Impact of bus rapid transit in shaping transit-oriented development: evidence from Lahore, Pakistan

ABSTRACT Transit-oriented Development (TOD) reduces traffic congestion and encourages high dense, mixed-use, and walkable environments. Whether Bus Rapid Transit (BRT) shaped TOD or not has not been well examined, though rail transit has been widely studied. Therefore, this study fills this research gap and investigates what kind of urban fabric has been created by BRT’s shaping of the TOD within 800 meters between 2012 and 2021 in Lahore using three TOD criteria: density, diversity, and design. This study conducted observation surveys and interviews with the officials of departments to evaluate the aforementioned TOD criteria, and the collected data were analyzed using ArcMap. This study concluded that population density, development volume, and land use for economic activities all increased after the BRT operation. Pedestrian paths were not improved or remain the same, signifying that the walkability and open space either remained the same or declined in the station areas. However, intersection density increased in some station areas. The evidence in this study indicates that density and diversity improved but design criteria remained the same or declined. Therefore, to achieve TOD, the government should give priority to improving the walking environment, for instance, by creating more pedestrian paths and open spaces.


Introduction
Recently, cities have been expanding unprecedentedly due to rapid urbanization (Perveen, Kamruzzaman, and Yigitcanlar 2017).Consequently, many problems associated with managing the built environment have emerged; some of the most visible of these issues are caused by over-dependence on automobiles, such as climate change, global warming, traffic congestion, air pollution, and socio-economic issues (Sharma and Kumar 2012).The introduction of public transit is one of the essential countermeasures to overcome this overdependence on automobiles (Holmgren 2007;Kanthavel, Sangeetha, and Keerthana 2021;Mulalic and Rouwendal 2020).However, many cities have revealed that introduction of public transit is not enough because even with a public transit system, people prefer to use automobiles as they are more flexible, fast, and comfortable (Bergstad et al. 2011;Redman et al. 2013).Hence, some measures to encourage public transit use need to accompany the introduction of public transit.Controlling land use associated with public transit can encourage public transit usage, and one of the ways is transitoriented development (TOD).
TOD is considered a form of urban development that enhances residential and commercial places within walkable distances from transit stations (Chen et al. 2021;Lamour, Morelli, and Marins 2019).Generally, TOD offers higher density, mixed-use, walkable urban development near mass transit stations, and encourages the use of public transit and walking rather than automobiles (Nasri and Zhang 2014;Rahman, Yasmin, and Eluru 2019).As TOD consists of various aspects such as urban development and infrastructure in the station area, it is created through land use control to restrict/guide urban development and infrastructure development by the private sector (Lyu, Bertolini, and Pfeffer 2016;Pojani and Stead 2014;Renne 2017).To consider how land use control and infrastructure development promote TOD, understanding what urban fabric is created and what aspects of TOD are lacking in the station area without these factors is essential.
TOD is not dependent only on the heavy rail transit (HRT) system, but also depends on light rail transit (LRT) and bus rapid transit (BRT) (Knowles, Ferbrache, and Nikitas 2020).There are examples of past literature that have examined the impact of the HRT system on the development of station areas.Notwithstanding the rising reputation of the BRT system, little is known about its influence on land development (Cervero and Kang 2011;Rodriguez, Vergel-Tovar, and Camargo 2016).Moreover, the effect of BRT on land development is context dependent (Rodriguez, Vergel-Tovar, and Camargo 2016).Most CONTACT Muhammad Nadeem nadeem-muhammad-mb@ynu.jp Graduate School of Urban Innovation, Yokohama National University, 79-5 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan studies on the interaction between rail transit and TOD have been conducted in cities in developed countries; there is little evidence on TOD particularly in developing nations.
This study focuses on Lahore, Pakistan, where the BRT system has been in operation since 2013 but there is no particular land use control and infrastructure development policy intending to create TOD.The main objective of this study is to examine (1) what kind of urban fabric has been created in BRT station areas and (2) whether the urban fabric has elements of TOD, which encourage public transit use and walking rather than private vehicle use.In addition, this study intends to suggest policy recommendations for encouraging TOD to make a BRT system possible in Lahore.This paper is organized as follows: the second section highlights the current literature on the influence of transit on TOD.The third section consists of methods and materials, including the study area description and research methodology.The fourth section outlines the study results and discussion.The last section provides the conclusions and recommendations.

