key: cord-310337-vw4cno4p
authors: Du, Yaxing; Mak, Cheuk Ming
title: Improving pedestrian level low wind velocity environment in high-density cities: A general framework and case study
date: 2018-08-04
journal: Sustain Cities Soc
DOI: 10.1016/j.scs.2018.08.001
sha: 
doc_id: 310337
cord_uid: vw4cno4p

An acceptable pedestrian level wind environment is essential to maintain an enjoyable outdoor space for city residents. Low wind velocity environment can lead to uncomfortable outdoor thermal experience in hot and humid summer, and it is unable to remove the pollutants out of city canyons. However, the average wind velocity at pedestrian level is significantly lowered by closely spaced tall buildings in modern megacities. To improve the low wind velocity environment at pedestrian level in high-density cities, a general framework and detailed guidelines are needed. This study is the first time to develop such a framework, and provide detailed guidelines for improving pedestrian level low wind velocity environment in high-density cities. Additionally, a detailed review and summarisation of evaluation criteria and improvement measures are presented in this paper, which provide additional options for urban planners. To investigate the performance of the framework, the Hong Kong Polytechnic University campus was utilised as a case study. Results showed that pedestrian level wind comfort was greatly improved with the developed framework. The outcomes of this study can assist city planners to improve the low wind velocity environment, and can help policy makers to establish sustainable urban planning policies.

Outdoor spaces are considered vital constituents of urban environments because they can host entertainment activities that are fundamental to the character of a city and the quality of life of city residents (Chen & Ng, 2012) . In particular, outdoor human comfort is extremely important in city planning because it can not only improve the physical and mental health of city residents, but also can help reduce power consumption in residential buildings (Amindeldar, Heidari, & Khalili, 2017; Chatzidimitriou & Yannas, 2016; Elnabawi, Hamza, & Dudek, 2016; Kong et al., 2017; Li, Huang, Wu, & Xu, 2018) . Actually, an enjoyable outdoor thermal comfort can encourage city residents to spend more time in outdoor spaces (Chen & Ng, 2012; Du, Mak, Huang, & Niu, 2017) . However, congested airflow at the pedestrian level has become a major concern in the high-density cities because it is the driving force for the transfer of pollutant, heat, and water vapour (Ai & Mak, 2015; Chatzidimitriou & Yannas, 2017; Ignatius, Wong, & Jusuf, 2015) . This problem is more serious in densely built cities at low or mid latitudes, such as Hong Kong and Singapore, which suffer from urban heat island and global warming Kong et al., 2017; Ng, 2009; O'Malley, Piroozfar, Farr, & Pomponi, 2015; Yang, Wong, & Jusuf, 2013 ). Therefore, improving pedestrian level wind environment in high-density cities has become a pressing issue for the establishment of comfortable, healthy, and sustainable cities.

A strong relationship between low wind velocity environments at pedestrian level and unfavourable outdoor living environments has been widely identified in high-density cities (Ignatius et al., 2015) . For sub-tropical and densely built cities, such as Hong Kong, the wind speed and radiant temperature are the most influencing factors of pedestrian thermal comfort in hot and humid summers . In particular, Cheng, Ng, Chan, and Givoni (2012) reported that even in shaded environments, the thermal sensation of respondents was not neutral when wind velocity was below 0.9 m/s during the hot and humid summer of Hong Kong. The cooling effect of wind flow can contribute to the reduction of heat stress in urban cities (Memon, Leung, & Liu, 2010; Priyadarsini, Hien, & Wai David, 2008; Yang & Li, 2011) . It was identified that a wind speed of 1 m/s to 1.5 m/s could lower the air temperature by nearly 2°C (Erell, Pearlmutter, & Williamson, 2012) . Thus, low wind velocity environment has a negative effect on outdoor thermal comfort, particularly in hot and humid summers. In addition, the dispersion and dilution of airborne pollutants in the urban environment depend strongly on wind flow because it can help remove pollutants to external surroundings (Ai & Mak, 2014a ,b, 2016 Hang & Li, 2011; Kim & Baik, 2004; N'riain, Fisher, Martin, & https://doi.org/10.1016/j.scs.2018.08.001 Received 2 May 2018; Received in revised form 31 July 2018; Accepted 1 August 2018 Littler, 1998) . The outbreak of severe acute respiratory syndrome (SARS) in Hong Kong urged the officials and city residents to consider the pollutant dispersion in their living environment (Li, Duan, Yu, & Wong, 2005) . By analysing the results from field measurements, Jones, Fisher, Gonzalez-Flesca, and Sokhi (2000) indicated that traffic pollutants could not be properly dispersed or diluted with weak wind conditions in urban areas. Moreover, low wind velocity environments in the street canyons could not form the coupling process to remove the pollutants from the street canyons to the upper atmosphere (Jones et al., 2000; Vignati, Berkowicz, & Hertel, 1996) .

