GIS-based approach evaluating sustainable spatio-functional accessibility to mosques

ABSTRACT The acceleration of urbanization leads to a geographical extension of cities, longer distances and travel times, increased motorization, and a constantly growing population. Faced with this situation, new urban policies are attempting to control the interaction between urban planning and urban mobility and to redefine a new dimension of proximity, which reduces excessive use of energy-consuming and polluting modes of travel, namely vehicles. This orientation is embodied in the 15-minute city model, which focuses on pedestrian accessibility from one’s home to nearby services and urban spaces. From this perspective, the paper proposes a GIS-based model to evaluate pedestrian accessibility to mosques. Using the open-source software QGIS, a total number of 30 mosques in the Agdal district in the city of Fez in Morocco were studied to investigate their pedestrian accessibility, by examining their spatial distribution and also their capacity to accommodate the worshippers of their catchment areas. Considering the land use and the population density of each mosque’s catchment area, the results showed that even if a mosque can be spatially accessible by walking, it may not be able to comfortably satisfy the potential worshippers in its area of influence. Many people are then forced to travel long distances, sometimes by vehicles, to reach other mosques that can eventually accommodate them. The proposed method helps city planners better understand the urban configuration in terms of spatial and functional pedestrian accessibility, for more inclusive and equitable cities.


Introduction
To create smart strategies for a sustainable and inclusive urban environment, it is more important than ever to focus on individuals and reaffirm a people centred planning vision (Rossetti et al. 2020). The acceleration of urbanization leads to a geographical extension of cities, longer distances and travel times, increased motorization and a constantly growing population. Faced with this situation, new urban policies are attempting to control the interaction between urban planning and urban mobility, and to redefine a new dimension of proximity, which reduces excessive use of energy-consuming and polluting modes of travel, namely vehicles.
Urban proximity induces numerous and widely shared benefits: promotion of sustainability, increased sociality and inclusiveness (particularly for disadvantaged groups), and improvement of psychophysical wellness (Weng et al. 2019;Cheng et al. 2019). There is then a direct relationship between the goal of improving sustainability and proximity, making enhancing proximity one of the most pressing issues confronting modern cities (Caselli et al. 2022).
The vision of urban proximity is embodied in the 15min cities concept. The concept of 15-min city is based on reducing travel distances between people's homes and their daily activities, seeking to humanize the scale of the city, abandon the motorized vehicle approach and combat urban sprawl (Correa-Parra, Francisco Vergara-Perucich, and Aguirre-Nuñez 2020). In this perspective, many recent studies have focused on urban walkable accessibility, developing indices to assess the performance of the pedestrian network and improve active mobility (Caselli et al. 2022). Linbo et al. (2017) In his work on the accessibility of metro stations, proposes and analyzes two concepts of accessibility: attraction accessibility, which is the ease of reaching a given station using certain modes of transport such as walking, cycling, bus, or cab, and radiation accessibility, which represents the ease of reaching other stations from a given station. The Xi'an, China metro transit system is used as a case study to develop the proposed accessibility concepts. The study shows that the attraction accessibility of the station's depends primarily on the pedestrian accessibility, so improving the pedestrian environment is the key factor in improving the overall accessibility of the station. Amini-Behbahani, Meng, and Ning (2020) present a study of walking distances between residential aged care facilities and services, as well as transportation, in the six largest Australian cities, which house more than half of the country's residential aged care centre capacity. Coffee shops, pharmacies, grocery stores, news agencies, lottery offices, and post offices were among the services available. According to the findings, most services are within walking distance of healthy inhabitants in all cities, however fewer mobile residents may have problems walking to services such as pharmacies. Campisi et al. (2021) develops a qualitative study based on the Commented Path Method (CPM) to generate underlying judgements from practical experience of individual paths to assess the walkability of urban landscapes from the standpoint of visually impaired people. Weng et al. (2019) proposes a modified method for assessing walkable neighbourhoods in Shanghai, China. The assessment examines the walking requirements of different pedestrian groups (i.e. the general population, children, adults, and the elderly), amenity qualities (size and category), and real traffic conditions based on amenity access.