Literature review
As noted before, a considerable amount of literature has explored the influence of rapid rail transit, particularly LRT and Metro, on land-use changes (Bhattacharjee and Goetz 2016;Higgins, Ferguson, and Kanaroglou 2014;Hurst and West 2014;Sahu 2018;Tang et al. 2020;Wu et al. 2020;Zhao and Shen 2019), density (Masoumi and Shaygan 2016;Waintrub, Greene, and Ortúzar 2016), and accessibility (Caset, Vale, and Viana 2018;Mulley et al. 2016;Yang et al. 2019;Zhai et al. 2021;Zhang et al., 2020) associated with transit infrastructure.Regarding BRT, Bocarejo, Portilla, and Pérez (2013) found that Transmilenio BRT experienced a significant increase in density relative to zones where the system was not operating in Bogotá.Jun (2012) determined the influence of BRT on development patterns in Seoul.They found that the BRT system has influenced a rise in development density in the urban areas, playing a vital role in attracting firms from the outskirts into urban areas.A study conducted by Rodriguez, Vergel-Tovar, and Camargo (2016) to investigate the land development impact of BRT around stations in Quito and Bogotá found a heterogeneous influence of BRT on land development which was, however, highly dependent on the local setting.Bocarejo, Portilla, and Pérez (2013) also found that several shopping centers have been constructed close to the stations of Transmilenio BRT in Bogotá.They also demonstrated that BRT occurrence had not encouraged a more significant increase in commercial and residential land use.Deng and Nelson found that the impacts of BRT were positive on residential and commercial development in Beijing (Deng and Nelson 2010).Basheer, Boelens, and Bijl (2020) found that commercial activities have increased due to landuse conversions close to the BRT stations in Lahore.Similarly, Adeel et al. (2021) found diverse impacts on commercial, apartment, and mixed-use conversions of BRT Lahore.Zhou et al. (2016) stated that BRT has significantly shaped land use in Brisbane.Some studies have examined the impact of metro rail transit systems on urban development in literature.Guan (2019) explored the spatial patterns of urban development for the metro system in Shanghai from 1994 to 2010 and concluded that buildings constructed after 2000 and buildings within the urban core had more significant spatial proximity to metro stations within 500 m.Yang, Su, and Cao (2020) identified the relationship between distance to metro stations and changes in floor area in Shenzhen, finding that distance to metro station shows a threshold effect on development intensity.Zhang and Wang (2013) found that investments in mass transit positively influenced urban land development in Beijing.Tang et al. (2020) broadly examined the development of urban land uses within 500 m to metro stations between 1981 and 2017 in Hong Kong and inferred that commercial-office and business uses are the most preferred land uses within a metro catchment.
Notwithstanding several studies investigating the significance of the built environment for rail transit, not many have focused on BRT, particularly its assimilation with the surrounding area.BRT is a renowned strategy for evolving good transit networks, and it would be a sound transport alternative if it was connected with walkability in the neighborhood (Pezeshknejad, Monajem, and Mozafari 2020).They concluded that convenient usage of BRT needs constructed walkway networks that provide various station ways in Tehran.Wu and Zhou (2022) measured the accessibility of metro stations in Tianjin and found that accessibility to the metro depends on the street network around stations.