There have been a substantial investigations into the improvement of pedestrian level wind environment in the past few decades (Blocken, Stathopoulos, & van Beeck, 2016; Du, Mak, Huang et al., 2017 , 2017c Hang, Sandberg, & Li, 2009; Kubota, Miura, Tominaga, & Mochida, 2008; Liu, Niu, & Xia, 2016; Mirzaei & Haghighat, 2010) . Tsang, Kwok, and Hitchcock, (2012) conducted wind tunnel tests to investigate the influence of building dimensions, separations, and podiums on pedestrian level wind environment. Their results indicated that wide buildings with podiums should be discouraged in building designs in high-rise cities, for resulting in large areas of low wind velocity environment around buildings. In the study of Kubota et al. (2008) , the relationship between the gross building coverage ratio and the wind environment at pedestrian level from actual Japanese city was quantitatively studied using wind tunnel tests. The results showed that a small gross coverage ratio should be adopted in city planning because a high gross coverage ratio results in low wind velocity at pedestrian level, especially for the detached houses. Niu et al. (2015) conducted field measurements in a university campus in Hong Kong, and their findings indicated that local wind amplification could be achieved by utilising an innovative building design (lift-up design). numerically studied the wind environment in idealized high-rise urban areas, and reported that building height variation could improve pedestrian level wind environment. Apart from the above mentioned measures, the Hong Kong government has issued the air ventilation assessment (AVA) scheme to enhance wind flow movement at pedestrian level (Ng, 2009) . Later, new wind comfort criteria were proposed by , which were aimed at assessing wind comfort in low wind velocity environments. However, no general framework that provides detailed application procedure has been presented. Therefore, a general framework that can be applied to improve pedestrian level wind environment is in urgent needed.

In this study, a general framework for improving low wind velocity environment at pedestrian level was developed, and this paper provides detailed guidelines for the framework. Moreover, measures that have been utilised in improving the low wind velocity environment in highdensity cities were thoroughly reviewed. Meanwhile, the evaluation criteria aimed at assessing low wind velocity environment at the city scale are summarized in this paper. A case study of the Hong Kong Polytechnic University (HKPolyU) campus model was used to elaborate this framework and demonstrate its performance. The remainder of this paper is structured as follows: Section 2 describes the general design framework for improving the low wind velocity environment and describes its detailed guidelines. Sections 3 and 4 summarise the evaluation criteria and improvement measures. The performance of the design framework was investigated with a case study of HKPolyU campus, which is presented in Section 5. Concluding remarks are given in Section 6.

This section describes the design framework and guidelines in detail. Note that the present study only concerns the improvement of low wind velocity environments in high-density cities. However, the general framework developed in this study can also be used to mitigate the nuisance caused by strong wind conditions. In doing so, the improvement measures and evaluation criteria need to be re-determined. A flow chart of the design framework for improving the pedestrian level low wind velocity environment in a high-density city is shown in Fig. 1 . Meanwhile, the basic parameters for the design framework are summarized in Fig. 2 . Based on Figs. 1 and 2, the detailed guidelines for the improvement of pedestrian level low wind velocity environment are presented as follows:

• Step 1: basic information on the target area should be collected before evaluating the wind environment at pedestrian level. This includes the target area, surrounding area, local wind climate, building information, and geomorphological information. It is well known that urban wind flow is very complex and is closely related to the urban morphology, especially in high-density cities. Thus, for scaled models, accurate reproduction of the buildings and geomorphological features of the target area are very important. These Y. Du, C.M. Mak Sustainable Cities and Society 42 (2018) 314-324 geometric data can be obtained from GIS or government departments. Moreover, the target area should include surrounding buildings and infrastructure, which have a significant impact on the pedestrian level wind environment in the target area. Furthermore, the local wind climate should be acquired from nearby weather stations.

•

Step 2: based on the obtained basic information, the wind environment at pedestrian level can be obtained by field measurements (Jones et al., 2000; Niu et al., 2015) , wind tunnel tests (Du, Mak, Huang et al., 2017; Kubota et al., 2008; Tsang et al., 2012) or computational fluid dynamics (CFD) simulations (Blocken, Janssen, & van Hooff, 2012 Du, Mak, Liu et al., 2017; Du, Mak, & Ai, 2018; Tominaga, Mochida, Murakami, & Sawaki, 2008) . The field measurements should be conducted for a long time due to the uncertainty and unsteadiness of meteorological conditions (Schatzmann & Leitl, 2011; Schatzmann, Rafailidis, & Pavageau, 1997) . Moreover, the building models in the wind tunnel tests or CFD simulations should be replicated as close to the real buildings as possible. In addition, the prediction process should rigorously follow the practice guidelines and regulations, such as the AWES and ASCE quality assurance for wind tunnel (ASCE, 1999; AWES, 2001) and COST and AIJ guidelines for CFD simulation (Franke, 2007; Tominaga, Mochida, Yoshie et al., 2008 ).

Step 3: the low wind velocity area that needs to be improved can be determined by combining the prediction results obtained in Step 2 and the evaluation criteria. The main features of the evaluation criteria and parameters for improving low wind velocity environments are summarized in Section 3.

Step 4: the low wind velocity environment can be improved by adopting improvement measures such as using the lift-up design to create a void space underneath the building (Du, Mak, Liu et al., 2017; Niu et al., 2015) or adopting openings in building design to guide wind flow through buildings (Fan, Chau, Chan, & Jia, 2017; Yuan & Ng, 2012) . The main findings of previous studies on improving weak wind environment by adopting innovative building designs are detailed reviewed in Section 4.