Several studies have therefore focused on the spatial accessibility of urban services; however, accessibility should be assessed not only spatially, but also functionally. In addition to spatial accessibility, the capacity of the destination facility should be assessed in relation to the number of potential users. The facility must accommodate the potential population of its catchment area, otherwise it cannot be considered accessible. In this sense van Heerden et al. (2022) presents a study of accessibility from different perspectives: Different modes of transport, different categories of provision, different cost thresholds, and different levels of access for different areas of the city of Tshwane, a metropolitan municipality in Gauteng Province, South Africa. Two categories are developed to explain the influence of capacity on accessibility and equity.
In all the works studying the accessibility of facility services, Alrasheedi et al. (2020) was the only one to study the accessibility of mosques. He used GIS to identify spatial accessibility by car to mosques and to determine the disadvantaged Muslim communities in the Melbourne metropolitan area. This study did not consider the issue of sustainability as it focused on analysing the accessibility of mosques by car, and it is a very simplistic study as the number of mosques as well as the number of Muslim worshipers are naturally very low in non-Muslim countries compared to a Muslim majority countries, where the mosque is in the heart of every urban and even rural centre.
The mosque is a building where Muslims perform their prayers. Therefore, easy accessibility to mosques is vital for Muslim communities. Besides, because users will need to access the mosque multiple times during the day for their daily prayers, mosque accessibility is critical to sustainability. Thus, the ease of access of user, as well as the mode of transportation required to access the facility, can have a significant impact on the mosque's environmental sustainability (Azmi and Zin Kandar 2019). However, there is a significant issue with congregational prayer accessibility in Islamic cities due to the random spatial distribution and the insufficient capacities of existing mosques. In this perspective, the spatial and functional organization of mosques must be rethought in terms of proximity. Mosques must be accessible by walking, and must have appropriate capacity to accommodate a maximum number of worshippers. Within this framework, the present paper proposes a GIS-based methodology to measure, in the light of the 15min city concept, the current performances of an existing mosque from the spatial and functional view, considering on one hand the walkable accessibility of the mosque, and on the other hand the capacity of the mosque to accommodate the potential worshippers in its catchment area.

Spatial pedestrian accessibility
The objective of this section is to determine the catchment area of a mosque M. From any point of this catchment area one can reach the mosque M at a maximum walking distance. The catchment area of mosque M is created from a network, which we named the mosque's service network. The Mosque M service network consists of the edges or parts of the edges of the urban blocks from which the mosque can be reached within a maximum walking distance. The spatial analyses of this study are performed using the open-source software QGIS.
To create the service network, we used the Interface Catchment (IC) plugin of the QGIS software. IC measures the total length of public/private interfaces accessible from a starting point within a given walking distance. IC is relevant to walking because most urban attractions, such as stores, workplaces, and especially mosques, are accessible through the public/private interface, where buildings meet the street. The IC takes into account the width of streets, and assumes that a person can walk across any open space that is not occupied by urban blocks (Majic and Pafka 2019). (a) The service network of the mosque M at a walking distance of x metres. To each walking distance x metres correspond a walking time of t minutes; (b) The catchment area of the mosque M, which is the convex envelope that encompasses the mosque's service network at a walking distance of x metres. The catchment area is created using the Minimum Catchment Geometry plugin in QGIS; (c) The area whose population can spatially access the mosque M by a maximum walking distance x metres; (d) The spatially disadvantaged area or whose population does not have access to the mosque under the same distance.

Functional accessibility
In addition to its spatial position, a mosque is characterized by its capacity C, i.e. the maximum number of people who can access it and pray at the same time.