The past studies also investigated the effects of TOD on travel behavior, real estate prices, and community life (Abdi and Lamíquiz-Daudén 2020;Ibraeva et al. 2020).For example, the propensity to use transit in non-TOD areas was 1.4 times lower than in TOD areas, while walking and cycling were four times less likely around BRT stations in Brisbane (Kamruzzaman et al. 2014).Yu, Pang, and Zhang (2018) evaluated the price changes of commercial properties around BRT stations in Austin and concluded that the effect of BRT proximity on commercial properties was modest.Similarly, BRT in Guangzhou increased real estate prices by 30% during the first two years of transit operations (Suzuki, Cervero, and Luchi 2013).It is evident that TOD can induce gentrification around transit stations (Padeiro, Louro, and da Costa 2019).Brown (2016) found that transit-oriented gentrification was linked with BRT in Los Angeles.Kamruzzaman et al. (2014) evaluated levels of community life around BRT stations in Brisbane.They found that residents of TOD areas had robust social links among themselves compared to non-TOD residents.
Several criteria and indicators have been used in the literature to evaluate TOD.Cervero and Kockelman (1997) established 3Ds of urban structures that influence travel behavior, including density, diversity, and design, as the main features of TOD.Later, two more dimensions were added, namely distance and destination, expanding to 5Ds (Ewing and Cervero 2001).After that, Ewing and Cervero (2010) suggested 7Ds, added demand and demographics in 5Ds.Some studies have evaluated the TOD using 3Ds in Dhaka and Melbourne (Aston, Currie, and Pavkova 2016;Haquea, Sharnab, and Khanc 2019).Similarly, 5Ds used by various researchers in their studies to measure TOD (Islam et al. 2018;Dirgahayani and Choerunnisa 2018;Zhang et al., 2020;Teklemariam and Shen 2020;Liu, Zhang, and Xu 2020;Niu et al. 2021;Su et al. 2021).As for the node-place model, it was developed by Bertolini in 1996 based on railway stations and their surrounding areas (Bertolini 1996).This model offers an analytical framework to delineate transit (node) and neighborhood (place), features of a location and their associations.In various studies, the node-place model has been applied for TOD classification around transit stations (Balz and Schrijnen 2009;Bertolini 1999;Reusser et al. 2008).
BRT has limited empirical evidence for shaping cities (Ogra and Ndebele 2014).Most studies on the interaction concerning urban development and rail transit have widely been conducted in the cities of developed nations, whereas studies from developing countries are limited (Wang et al. 2019).
To conclude, while the literature provides plentiful findings regarding the influence of rapid transit on urban development in advanced nations, these outcomes are not handy for developing economies due to the varying contexts.However, whether BRT infrastructure shaped TOD or not has not been well examined.This is evident as most studies investigated rail transit rather than BRT for supporting TOD.Moreover, the BRT and its role compared to other transit infrastructures merit investigation to apprise the prioritization of transport investment in developing nations.To fill this gap in research, this study investigated whether BRT investment created urban fabric and what aspects of TOD are lacking in the station areas of Lahore.