• Step 5: the pedestrian level wind environment should be re-evaluated after adopting improvement measures. If the improved wind environment meets the requirements of the evaluation criteria, the wind environment in the target area is then viewed as an acceptable wind environment. Otherwise, new improvement measures should be adopted in the target area as illustrated in Fig. 1 .

Similar to the fact that strong wind conditions can be evaluated by using the existing wind comfort criteria (Hunt, Poulton, & Mumford, 1976; Isyumov & Davenport, 1975; Lawson, 1978; Melbourne, 1978; Willemsen & Wisse, 2007) , the low wind velocity environment should be assessed by suitable evaluation criteria. The wind comfort criteria aimed at assessing low wind velocity environments in high-rise cities have been established in past years. This section presents a detailed review and summarisation of the evaluation criteria for low wind velocity environments, as well as the evaluation parameters for outdoor wind environments.

In order to enhance pedestrian level wind ventilation, the Hong Kong government has stipulated elaborate technical guidelines and protocols, titled "Feasibility study for establishment of air ventilation assessment (AVA) system" (Ng, 2009 ). This AVA scheme was established specifically for low wind conditions in the high-density urban area. The main purpose of the AVA scheme is to improve wind flow movement at pedestrian level and to assess the influence of newly built developments on the neighbouring wind environment. The principle of the AVA scheme is "the more, the better"; and thus only a low wind threshold value was established in this scheme (Ng, 2009 ). This means that the pedestrian level wind environment is acceptable when the wind velocity is over 1.5 m/s for 50% of the time.

It can be seen that the AVA scheme aiming for enhancing pedestrian level wind environment has no specific regulations on wind comfort.

Unlike the wind comfort criteria adopted from overseas, which mainly focuses on strong wind conditions (Hunt et al., 1976; Isyumov & Davenport, 1975; Lawson, 1978; Melbourne, 1978; Willemsen & Wisse, 2007) , the new wind comfort criteria proposed in our earlier study are intended specifically for low wind velocity environments . Additionally, the new wind comfort criteria are established seasonally in consideration of the subtropical weather conditions in Hong Kong. A low wind threshold parameter from AVA scheme was adopted for hot season criteria. Moreover, the parameters in the new wind comfort criteria were chosen by thoroughly reviewing the existing wind comfort criteria to adapt to low wind conditions. Similar to the indicators in the AVA, the mean wind velocity ratio (MVR) and overall mean wind velocity ratio (OMVR) were utilised as indicators in the new wind comfort criteria. These two indicators are defined as follows:

here, i is the approaching wind direction; MV pi is the mean wind velocity at any area of pedestrian level when the approaching wind direction is i ; MV r is the mean wind velocity at a reference height (200 m in the prototype) when the approaching wind comes from i direction; and P i is the occurrence probability of approaching wind coming from i direction.

As mentioned earlier, the OMVR is used directly instead of using actual wind velocity because of its convenience for scaled models, namely CFD simulation which can provide sufficient spatial data or wind tunnel measurement when sufficient measurement points are used. Note that the application of wind comfort assessment can also be achieved by on-site measurement on the condition that the monitored points are sufficient enough and the wind data has been monitored for a long time, i.e. several months or years. Details of the new wind comfort criteria are listed in Table 1 .

Apart from the above two criteria that focus on low wind environment at pedestrian level, different parameters for evaluating the outdoor wind environment in urban areas have also been proposed and utilised in current researches. Some of them were adopted from indoor ventilation research. Detailed descriptions of the commonly used assessment parameters and their corresponding studies are summarized in Table 2 .

In view of the congested airflow in high-rise cities, many innovative building designs (passive and active) have been intensively investigated in the past few years. Since the geometric characteristics of buildings can modify airflow patterns at pedestrian level, specific building designs should be deployed to improve pedestrian level wind environment. This section provides a detailed review of the improvement measures in current studies, which also provides additional options for city planners to improve the low wind velocity environment at pedestrian level.

The lift-up design, in which the main building structure is elevated off the ground by pillars or a combination of columns and shear walls, is a potential solution for improving the low wind velocity environment at pedestrian level. A schematic diagram and a photo of the lift-up design in a university campus are shown in Fig. 3 . The void place underneath the elevated building is known as the lift-up area. The main purpose of the lift-up area is to minimize the wind flow obstruction at pedestrian level and to function as a corridor for the wind flow. Thus, the wind environment at pedestrian level is improved significantly. Moreover, the lift-up area can be utilised as a passageway for pedestrians, recreational zone, and parking zone.

Several studies have investigated the effect of the lift-up design on improving low wind velocity environment in high-density cities. They reported a clear correlation between wind speed enhancement and the implementation of lift-up design, and suggested that it was necessary to adopt lift-up design in urban planning and architectural design. A field Table 1 New wind comfort criteria for Hong Kong .

Wind comfort criteria for cold seasons Category Threshold velocity Exceedance probability Activity description Category Threshold velocity Exceedance probability Activity description Flow rate (Q) Calculated as the ratio between the flow entering the target area and the "free" flow rate far upstream of the same area in windward opening. A small Q value means that the domain has a low ventilation rate.

