 0 and 1. Betweenness centrality measures the number of times a node or link falls on the shortest path between other nodes in the network. As a result, a grid of cells with corresponding thermal comfort values is created while the centrality values are obtained by measuring and finding the shortest paths for all possible origin-destination pairs. In consequence, each segment of the street network has associated centrality values. To couple the values mentioned above, street segments are determined that pass through the cells, and the mean value is calculated. Thus, for each cell, a pair is associated with thermal comfort and centrality values.

Case study
Ajdan Walk is selected as the case study, which is a mixed-use neighbourhood in diverse spaces in Al-Khobar, Saudi Arabia with a total area of 17,000 square meters. Ajdan lies in the hot and humid climate area of AlKhobar in Saudi Arabia. Fig. 4 shows that Ajdan Walk has a strong pedestrian connection to other development parcels; the framework developed in this study involves developing a 3D geometric model of the neighbourhood considering building typology around Ajdan Walk, height, and orientation of principal facades.This section involves the integration of the UTCI results with the Space Syntax findings for several parameters analysed. Before combining the analysis results, all values are remapped to a range between 40°C and 45°C. The UTCI results correspond to 9am in August. The combined results are only presented for UTCI grids that intersect with any of the Space Syntax pathways, and all values within a UTCI grid are averaged to provide a single value. There are a total of 1,232 cells in the generated grid with the size of 10m as it can be seen in Table 2. The street segment lines are used and proj ected to the same cells. A total of 1,200 grid intersections are obtained with 67 lines. The outcome is an example of the combined results, which only consider the grid cells that intersect with pathways.  Weather data is imported from Ladybug's EPW map to collect climate information and investigate sunlight hours, amount of insolation, nearby shade, and wind speed. In the Fig. 5, UTCI of the case study is generated. The study is focused on a single calendar year, from January 1st to E3S Web of Conferences 396, 05005 (2023) https://doi.org/10.1051/e3sconf/202339605005 IAQVEC2023 December 31st, from 12:00 am to 2:00 pm. This is the worst-case scenario for an examination of pedestrian discomfort brought on by sweltering urban air temperatures during the month, given that the number of sunshine hours only pertains to the time when the sun is visible. Note that the case study does not include any mountains, clouds, or fog to block the view. In Saudi Arabia, the Eastern Province area has eight hours of sunlight per day in October.

Fig. 5. UTCI data maps in Ajdan walk from January 1st to
December 31st, from 1:00 am to 24:00 pm) Fig. 6 shows the tourism score, temperature variation, cloud cover, and precipitation variation. The filled area of the chart represents the overall tourism score, which is made up of three separate scores, including the temperature score represented by the red line, the cloud cover score represented by the blue line, and the precipitation score represented by the green line (weatherspark, 2023). The graph demonstrates that the optimum travel seasons are from the end of January to the beginning of April and from the middle of November to the end of December. The main focus of this study is to examine UTCI and space syntax in relation to a specific area in Al Khobar, Saudi Arabia. The investigation involves analysing how the UTCI and space syntax metrics behave when there are changes to building heights and orientation parameters. The study compares and evaluates the individual and combined effects of these metrics.    https://doi.org/10.1051/e3sconf/202339605005 IAQVEC2023 Fig. 9 illustrates the wind rose of the area in the year while the prevailing wind is mostly from North and Northwest. Fig. 10 demonstrates the results of the measuring index that captures the maximum thermal comfort and movement flow inside the spatial configuration, representing the pathway segment that has a high value for this metric; it indicates that the street or public space is a crucial area that should be prioritized for investment or improvement, as it could potentially become a congestion point. Furthermore, the analysis involves exploring the spatial relationships within a space and examining human behaviour and pedestrian movement. Fig. 11 shows a scatterplot of thermal comfort against space syntax. The Pearson's correlation for these variables is equal to 0.66, and the result of Spearman's correlation is 0.67 (p-value < 0.01). This result indicates that positive values of space syntax are related to positive values of comfort. These findings indicate that a positive (direct) association exists between syntax and thermal comfort and that the null hypothesis of no association between comfort and space syntax can be rejected with a significance level of less than 1%.   z Urban geometry and buildings' morphology affect pedestrian thermal comfort. Increasing the shading effects by changing the block's orientation and buildings' height diversity. Comparing the coupled UTCI and BW centrality metrics in the same scaled-size neighbourhood model with increasing building height shows a significant change in UTCI and betweenness centrality. As seen in Fig. 12, an urban configuration with significant variation in the buildings' height results in more reduction in outdoor air temperature compared to a configuration with low-rise buildings' heights. Orienting the significant variation in buildings' heights along the short axis of the urban block results in a reduction of the air temperature. The findings show that the preferred pathways highlighted as red for the greatest thermal comfort and mobility flowing inside the spatial configures. Fig 13 is a 3d outcome of the coupled model exploring the spatial relationships within a space and examining human behaviour and pedestrian movement.  Fig. 13 shows a scatterplot of average syntax and comfort and average syntax and comfort after changing height and orientation. These two values have a Pearson correlation of 0.12 with a Spearman correlation equal to 0.13 and a pvalue of 0.0174, statistically significant at the 5%, but not the 1% level. Hence, it can be concluded that a positive correlation exists between the average remapped syntax and comfort and the average syntax and comfort after changing height and orientation.

Conclusion
UTCI and the betweenness centrality are metrics that calculate outdoor comfort and space, respectively. An analysis was conducted to examine how building orientation and height affect UTCI and space syntax. Two experiments were carried out, one involved the rotation of all buildings in the region by the same degree, and the other involved a uniform increase in the height of all buildings. The result is a grid of cells connected with thermal comfort metric values. On the other hand, centrality values are determined by calculating the shortest routes for all possible origin-destination pairings in the current situation. The result is a geometry of the street network that has associated centrality values with each segment. A highresolution spatial model is developed to determine street segments that travel through each cell; the results are combined, and a mean value is obtained. For each cell, a pair of thermal comfort and centrality values are linked with each cell. The estimated values are then used to identify the link between spatial arrangement and thermal comfort. Combined UTCI and space syntax analysis results are investigated for varying building height and orientation. A logical result is produced that taller buildings have the potential to create more shade, which can result in a decrease in the perceived temperature. The extent of shading at different times of the day is heavily influenced by the orientation of the buildings. Consequently, the shading provided by buildings and their spatial configuration can significantly impact the flow of mobility within an area. Observations show that there is a positive correlation between OTC and space syntax after varying building heights and orientation.

Future research
The concept of points of interest and working hours may be included as a future extension of the proposed approach to give added details to the analysis. For example, the commute times can be considered for a school to determine the shortest route and to travel through pleasant, shaded areas. Since the shading analysis is limited to buildings, for more accurate findings, trees, shading equipment, and other urban furniture may also be considered during the modelling.