Research setting
This study focuses on Lahore.The city's total population was about 11.12 million in 2017, with a land area of 1772 km 2 (Punjab Bureau of Statistics 2017).Due to the rapid urbanization process, the city is sprawling in a less-controlled manner and has crossed a radius of 38 km in 2017.In Lahore, the BRT system was introduced in 2013 to reduce the growing congestion, serving 27 stations with a total length of 27 km (see Figure 1), integrated with Speedo Bus Service (Punjab Mass transit Authority 2021).The BRT currently carries approximately 135,000 commuters daily.According to the Punjab Mass transit Authority, it has attained the highest daily ridership of about 180,000 people.However, according to the official's interview, no policy, regulations, rules, and incentives regarding BRTbased TOD have yet been prepared by the government.

Criteria of TOD and the data source
In literature, various approaches were used for measuring TOD.Cervero and Kockelman (1997) established the 3Ds of urban structures influencing travel behavior, namely density, diversity, and design.These 3Ds are also useful in explaining how land use patterns influence travel patterns (Sarkar and Mallikarjuna 2013).Singh et al. (2017) identified eight criteria to measure TODness, including density, land use diversity, walkability and cyclability, economic development, userfriendliness of a transit system, accessibility, transit capacity utilization, and parking at the station.Moreover, Su et al. (2021) identified various indicators under the 5Ds framework of TOD, such as design, including intersection density and street network connectivity; density, which indicates the population and road density; diversity, which measures the land use proportion and mixture pattern; and destination, which represents the trip attractions accessibility; and demand management, which denotes the traveling services.We chose the 3Ds criteria to measure how close the urban fabric in a BRT station area is to TOD because the 3Ds of urban structures influence travel behavior, as a feature of TOD (Cervero and Kockelman 1997;Sarkar and Mallikarjuna 2013) (see Table 1).Moreover, the selected set of criteria and indicators are measurable, evaluate the previous and current situations, and for which data is readily available and collectible.
Higher density around the stations can encourage the use of BRT.Greater diversity can attract people to visit BRT station areas by BRT and walk around in the station area.Walking friendly design can promote walking in the proximity of BRT.The indicators in each criterion are selected by the below reasons and availability of data.To understand the impact of BRT, this study compares the situation before BRT operation (2012) and the current one (2021) and examines the change of the indicators.
In terms of density, population density is used as criteria because if more people live in the station area, more people will use the BRT, and the population density attracts various urban development in the station area.The population data for the years 2012 and 2016, pertaining to the study area were collected from Punjab Development Statistics.The collected population was available at Union Council (UCs) level, also known as blocks.That is why average population density was considered at the UCs level for observed BRT stations.The concept of Floor Area Density (FAD) was introduced and calculated under density criteria.FAD is calculated by the following formula: where l(i) is the land plot area of developed land i and f(i) is the number of building floors in the developed land i.The data for floor area ratio (FAR), the total floor area of building divided by the area of the land plot, is more accurate to show the volume of the development.However, the data size of each building is not available.Therefore, FAD is used to show the total volume of development in each station area in this study.Higher FAD means more development and activities; this attracts more people to settle down and visit the station area and promote BRT use to visit this area.The floor area for 2012 and 2021 was calculated by multiplying the plot's total size with building's height.After that, FAD was determined by dividing the total floor area by the total station area in each station.
In terms of diversity, land use diversity is used as a criterion, because a range of land use leads to various activities and increase of convenience in the area.It attracts more people to the area and therefore more  where Pi represents the proportion of each land-use type, and n shows the total number of land uses in the observed stations.The resulting value of the index falls between 0 and 1. Complete homogeneity of land use is indicated by 0, and 1 denotes absolute heterogeneity of land use.A higher entropy index value demonstrates higher land-use diversity, thus signifying higher levels of TOD (Huang et al. 2018).
Design is a critical part of establishing accessible environments for encouraging walking, a vital element for TOD (Teklemariam and Shen 2020).In addition, the design element is not only associated with the travel choice; it also develops a critical link in the spatial organization between residents and stations, where it supports creating high-quality open spaces (Niu et al. 2021).In our study, the following criteria were selected; pedestrian paths (m), the area of open spaces and parks (ha), and intersection density (number of intersection/ha).In the 2021 observation survey, pedestrian paths were identified, and the length of footpaths was calculated using GIS.After that, interviews were carried out with LDA, local government, and transport department officials about whether pedestrian paths were developed/improved or not after the BRT infrastructure.Moreover, the ratio of pedestrian paths based on existing footpaths shows walkability and accessibility, and was determined using the following formula.
In addition, the open spaces ratio is used as one of the indicators of walkability in this study.According to McCormack et al. (2010), the presence of high-quality green open spaces and parks encouraged walking in the neighborhood.The quality and size of open spaces influenced the neighborhood's pedestrian activities (Koohsari et al. 2013).Zhang et al. (2020) and Zlot and Schmid (2005) found that more walkable neighborhoods have more parks.Similarly, Adams et al. (2011) determined that transportation walking was higher proximate to parks.The area of open spaces and parks was identified from the land use data in 2012 and the observation survey in 2021 and calculated using GIS.In our study, the open space ratio was calculated using the following formula.
To calculate intersection density, at first, the road network for 2012 was digitized by using the historical imagery option in Google Earth, and the road network for 2021 was taken from the Open Street Map.In addition, the road network for 2021 was corrected using Google Earth.We computed the intersection density using the road network considering the three or more-way intersection located in the study area using GIS.Cul-de-sacs were excluded from the analysis in this study.According to Islam et al. (2018), cul-desacs reduce accessibility to destinations and increase the commuting distance.Islam et al. (2018) and Cervero, Murakami, and Miller (2010) determined the intersection density using the following equation.
where TI represents the total number of three or moreway crossings, and A shows the land area (ha) at observed BRT stations.