Generic/3D/Circular block with 2 or 4 sectors (Skote, Sandberg, Westerberg, Claesson, & Johansson, 2005 ) Generic/3D/Parallel buildings (Blocken, Carmeliet, & Stathopoulos, 2007 ) Generic/3D/Round shaped, Square, rectangular city model (Hang, Li, Sandberg, & Claesson, 2010 , 2009a Generic/3D/Courtyard (Moonen, Dorer, & Carmeliet, 2011) Purging Flow Rate (PFR)

The net flow rate needed to remove pollutants. A small PFR value means that the domain has limited airflow movement.

Generic/3D/Two-building model and building arrays (Bady, Kato, & Huang, 2008) Generic/3D/High-rise square arrays Generic/3D/Aligned arrays, semiopen street roof (Hang et al., 2013) Generic/3D/Aligned and staggered building arrays (Lin, Hang, Li, Luo, & Sandberg, 2014 ) Generic/3D/Street void model (Bu, Kato, Ishida, & Huang, 2009 ) Air change rate (ACH)

The value of air change in volume per hour for the domain. A small ACH value means that the domain has limited airflow movement.

Generic/2D/Street canyon (Li, Liu, & Leung, 2005 ) Generic/2D/Street canyon Generic/3D/Long street canyon (Hang et al., 2010) Generic/3D/Aligned arrays of cubes (Hang & Li, 2010a,b) Local mean age of air (τp)

The time required for a portion of the airflow to get to a specified point after entering the domain. A small τ p value means that the domain is poorly ventilated.

Generic/3D/Round, square, long rectangular city model (Hang et al., 2009a) Generic/3D/Long street canyon ) Generic/3D/Aligned arrays, semiopen street roof (Hang et al., 2013) Generic/3D/Building array of cubes (Buccolieri et al., 2010) Y. Du, C.M. Mak Sustainable Cities and Society 42 (2018) 314-324 measurement study in the HKPolyU campus found that wind speed was amplified by the lift-up design in a local precinct . Xia, Liu, Niu, and Kwok, (2015) assessed the influence of the lift-up design on the pedestrian level wind environment using wind tunnel measurement, and found that integrating the lift-up design in existing building configurations can enhance the wind speed around buildings. In a study of Du, Mak, Liu et al. (2017) , the effects of the lift-up design on pedestrian level wind environment in different building configurations, namely the slab shaped building, the corner shaped building, the semi-open shaped building and the close shaped buildings, were quantitatively evaluated using CFD simulation. The results showed that the lift-up design could effectively improve the low wind velocity environment and its enhancement highly relied on the approaching wind direction. Tse et al. (2017) found that the lift-up core dimensions govern the pedestrian level wind speed around an isolated building using wind tunnel measurement, and suggested that proper core dimensions should be selected before implementing the lift-up design in architectural design. Similarly, Du, Mak, and Li (2018) proposed a multi-variable optimization method to determine the appropriate liftup dimensions for an optimum wind environment at pedestrian level using CFD simulation. Liu et al. (2016 Liu et al. ( , 2017 studied the pedestrian level wind environment around isolated buildings with and without the lift-up design using CFD simulation, and found that the lift-up design could improve the wind speed around isolated buildings, but the effects were only limited to the neighbouring area. The above mentioned studies confirm that the lift-up design is a promising solution for low wind velocity environments in densely built urban areas, especially at pedestrian level.

The arcade design, known as "semi-open spaces", in which an outcorridor is formed from a line of arches on the side of a main building, as shown in Fig. 4 . It is frequently utilised as a pathway for pedestrians because it can provide shelter from direct solar radiation and rain.

A number of studies have been conducted about the effect of the arcade design on improving the low wind velocity environment in densely built-up areas. The CFD simulation study conducted by Hang, Luo, Sandberg, and Gong (2013) showed a direct relationship between the arcade roof designs and the natural ventilation characteristics in two ideal urban models with different wind directions. Wen, Juan, and Yang (2017) studied the effectiveness of the arcade design on the improvement of flow aeration in ideal urban canyons using CFD simulation, and concluded that the arcade design had the ability to improve low wind velocity environments at a city scale. Juan, Yang, Wen, Lee, and Wang (2017) optimized the wind environment in real urban streets by manipulating the configurations of the arcade design in CFD simulation. The results indicated that a wider arcade provided a better wind environment around buildings in a high-density city, which corresponds with the results obtained by Du, Mak, Li et al. (2018) . It can be concluded from the above studies that the arcade design has the potential to improve the low wind velocity environment in high-rise urban cities with proper arrangement.

The building opening design is a solution that enhances wind speed by increasing the permeability of buildings. Fig. 5 shows a schematic diagram of a building opening and a photo of the opening design utilised in a real building. Based on the CFD simulation results, Yuan and Ng (2012) concluded that building openings could enhance wind speed in compact urban canyons without losing land use efficiency. Their study also suggested that a higher pedestrian level wind velocity could be achieved when the opening was closer to the ground level and when the size of the opening was larger. In a study conducted by Fan et al. (2017) , different types of building openings inside canyons were investigated using CFD simulation. The study concluded that building openings could effectively improve the low wind velocity environment, and considering the values in the development floor area, an opening value of 10% was enough to improve the low wind velocity environment. Moreover, the CFD simulation studies by Hang and Li (2010a) , Hang and Li (2011) indicated that the airflow ventilation rate could be improved significantly when buildings in a compact array were openbased (have openings underneath the building). Thus, the building opening design presents prominent advantage in improving the low wind velocity environments in compact high-rise cities.