The functional accessibility of a mosque is then the ratio between the capacity of this mosque and the average target population that, theoretically, should have a comfortable access to the mosque at prayer time. We call P cx the average target population of the mosque in a catchment area of x metres. P cx is calculated from the total population P x in a catchment area of x metres by the following formula: The average target population of the mosque within a catchment area of x metres; P x : The average number of inhabitants within a catchment area of x metres; α : The ratio of practicing Muslims among the total Muslim population in a mosque's catchment area. α depends on the mosque's location, but especially on the type of the prayer. In the Moroccan case, the Ministry of Habous and Islamic Affairs, which is the state body in charge of the construction and management of mosques in Morocco, considers that α ¼ 10% of the population of the mosque's catchment area are supposed to go to the mosque to pray. For the weekly Friday prayer, which is a large gathering of worshippers to hear the Friday speech and perform the midday prayer, this ration reaches 20% ("Equipment in Mosques: New Urban Standards" 2022).
The total average population of the x metres catchment area is calculated by the following formula: S xi : The area of the urban area within the x metres catchment area of the mosque; D xi : Average number of inhabitants per dwelling in an urban area; L xi : Average number of dwellings per hectare in an urban area; The calculation of the average population of an x metres catchment area is done as follows: • Digitization of the study area development plan; • Intersection of the study area development plan with the catchment area, to obtain the urban areas within the catchment area and subsequently calculate the total average population of the catchment area; • Difference between the study area development plan and the catchment area, to obtain the urban disadvantaged areas and subsequently be able to calculate the total average population of the disadvantaged areas; • Each urban area is characterized by an average number of dwellings per hectare L xi , and an average number of inhabitants per dwelling D xi , these data are respectively taken from the urban planning regulations and the results of the general population census, and introduced in QGIS as attribute data of the urban areas; The images a, b et c in the Figure 2 show respectively the land use plan of the study area, the urban areas included in the catchment area and the deprived urban areas. To assess the functional accessibility of the mosque, we compare the capacity C of the mosque to the target population P cx in the x metres catchment area: If C ≥ P cx , the mosque is spatially-functionally accessible at x metres walking distance; If C < P cx , the mosque is spatially accessible, but functionally inaccessible at x metres walking distance.
This assessment aims to propose an intervention plan to the urban planning decision maker to make mosques spatially-functionally accessible.
We applied our method to measure the spatiofunctional accessibility of mosques in the Agdal district of the city of Fez in Morocco. Fez is a city in north-eastern Morocco that covers an area of about 332 km 2 . It is often considered the cultural and spiritual capital of the country. According to the 2014 General Census of Population and Housing (RGPH), the city of Fez, composed of six urban districts and one urban commune, is the second largest city in Morocco after Casablanca, with a population of 1 112 072 inhabitants, and with a projected average population growth of 14 000 inhabitants per year, to reach 1 252 000 inhabitants by 2024 (High Commission for Planning 2021).
According to data provided by the Ministry of Habous and Islamic Affairs in Morocco, the institution in charge of the construction and management of mosques in the country, the city of Fez has 451 mosques, 393 mosques in the urban area and 58 mosques in the rural areas of the city. The total area of the prayer halls of the said mosques is 201 000 m 2 , which can accommodate about 201 000 worshipers during the 5 daily prayer, considering that the unit area necessary for a person to perform the prayer comfortably is 1 m 2 ("Equipment in Mosques: New Urban Standards" 2022). The district of Agdal contains thirty mosques with different capacities, which are spatially distributed throughout the territory of the district. To determine the capacity of mosques in the study area we proceeded to take measurements of the surfaces of the prayer halls of each mosque, either on the plans of the mosques which were provided to us by the ministry of habous and Islamic affairs, or by going to the places of the mosques which do not have plans, and by taking the measurements of the surfaces of the prayer halls on the site. Table 1 gives the code and the capacity of each mosque, and the Figure 3 shows the spatial distribution of mosques in the Agdal district.
Our work is divided into two steps, the first consisted in the evaluation of the spatial accessibility of the mosques, the second had for objective to compare the capacity of mosques to their target population.