Study area
As this study collected some data via observation survey, we needed to select stations for the study.First, a discussion was conducted with five professionals of the Lahore Development Authority (LDA), local government, and the transport department officials in Lahore in August 2021 to select BRT stations for this study.From the discussion, we expected to learn about location, population density, station ridership, development type (controlled/less-controlled), development age, and area characteristics effects on the density, diversity, and design in the station area.Lesscontrolled area refers to areas which are regulated and controlled by government authorities.Most of the city area is regulated and controlled by the local government but overlapping functions and power among various local government departments and lack of government staff results in weak control.The land use of some area is regulated and controlled by the "cooperative housing societies." A cooperative housing society is a community organization, is regulated by the Cooperative Society Act, 1925, has the authority to formulate their own regulations, and control the area following the regulations.Model town and Naseerabad have cooperative housing societies, and their land use is regulated by both local government and housing society's regulations.It results in a well-planned and controlled area and we refer to as controlled area.On the other hand, land use in other areas is regulated only by local government, refer as a less controlled area.These criteria (see Table 2) were applied to all BRT stations as selection criteria and we chose the eight stations ensuring heterogeneity in criteria weightage.The selected stations were Shahdara, Bhatti Chowk, Ichra, Model Town, Naseerabad, Kamahan, Nishtar Colony, and Dullu Khurd (see Figure 1), and their characteristics are summarized in Table 2.
In this study, the urban fabric in the station catchment area are investigated.Schlossberg (2006) considered a 10-minutes walking distance of 800 m to evaluate the TOD level, which is pertinent to TOD planning.Our study also selected an 800 m radius from BRT stations in Lahore.

Population density
Table 3 shows the average population density within an 800 m radius for the selected eight BRT stations between 2010 and 2016.There are great differences in population density from station to station.The two stations in the city center, Bhatti Chowk and Ichra have the highest population density and the two stations in the suburban area, Nishtar Colony and Dullu Khurd station have the lowest population density consistent across time.Population density in all of the eight stations increased from 2010 to 2016.The biggest change was 14% in Kamahan and the smallest change is 11% in the two suburban stations such as Nishtar Colony and Dullu Khurd.The differences in rate of change is not as large as the differences in population density.The average population density of Lahore accounts for 45 PPH in 2010 and 51 PPH in 2016, with a percentage change of 13%.Comparing with this rate of population change in the whole city, the population density in the eight BRT station area has not increased significantly.Our findings are inconsistent with those of Masoumi and Shaygan (2016).They found that the population density of the observed metro stations significantly increased between 2005 and 2015 in Tehran, Iran.Similarly, Bocarejo, Portilla, and Pérez (2013) found that BRT has a significant increase in density relative to zones where the system was not operating in Bogotá.

Floor area density
Table 4 shows the Floor Area Density (FAD) in the eight BRT station areas in 2012 and 2021.Figure 2 shows the station-wise land-use changes in the study area between 2012 and 2021.FAD was increased in all stations but there are great differences in FAD between the stations both in 2012 and 2021.In 2021, the two stations in the city center have the highest FAD, 2.14 and 1.92 and the Shahdara and Dullu Khurd have the lowest FAD, almost 1.15.The rate of change in FAD is highest in Nishtar Colony station.In 2012 the Nishtar Colony station area had large vacant and agricultural land.This type of land use has greatly decreased in proportion, and other land use such as mixed use and industry have increased (see Figure 2).The FAD has greatly increased in these areas.The station area in the city center, Bhatti Chowk station, also increased in FAD significantly, by 18.5%.This station area has almost no agricultural land and a very small proportion of vacant land (see Figure 2), so floor area was expected to be increased by rebuilding or adding additional floors.The two controlled areas, Model Town and Naseerabad, have the lowest rate of FAD change, 6.50% and 2.59%, respectively.This was expected because of strict land use control under building and zoning regulations.Our findings are consistent with those of Deng and Nelson (2013).They concluded that BRT has some positive impacts on land development around the stations in Beijing.