Aside from the specific architectural designs that can be integrated into building configurations, building disposition has been proven to improve pedestrian level wind environment around compact and tall buildings. To improve the low wind velocity environment, strategies including low building area density (2011, Buccolieri, Sandberg, & Di Sabatino, 2010; Hang & Li, 2010a) , low building aspect ratio Ho, Liu, & Wong, 2015; Liu, Leung, & Barth, 2005) , non-uniform building height (Gu, Zhang, Cheng, & Lee, 2011; Hang & Li, 2010a; Nelson, Pardyjak, Klewicki, Pol, & Brown, 2007) , street orientation (Hang, Sandberg, & Li, 2009 , b, Ramponi, Blocken, de Coo, & Janssen, 2015 and aligned arrays (Coceal, Thomas, Castro, & Belcher, 2006; Hanna, Tehranian, Carissimo, Macdonald, & Lohner, 2002) have been used in precinct planning. The key findings of the building arrangements and their contribution towards improving low wind velocity environments are summarized in Table 3 .

It is noteworthy that the above mentioned measures are passive strategies to improve low wind velocity environments in high-density cities. In studies conducted by Mirzaei and Haghighat (2010) , Mirzaei and Haghighat (2011) , an active pedestrian level ventilation system was proposed. The basic idea of this system was that it induced airflows from the roof of the building through ventilation ducts to pedestrian level in the street canyon. The performance of the ventilation system under different atmospheric conditions was numerically studied using CFD simulations of stable, neutral, and unstable conditions. Additionally, different ventilation strategies had been proposed for the ventilation system. Their investigation results indicated the application feasibility of the pedestrian level ventilation system. More detailed information on this active pedestrian ventilation system can be found in the studies of Mirzaei and Haghighat (2010) ; Mirzaei and Haghighat (2011) .

From the above the above literature review, the effectiveness of different improvement measures can be summarized as follows: (i) the building dispersion can improve wind environment in the whole city, and the pedestrian level wind environment can be affected subsequently. However, this improvement measure may lead to low land use efficiency and it is impractical for the existing urban areas. (ii) The arcade design and the lift-up design can enhance wind ventilation at pedestrian level directly without sacrificing any land use, but the result of the arcade design is less effective than the lift-up design. (iii) The position and size of the building opening can affect the effectiveness of the improvement. (iv) The active strategy of the pedestrian level ventilation system is more controllable than the above passive building design, but it requires energy consumption. Thus, the passive strategy should be preferred over active strategy for the sustainability of city development.

The HKPolyU campus was selected as the study area in this study. The campus is located in the downtown area of Hong Kong, which is closely surrounded by tall and dense buildings. Despite the meteorological advantage of being adjacent to the Victoria Harbour, the pedestrian level wind environment in the campus is low due to the compact neighbouring buildings. To accurately reproduce the wind environment in the campus, surrounding buildings that were up to 1 km from the campus were also included in the model. The campus building information was obtained from architectural drawings, and the geographical data of surrounding buildings were acquired from government department.

To obtain the pedestrian level wind environment in the campus, the wind tunnel test approach was utilised in this study. The CLP Power Wind/Wave Tunnel Facility (WWTF) at the Hong Kong University of Science and Technology (HKUST) were used to conduct the wind tunnel tests. The scaled ratio for the campus model with its surroundings was 1:200, and the campus model was reproduced in great detail (see Fig. 6(a) ). 16 wind directions were measured during the wind tunnel tests at an interval of 22.5°from 0°(North) to 360°. A reference sensor was located at 1 m high (200 m in full-scale) above the edge of the turn table on the windward side. The approaching wind profiles were acquired by grouping and fitting from previous topographic study, which could represent local wind conditions in real life. The measured and target values of the wind profiles and their corresponding approaching wind directions are presented in Fig. 6(b) , where the mean wind velocity at the reference height (200 m in prototype) was around 8 m/s. The measured errors were within 5%, which confirmed the confidence in the measurement data. The similarity requirements were strictly examined during the wind tunnel tests, and the blockage ratio was 4.5% with Reynolds Numbers (Re) over × 7 10 4 . These met the requirements of the quality wind tunnel test standards ASCE (1999) and AWES (2001) manuals). The Kanomax velocity sensors were utilised for the measurements of pedestrian level wind environment, and they were Y. Du, C.M. Mak Sustainable Cities and Society 42 (2018) 314-324 evenly distributed in the campus model. Moreover, the Kanomax velocity sensors were carefully calibrated before the wind tunnel measurements.

As indicated in Section 3.2, the new wind comfort criteria proposed by were developed specifically for low wind velocity environment. In particular, the parameters in the new wind comfort criteria were chosen explicitly to adapt to the low wind velocity environment in Hong Kong. Moreover, the new wind comfort criteria were established seasonally in consideration of the subtropical weather conditions in Hong Kong. Thus, the proposed new wind comfort criteria are adopted as the evaluation criteria in this case study. A detailed description of the new wind comfort criteria is presented in Section 3.2. With respect to the fact that the hot and humid summer is the critical season in Hong Kong, only the wind comfort for summer is considered in this study. It should be mentioned that the mean wind velocity at 200 m (reference height) in Hong Kong is 5 m/s in the summer season (Planning Department of the HKSAR, 2018). The wind comfort levels and their corresponding threshold values for the HKPolyU campus are presented in Table 4 .