Spatial accessibility to the mosques
In his work presenting a methodology for assessing the potential of an urban area to become a 15-minute city based on its existing core urban functions, (Correa-Parra, Francisco Vergara-Perucich, and Aguirre-Nuñez 2020) considered that 1200 m and 600 m walking distances correspond respectively to 15 min and 30 min walking times. By the same logic we have opted in our study for 15 min, 10 min, and 5 min as respective walking times corresponding to 600 m, 400 m, and 200 m walking distances.
The spatial accessibility of the mosques was then evaluated with regard to 3 walking distances: 200 m, 400 m and 600 m. Figure 4 illustrates respectively the service networks and the catchment areas of each mosque M at 200 m, 400 m and 600 m walking distance.
Intersecting the catchment areas with the land use plan will give us the land use types in each catchment area, which will allow us to deduce the number of dwellings and thus the population in the catchment area of each mosque. However, we found, as shown in Figure 4 b), an overlap between the catchment areas of some mosques. For the 200 m catchment areas, this overlap is marginal, but it becomes more pronounced when the catchment area is extended to 600 m, making it impossible to define the actual catchment area of each mosque, as well as the population in each catchment area, which is the objective of this part of the study, since we are trying to calculate the target population for each mosque. To solve this problem, we used Voronoi polygons. Voronoi polygon is a tiling of the plane  into cells from a discrete set of points. Each cell encloses a single point and forms the set of points in the plane closer to that point than to any other. The discrete points in our case are the mosques. We then generated the Voronoi polygons corresponding to the mosques. The Voronoi polygons have thus allowed us to assign the population in the overlapping areas to their nearest mosques. We then performed the intersection between the Voronoi polygon corresponding to the mosques of our study, with the catchment areas of these mosques at 200, 400, and 600 m respectively. This operation allowed us to determine the catchment area specific to each mosque independently of the others, which facilitated the rest of the study. The results of this intersection are shown in Figure 5.
Then we made another intersection. This time between the new catchment areas of the mosques, and the land use plan of the Agdal district. The objective of this operation was on one hand to determine the types of land use, the number of housings, and especially the population of each catchment area, and on the other hand to determine the deprived areas, i.e. whose population cannot access spatially to the mosques at 200, 400, and 600 m respectively.
We first digitized the land use plan of the study area, Agdal district, then we deduced from this plan the nonresidential areas, namely: green spaces, non-aedificandi areas, reforestation areas, activity areas, parking lots, railway areas, etc. Figure 6 shows an excerpt of the land use plan of the study area with only residential areas. This analysis has identified three levels of spatial accessibility and three additional levels of spatial inaccessibility.
• First level of spatial accessibility: Refers to areas where the population can access the nearest mosque within a 5-minute walk; • Second level of spatial accessibility: Refers to areas where the population can access the nearest mosque within a 10-minute walk; • Third level of spatial accessibility: Refers to areas where the population can access the nearest mosque within a 15-minute walk; • First level of inaccessibility: Whose population needs more than a 5-minute walk to reach the nearest mosque; • Second level of inaccessibility: Whose population needs more than 10 minutes of walking to reach the nearest mosque; • Third level of inaccessibility: Whose population needs more than 15 minutes of walking to reach the nearest mosque; The government's and urban planners' strategies must take into account the three levels of spatial inaccessibility. Efforts to achieve spatial equity and proximity to worship facilities must act to eliminate or at least reduce this inaccessibility, starting with the third level of inaccessibility.