Land-use change
Table 5 shows the land-use changes within the study area in 2012 and 2021.Residential land use is continuously declining from 613 ha in 2012 to 523 ha in 2021, a decrease of 14.6%.Residential use has decreased significantly in Bhatti Chowk, Ichra and Kamahan stations.In these three stations, mixed-use has significantly increased.Mixed-use is defined as "land use which enables a range of land use including residential, commercial, and institutional to be co-located in an integrated way" in the Lahore Development  Authority's land use rules.Therefore, residential land use has not declined as much as 14.6%, because it is included in the mixed-use category, which increased significantly by 83.3%.From the increase in population density, residential floor area, including both residential use and residential part in mixed-use, is expected to have increased in the study area.A significant increase is noticed in commercial use, by 54.1%.As commercial floors are also included in mixed-use, the floor area for commercial use is expected to have increased more than previously estimated.The commercial use has significantly increased in Shahdara, Ichra, and Kamahan stations.Industrial use has also increased by 31.4% and significant changes were    Bocarejo, Portilla, and Pérez (2013) concluded that some vital shopping centers had been built around the terminals of the Transmilenio BRT in Bogotá, however, the BRT occurrence does not influence a greater rise in built-up areas for commercial and residential use.Our some study findings are consistent with those of Deng and Nelson (2010).They found that BRT positively impacts residential and commercial development in Beijing.However, in our study BRT investment does not influence a rise in residential land use.

Entropy index
In our study, the entropy index was used to estimate the diversity of land use close to BRT stations.A higher value of the entropy index shows higher land use diversity.

Pedestrian paths
Table 7 illustrates the pedestrian paths and their ratios at the eight stations in 2021.Due to lack of past data for pedestrian paths, we cannot calculate the ratio of pedestrian paths in 2012.However, according to interviews with officials of LDA, local government, and transport department, after the introduction of BRT, pedestrian paths were not developed in this study area.In addition, extensive changes did not occur in the road network.Therefore, the ratio of pedestrian paths was the same for each station between 2012 and 2021.The pedestrian path ratio in all stations is quite low, representing the lack of walkability and accessibility.The ratio of pedestrian paths was especially low in older areas such as Shahdara, Bhatti Chowk, Ichra, and Kamahan.

Intersection density
Table 8 shows intersection density for 2012 and 2021.
There is a wide range of intersection density, from 1.67 to 8.05 in 2012 and from 1.74 to 8.05 in 2021.Jacobs (1993) determined the intersection density in Venice was 5.79/ha, that of downtown Los Angeles was onetenth that of Venice, and that of Irvine, California, was one-tenth that of downtown Los Angeles.Compared with the intersection density in these Western cities, some station areas are more walkable than Venice (Ewing 1999;Jacobs 1993).
Bhatti Chowk and Ichra have the highest intersection density of 8.05 and 7.80, respectively, due to the old built-up area and several public facilities.These station areas have association between land use mix and street network.However, there was no