The local wind climate for the HKPolyU campus was acquired by calibrating hourly wind data from 2013 to 2015 from a nearby weather station, namely the King's Park Meteorological Station of the Hong Kong Observatory. Detailed information on the meteorological station can be found in the website (Observatory, 19 December 2012). The wind rose in summer for HKPolyU campus is presented in Observatory (2018) Fig. 7 .

Studies on the building arrangements and improvements of low wind velocity environments.

Model configuration Results Reference

Low building area density Generic/3D/ Aligned square building arrays

The city model with lower building area density had faster wind flow movement inside street canyons than that with higher building area density. (Hang & Li, 2011) Generic/3D/ Aligned square building arrays

Lower building density provides a better wind environment in both main streets and secondary streets, especially in downstream regions. (Hang & Li, 2010a) Generic/3D/ Cubical buildings arrays

The wind environment inside the city model worsened with increasing packing density, and the model with low packing density had a good wind environment. (Buccolieri et al., 2010) Low building aspect ratio Generic/3D/ Long street model The larger the value of the street aspect ratio was, the weaker of the wind environment inside the street model was when the approaching wind was parallel to the street model. Generic/2D/ Street canyon model A smaller street aspect ratio presented lower aerodynamic resistance to the approaching wind, and thus the wind environment was better than that of the model with a larger street aspect ratio. (Ho et al., 2015) Generic/2D/ Street canyon model A smaller street aspect ratio had better air ventilation performance and a greater pollutant removal rate than that of the model with a larger street aspect ratio. Non-uniform building height Generic/3D/ Square building array Larger standard deviations in building height variability could lead to better pedestrian level ventilation. Generic/3D/ Aligned square building array Building height variation improved the low wind velocity environment in secondary streets. (Hang & Li, 2010a) Generic/3D/ Street canyon Large-scale wind flow change was found in non-uniform street canyons, which in turn improved the wind environment inside street canyons. (Gu et al., 2011) Actual street canyon/3D/ Actual street canyon Field measurements in an actual street canyon indicated that heterogeneous building heights could enhance wind speed by means of horizontal transport and enhanced turbulence. (Nelson et al., 2007) Street orientation Generic/3D/ Round-shaped model

For the round-shaped city model, an oblique approaching wind to the main street produced a better wind environment than that with a parallel approaching wind direction. (Hang et al., 2009a, b) Generic/3D/ Generic urban model

The wind environment was better with oblique and perpendicular approaching wind directions to the main street than that with a parallel approaching wind direction. (Ramponi et al., 2015) Aligned array Generic/3D/ Cubical obstacle Staggered square building arrays had a stronger form drag than aligned arrays. (Coceal et al., 2006 ) Generic/3D/ Simple obstacle array Higher wind speed was found in between the building obstacles in aligned arrays compared to that in staggered arrays. (Hanna et al., 2002) Y. Du, C.M. Mak Sustainable Cities and Society 42 (2018) 314-324 

To facilitate the illustration of the developed framework, the lift-up areas in the campus model were firstly blocked during the wind tunnel tests, as shown in Fig. 8 . This means that there was no lift-up design in the campus model for this case (preliminary case).

The evaluation results of wind comfort in summer for the preliminary case are presented in Fig. 9 . It can be seen that in the central part of the campus, the areas around building Y and the area between core A and core B had higher OMVR values than the other parts of the campus. Thus, these areas had acceptable wind comfort. Besides, the majority of the campus had low OMVR values, which lead to unfavourable wind comfort. Furthermore, nearly 80% of the spaces in the campus were unfavourable for any pedestrian activity. In general, the wind comfort in the campus was not suitable for pedestrian activities in hot and humid summers.

Taking the sustainability of the city development into consideration, the passive building designs are used in this case study. As concluded in Section 4.6, the arcade design and the lift-up design can both improve the pedestrian level wind environment directly, but the lift-up design is more effective than the arcade design. Thus, the lift-up design is adopted as the improvement measure in this case study. Buildings that adopted the lift-up design are indicated on the campus map by blue dashed lines in Fig. 10 . Fig. 11 presents the wind comfort evaluation results after adopting the lift-up design in the campus. As shown in Fig. 11 , the central parts of the campus had higher wind speeds than other parts of the campus, and most of the lift-up area had acceptable wind comfort. For the podium area in the campus, the area of acceptable wind comfort was larger than that in the preliminary case. Overall, less than 60% of the spaces in the campus were unfavourable for pedestrian activities. Compared to the preliminary case, the wind comfort in the campus was greatly improved after adopting the lift-up design. However, large areas in the campus still had unfavourable wind comfort.