Functional accessibility of the mosques
After analysing the spatial accessibility of the mosques, we moved on to assess their functional accessibility. The objective of this section is to compare the capacity of each Mosque to the target population within its specific catchment area. Table 2 shows the designation of each urban area in the Moroccan urban regulations, the number of dwellings per hectare in each area Li, and the average number of inhabitants per dwelling Di. For each catchment area (200, 400 or 600 m) we have calculated the Area of each urban zone and the average population of each urban zone, then we deduce the average population of each catchment area. The following table represents the results of the calculations. It gives for each mosque the number of inhabitants in the three catchment areas, 200 m, 400 m, and 600 m away, and also the satisfaction rate of the inhabitants of each area, this satisfaction rate is the capacity of the mosque compared to the number of Table 3 is a dashboard for policymakers and urban planners to assess the level of functional accessibility for each mosque in its different catchment areas. The assessment looks at each mosque individually, but also in relation to other mosques, for a holistic assessment that will allow policymakers to determine priorities for intervention with the goal of ensuring equitable access to mosques. policymakers can start, for example, by reducing the inaccessibility of mosques where the satisfaction rate is less than 10%, coloured yellow in Table 3, and starting with the distant surroundings of each mosque, which in our case of study is 600-metre of walking distance, before moving to the close surroundings. The effort of the decision-makers must be focused on the analysis of the potential for enlarging the current mosques, of inadequate reception capacities, or to mobilize land and plan the construction of new mosques to absorb the need in place of worship.
In this regard, Table 3's findings demonstrate that 15 mosques out of the thirty mosques in the study area can accommodate more than 10% of the population within a 600-metre walking distance. The remaining 15 mosques must be enlarged, and if that is not possible, new mosques must be built. Once the accessibility of mosques is ensured for at least 10% of the population within a 600-metre walking distance, the focus should be on improving the accessibility of mosques at 400 and 200metre walking distances. However, before proceeding with this second and third step, the analysis conducted in this study will need to be redone, as it may be that improving accessibility at 600 m will automatically improve accessibility for some mosques at 400 and 200 m.

Discussion
The present paper proposes a GIS-based methodology to measure, in the light of the 15-min city concept, the current performances of existing mosques in Agdal district in Fez-Morocco from the spatial and functional view, considering on one hand the walkable accessibility of each mosque, and on the other hand its capacity to accommodate the potential worshippers in its catchment areas. Besides exploring pedestrian and functional accessibility to mosques in urban areas, this study can help urban planner to identify deprived and inaccessible areas in term of mosques, where new mosques should be built or existing mosques should be extended.
The outcomes of this research show that the notion of accessibility of a public facility, such as a mosque, must take into consideration not only spatial accessibility or walking distance, but also the ability of the facility to accommodate the target population of its catchment area. Planning proximity facilities is fine, but it will not help promote environmental sustainability, inclusiveness, and energy conservation if the user is forced to travel long distances because the nearby facility is unable to comfortably accommodate them.
The methodology proposed in this study was applied to mosques in the Agdal district of the city of Fez. The Agdal district is in the centre of the city and is surrounded by other districts to form the city of Fez. The spatial distribution of mosques in Agdal shows that several mosques are located almost at the edge of the district. Naturally, these mosques have a role for the population of the bordering districts and it is the same for the mosques of the bordering districts. It is therefore necessary to expand the scale of the study. Analysing the spatial distribution and spatial-functional accessibility of the mosques at the scale of the entire city will allow for more realistic results.
The GIS tool used in this work allowed us to create a service network for each mosque. Except that this service network does not take into account the topography of the land, flat or sloping, or the age or gender of the person walking. These factors strongly influence the walking time to reach the facility. It would therefore be wise to integrate the topographical data of the study area into our system, with a walking fluidity coefficient that reduces the walking time in the case of flat terrain and increases it in the case of uneven terrain, and to study in more detail the composition of the population of the study area, in order to have a clearer vision of the target population of each mosque.
To the above, we add another factor that seems to us extremely important in the analysis of the spatialfunctional accessibility of mosques, this factor we will call it the attractiveness of the mosque. This attractiveness is divided into two main axes, the comfort provided by the mosque, and the quality of the religious attendants who manage the Islamic rituals in the mosque. For the first axe, which is the comfort, it depends on several elements, such as thermal comfort, sound comfort, quality of equipment, etc. For the second axe which is the quality of the religious attendants, it depends on their level of studies, their level of learning the Koran, the beauty of their voice when they recite, etc. These two factors must in our opinion be studied and added to the factors previously mentioned to evaluate the level of accessibility of each mosque.

Disclosure statement
No potential conflict of interest was reported by the authors.