Conclusions and recommendations
This study investigates the impacts of BRT on the urban fabric in Lahore from the viewpoint of 3Ds of TOD.The results shows that the urban fabric has changed after BRT operation, but it is dependent on the characteristics of each BRT station area.After the BRT operation, population density and development volume increased.However, the population density in the study area has not increased significantly compared to Lahore's population density rate.Additionally, land use for economic activities, such as commercial and industrial use, also increased.On the other hand, vacant land, open space, and agricultural land declined significantly in terms of area.Land use diversity also increased in most of the target stations.Thus, in density and diversity, elements of TOD were encouraged in the station area.Regarding design, which shows walkability in the station area, the three indicators declined or slightly increased.Before the BRT operation, the pedestrian paths were mostly well equipped.There was no improvement after BRT operation.The intersection density of the eight station areas lies almost in the same range as that of Venice, Italy and it increased in some areas.The area of parks remained the same or declined in all areas.The area of parks declined significantly in a controlled area, Model Town.From this result, after BRT operation, the urban fabric created in the BRT station areas to some degree, such as density and diversity improved but design, especially pedestrian paths and open space remained the same or declined.Moreover, the urban fabric around BRT station areas has elements of TOD to some extent.In some measures, TOD elements such as density are satisfied but do not significantly grow and diversify are fully satisfied, while in other elements such as walking are not fully satisfied.It is expected that the diversity element can encourage the use of BRT rather than private vehicle use around station areas.Therefore, to achieve TOD, the government should give priority to improving the walking environment, for instance, by creating more pedestrian paths and open spaces.Comparing change in population density in the station area with the whole city, the growth of population density in the station area is almost same as in the whole city.Although the station area became more convenient because of increase in FAD and land use diversity, the station area did not attract people as a residential area would.Thus, to increase population density, improvement of convenience is not enough, and improvement of living environment is also very important.
This study offers the following recommendations for urban planning/policy in Lahore: in city centers and some urban areas, there is almost little to no vacant land.In order to secure space for open space and pedestrian paths, introduction of the incentive system where developers receive incentive such as a FAR bonus as a reward of providing open space or pedestrian paths in their development plots, would be effective.In suburban areas and some urban areas, vacant land remains and securing space for pedestrian paths and open space in the vacant land before development should be necessary.In the city centers and urban areas, the new development will be infill and vertical development because of the unavailability of more vacant land.To achieve higher density, high-rise redevelopment is to be necessary in a subdivided lot.
In such places, it would be effective to introduce an urban redevelopment scheme that facilitate subdivided lands to be developed in an integrated manner.
In the controlled area, there were large open spaces, but parts of the open space were developed.To maintain open space, encouraging infill and vertical development is also essential in the controlled area.In lesscontrolled areas more development proceeded than in controlled areas.Accumulation of new development might harm the living environment in the lesscontrolled area.Therefore, policy guidelines should be necessary to prevent environmental deterioration due to urban development.This study has some limitations: we compared station areas and the whole city only by population density.This study revealed that FAD and land use diversity increased in the station areas, but we were unable to determine whether this was a particular tendency in the station areas or not.Further, the number of 3D criteria is smaller than in other studies because of data availability.
In future studies, comparing station area and the whole city or non-station area is necessary to understand the characteristics of station area.This study can also be enhanced by considering more criteria and indicators and the whole route of BRT, which would offer a more thoughtful identification of TOD areas after BRT investment in Lahore.

Figure 1 .
Figure 1.(a) Lahore city with administrative zones and study area (b) A study area within a catchment of 800 m from BRT stations.

Figure 2 .
Figure 2. Percentage of land use in each station within the study area between 2012 and 2021.

Table 1 .
A set of criteria and indicators used in the study.
(Cervero and Kockelman 1997;Frank et al. 2006;Kockelman 1997;Nasri and Zhang 2014;Niu et al. 2021 floor having commercial use and subsequent floors having residential or institutional use.We have included such buildings in the mixed-use category.The number and the types of land use categories were the same as the 2012 data, and we applied the same survey methods used then.To measure the land-use diversity, "entropy index" is applied.Cervero & Kockelman and other researchers used the land use entropy index to measure land use diversity in their study(Cervero and Kockelman 1997;Frank et al. 2006;Kockelman 1997;Nasri and Zhang 2014;Niu et al. 2021).

Table 2 .
Characteristics of the selected BRT stations in Lahore.

Table 3 .
Average population density of 2010 and 2016 with average percentage change in density.

Table 4 .
Percentage change in floor area density.

Table 5 .
Percentage change in land use area of eight BRT stations for 2012 and 2021.
Table 6 shows the entropy index values around BRT stations.The entropy index value in the two stations were the lowest in the two controlled areas, Model Town and Naseerabad both in 2012 and 2021, due to the strict land use control.The entropy index increased for all stations between 2012 and 2021, except Bhatti Chowk and Nishtar Colony, which had relatively high index values in 2012.

Table 6 .
Entropy index in BRT station areas.

Table 7 .
Ratio of pedestrian paths in BRT station areas.

Table 8 .
Intersection density in station areas 2012 and 2021.
Table 9 shows the area of open spaces and parks (ha) with open space ratio in the eight BRT station areas for 2012 and 2021.The two controlled areas have the largest open spaces both in 2012 and 2021.The area of open space decreased in four stations and remained the same in the other four stations from 2012 to 2021.

Table 9 .
Open space ratio with percentage rate of change for 2012 and 2021.