The opening design is also effective in improving pedestrian level wind environment. As summarized in Section 4.6, the location and size of the building opening are important in improving pedestrian level wind environment. Thus, the opening design is adopted in building Y because the building Y is the highest and slab-shaped building in the campus. Since the building Y is on the edge of the campus, the opening design is located in the centre of the building Y. Fig. 12 shows the photo of the opening design in building Y during wind tunnel tests. Note that the dimensions in Fig. 12 are in prototype. Fig. 13 shows wind comfort evaluation results after adopting the lift-up design and the building opening in the campus model. Compared to Fig. 11 , the wind comfort of the areas between building M and building Y had significantly improvement after adopting the opening design in building Y. Besides, more than 50% of the places in the campus had acceptable wind comfort, which suggested satisfactory wind comfort for pedestrian activities. Thus, the wind comfort in the campus for summer was significantly improved after adopting lift-up design and opening design. Considering that the majority areas in the campus had acceptable wind comfort, the wind comfort in the campus could be viewed as acceptable in general.

To quantitatively assess wind comfort by adopting the above mentioned improvement measures, this section presents the results of OMVR in three cases. The results of OMVR in three cases are given in Fig. 14. It can be seen that when the campus had no improvement measures (preliminary case), nearly 80% of OMVR values were below 0.3, which meant that nearly 80% of the areas in the campus had unfavourable wind comfort. After adopting lift-up design in the campus, the values of OMVR had been greatly improved. However, half of the OMVR values were still below 0.3. Nevertheless, by adding opening in Y building, the values of OMVR were further improved and more than half of the OMVR values were over 0.3. This suggested that more than half of the areas in the campus had acceptable wind comfort.

To stress the importance of improving low wind velocity environments in high-density cities, the influences of low wind velocity environments on pollutant dispersion, thermal comfort, and wind comfort were presented in this paper. A general design framework was developed, and detailed guidelines on improving the pedestrian level low wind velocity environment in high-density cities were offered. The general design framework was presented in the form of a flowchart, which was systematic and effective. Evaluation criteria that aim to improve the low wind velocity environment were also reviewed in detail, namely the AVA scheme, the new wind comfort criteria for Hong Kong, and city ventilation parameters. Moreover, the strategies used to improve low wind velocity environments were reviewed in detail, including the lift-up design, the arcade design, the building opening, building disposition, and the active pedestrian ventilation system.

A case study of an existing university campus was used to investigate the performance of the proposed design framework. Wind tunnel tests were conducted to obtain the pedestrian level wind environment in the campus model. The choice of the evaluation criteria were chosen based on the review results presented in Section 3, and the new wind comfort criteria were selected. The wind comfort of the campus model was improved by vigorously following the proposed design framework, and the improvement measures were chosen based on the review results presented in Section 4. Specifically, three distinctive cases were examined, namely the campus model without the lift-up design, the campus model with the lift-up design, and the campus model with an opening in building Y. Results showed that the proposed design framework could help in improve the low wind velocity environment at pedestrian level. This study mainly focused on improving congested wind flow caused by tall and dense building arrays. The design framework and guidelines described in this paper were established under the presumption that low wind velocity environments have resulted in serious issues for urban residents. However, this design framework can also be applied to other wind environments, such as strong wind conditions in sparsely populated cities. This can be achieved by replacing the improvement measures and evaluation criteria with the current measures and criteria. Moreover, this study indicated that favourable wind environments in local precincts could be achieved by integrating innovative building designs.

The low wind velocity environment can be improved by following the proposed design framework, which also corresponds with government policies of enhancing ventilation in high-rise cities. In particular, the evaluation criteria and the improvement measures reviewed in this study can be used in practical developments. For example, the lift-up design could be used to improve the pedestrian level wind environment for an existing urban area. These findings can help urban planners and policy makers improve low wind velocity environments at pedestrian level, and build a sustainable city.

Large eddy simulation of wind-induced interunit dispersion around multistory buildings

From street canyon microclimate to indoor environmental quality in naturally ventilated urban buildings: Issues and possibilities for improvement

Determination of single-sided ventilation rates in multistory buildings: Evaluation of methods

Modeling of coupled urban wind flow and indoor air flow on a high-density near-wall mesh: Sensitivity analyses and case study for single-sided ventilation

The effect of personal and microclimatic variables on outdoor thermal comfort: A field study in Tehran in cold season

American society of civil engineers (ASCE) manuals and reports on engineering practice No

Assurance manual for wind-engineering studies of buildings (AWES-QAM-1)

Towards the application of indoor ventilation efficiency indices to evaluate the air quality of urban areas

CFD evaluation of wind speed conditions in passages between parallel buildings-Effect of wall-function roughness modifications for the atmospheric boundary layer flow

CFD simulation for pedestrian wind comfort and wind safety in urban areas: General decision framework and case study for the Eindhoven University campus. Environmental Modelling & Software

Pedestrian-level wind conditions around buildings: Review of wind-tunnel and CFD techniques and their accuracy for wind comfort assessment

New criteria for assessing local wind environment at pedestrian level based on exceedance probability analysis

City breathability and its link to pollutant concentration distribution within urban-like geometries

Microclimate design for open spaces: Ranking urban design effects on pedestrian thermal comfort in summer

Street canyon design and improvement potential for urban open spaces; the influence of canyon aspect ratio and orientation on microclimate and outdoor comfort

Outdoor thermal comfort and outdoor activities: A review of research in the past decade

Outdoor thermal comfort study in a subtropical climate: A longitudinal study based in Hong Kong

Mean flow and turbulence statistics over groups of urban-like cubical obstacles

Modelling of pedestrian level wind environment on a high-quality mesh: A case study for the HKPolyU campus

Towards an integrated method to assess effects of lift-up design on outdoor thermal comfort in Hong Kong

New criteria for assessing low wind environment at pedestrian level in Hong Kong

Application of a multi-variable optimization method to determine lift-up design for optimum wind comfort

Effects of lift-up design on pedestrian level wind comfort in different building configurations under three wind directions

Thermal perception of outdoor urban spaces in the hot arid region of Cairo

Urban microclimate: Designing the spaces between buildings

A decision support tool for evaluating the air quality and wind comfort induced by different opening configurations for buildings in canyons

Best practice guideline for the CFD simulation of flows in the urban environment

Effect of uneven building layout on air flow and pollutant dispersion in non-uniform street canyons

Age of air and air exchange efficiency in high-rise urban areas and its link to pollutant dilution

Ventilation strategy and air change rates in idealized high-rise compact urban areas

Wind conditions in idealized building clusters: Macroscopic simulations using a porous turbulence model

Wind conditions and ventilation in high-rise long street models

Natural ventilation assessment in typical open and semi-open urban environments under various wind directions

Age of air and air exchange efficiency in idealized city models

Effect of urban morphology on wind condition in idealized city models

On the contribution of mean flow and turbulence to city breathability: The case of long streets with tall buildings

The influence of building height variability on pollutant dispersion and pedestrian ventilation in idealized high-rise urban areas

Comparisons of model simulations with observations of mean flow and turbulence within simple obstacle arrays

Preliminary study of the parameterisation of street-level ventilation in idealised two-dimensional simulations

The effects of wind on people; new criteria based on wind tunnel experiments

Urban microclimate analysis with consideration of local ambient temperature, external heat gain, urban ventilation, and outdoor thermal comfort in the tropics

The ground level wind environment in built-up areas

The use of measurement programmes and models to assess concentrations next to major roads in urban areas

Optimization procedures for enhancement of city breathability using arcade design in a realistic high-rise urban area

A numerical study of the effects of ambient wind direction on flow and dispersion in urban street canyons using the RNG k-ε turbulence model

Regulation of outdoor thermal comfort by trees in Hong Kong

Wind tunnel tests on the relationship between building density and pedestrian-level wind velocity

Development of guidelines for realizing acceptable wind environment in residential neighborhoods

The widn content of the built environment

Feasibility study of a PCM storage tank integrated heating system for outdoor swimming pools during the winter season

Multi-zone modeling of probable SARS virus transmission by airflow between flats in Block E

Development of a model for the determination of air exchange rates for street canyons

Quantitative ventilation assessments of idealized urban canopy layers with various urban layouts and the same building packing density

On the prediction of air and pollutant exchange rates in street canyons of different aspect ratios using large-eddy simulation

Combining measured thermal parameters and simulated wind velocity to predict outdoor thermal comfort

Criteria for environmental wind conditions

Effects of building aspect ratio and wind speed on air temperatures in urban-like street canyons

A novel approach to enhance outdoor air quality: Pedestrian ventilation system

Pollution removal effectiveness of the pedestrian ventilation system

Evaluation of the ventilation potential of courtyards and urban street canyons using RANS and LES

Flow field and pollution dispersion in a central London street

Properties of the wind field within the Oklahoma City Park Avenue Street Canyon. Part I: Mean flow and turbulence statistics

Policies and technical guidelines for urban planning of high-density cities -Air ventilation assessment (AVA) of Hong Kong

A new method to assess spatial variations of outdoor thermal comfort: Onsite monitoring results and implications for precinct planning

Urban Heat Island (UHI) mitigating strategies: A case-based comparative analysis

Observatory King's park meteorological station

Microclimatic modeling of the urban thermal environment of Singapore to mitigate urban heat island

CFD simulation of outdoor ventilation of generic urban configurations with different urban densities and equal and unequal street widths

Issues with validation of urban flow and dispersion CFD models

Some remarks on the validation of small-scale dispersion models with field and laboratory data

Numerical and experimental studies of wind environment in an urban morphology

Comparison of various revised k-ε models and LES applied to flow around a high-rise building model with 1: 1: 2 shape placed within the surface boundary layer

AIJ guidelines for practical applications of CFD to pedestrian wind environment around buildings

Wind tunnel study of pedestrian level wind environment around tall buildings: Effects of building dimensions, separation and podium

Adopting 'lift-up' building design to improve the surrounding pedestrianlevel wind environment

Comparison of air quality in streets of Copenhagen and Milan, in view of the climatological conditions

Enhancement of city breathability with half open spaces in ideal urban street canyons

Design for wind comfort in the Netherlands: Procedures, criteria and open research issues

Effects of building lift-up design on the wind environment for pedestrians

Thermal conditions and ventilation in an ideal city model of Hong Kong

Thermal comfort in outdoor urban spaces in Singapore

Building porosity for better urban ventilation in high-density cities -A computational parametric study

The work described in this paper was fully supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. C5002-14 G).