Characterization of different heat mitigation strategies in landscape to fight against heat islandand improve thermal comfort in hot–humid climate (Part I)Measurement and modellingZhao, T.F.; Fong, K.F.

Published in:Sustainable Cities and Society

Published: 01/07/2017

Document Version:Post-print, also known as Accepted Author Manuscript, Peer-reviewed or Author Final version

License:CC BY-NC-ND

Publication record in CityU Scholars:Go to record

Published version (DOI):10.1016/j.scs.2017.03.025

Publication details:Zhao, T. F., & Fong, K. F. (2017). Characterization of different heat mitigation strategies in landscape to fightagainst heat island and improve thermal comfort in hot–humid climate (Part I): Measurement and modelling.Sustainable Cities and Society, 32, 523-531. https://doi.org/10.1016/j.scs.2017.03.025

Citing this paperPlease note that where the full-text provided on CityU Scholars is the Post-print version (also known as Accepted AuthorManuscript, Peer-reviewed or Author Final version), it may differ from the Final Published version. When citing, ensure thatyou check and use the publisher's definitive version for pagination and other details.

General rightsCopyright for the publications made accessible via the CityU Scholars portal is retained by the author(s) and/or othercopyright owners and it is a condition of accessing these publications that users recognise and abide by the legalrequirements associated with these rights. Users may not further distribute the material or use it for any profit-making activityor commercial gain.Publisher permissionPermission for previously published items are in accordance with publisher's copyright policies sourced from the SHERPARoMEO database. Links to full text versions (either Published or Post-print) are only available if corresponding publishersallow open access.

Take down policyContact lbscholars@cityu.edu.hk if you believe that this document breaches copyright and provide us with details. We willremove access to the work immediately and investigate your claim.

Download date: 23/05/2022

1

Characterization of different heat mitigation strategies in landscape to fight against heat island

and improve thermal comfort in hot-humid climate (Part I): Measurement and modeling

T.F. Zhao1, K.F. Fong

2*

1Department of Architecture and Civil Engineering, College of Science and Engineering, City

University of Hong Kong

2Division of Building Science and Technology, College of Science and Engineering, City University

of Hong Kong

*Corresponding author

Email address: bssquare@cityu.edu.hk

Postal address: 8 Tat Chee Avenue, Kowloon Tong, Hong Kong, China

http://dx.doi.org/10.1016/j.scs.2017.03.025

Abstract

This study aims to consolidate and characterize the cooling potentials of heat mitigation strategies

(HMSs) in landscape from both combating urban heat island (UHI) and alleviating human heat stress.

This is Part I of the study, which focuses on measurement and modeling of all the possible HMSs

(vegetation, albedo and water body) in landscape. The field measurement included ground level

greening, different albedo pavements and water pools, which can be categorized into green, grey and

blue areas. Through the microclimate simulation tool ENVI-met V4, a landscape model was

developed and its performance was assessed. It was found that the variations of air temperature and

mean radiant temperature within a campus landscape would be large, up to 4.5 ºC and 32 ºC

respectively. The green and blue areas showed certain heat mitigation potential, but the grey areas

had the possibility to generate uncomfortable thermal environment. The close relationship between

microclimate and different HMSs in landscape was identified. The landscape model was capable of

2

reproducing the main features of temporal and spatial distributions of intra-urban microclimate.

Hence, the validated settings of such landscape model would be passed on for prediction of

mitigation potentials of different HMSs in Part II of this study.

Keywords: Heat mitigation strategy; Urban heat island; Outdoor thermal comfort; Field

measurement; Landscape modeling; Hot and humid climate.

1. Introduction

Urban heat island (UHI) phenomenon has a clear impact on local climate, which results in higher air

temperatures in dense urban areas compared to their rural surroundings. Daily mean UHI would

typically range from 2 to 5 C [1]. In Hong Kong, some urban areas are experiencing an UHI of 5 C

[2]. Severe urban living environment demonstrates strong demand to fight against UHI and improve

human thermal comfort by appropriate heat mitigation strategy (HMS). Via the interplay of design

creativity and scientific knowledge on the natural elements, landscape approach as a soft technology

has been received growing research interest [3,4]. In the landscape scale, three heat mitigation

strategies (HMSs) have been commonly acknowledged as useful measures [5,6]: Vegetation, high

albedo and water body.

Vegetation, i.e. the green space, can moderate climate mainly through the combined effect of (a)

shading, which lowers the air and surface temperatures through interception of incoming solar

radiation; (b) evapotranspiration, which requires significant amount of heat taken from surroundings

[6]. It has been found that urban greening contributes a lot to heat mitigation under various climates,

even in subtropical and tropical climatic conditions. Through field measurement and microclimate

modeling, Ng et al. [7] found that the pedestrian air temperature could be reduced by around 1 C

when the tree planting coverage area reached 33% of the site in Hong Kong. By the similar

methodology, Srivanit et al. [8] found that the average daily maximum temperature would be

decreased by 2.27 C at the pedestrian level when the quantity of trees was increased by 20% in a

3

university campus in Japan. Wong et al. [9] indicated that the peak temperature difference between

dense, less dense and sparse greenery area could be as high as 4 C in Singapore via field

measurement. However, most of the previous studies only focused on the cooling effect in air

temperature Ta. Its effect on human-biometeorological aspect has only been addressed in a few

studies [10,11]. Previous study demonstrates that green coverage ratio may be closely related to its

mitigation potential [6]. In addition, Givoni [12] claimed that the type and features of the plant can

affect cooling effect. Leaf area index (LAI) is a conceptual modeling parameter in studying plant’s

heat exchange with environment, and serves as a key measure in comparing plant canopies [13].

Kotzen [14] concluded that trees with low and high LAI could bring only 21% and 7% radiant heat

underneath the canopy, respectively.

High albedo, another heat mitigation measure, is a key thermal property and an important indicator

of the reflecting power of a surface. Corresponding to lower solar radiation absorption, high albedo

materials gains lower surface temperature, which would lead to lower ambient temperature through

the mechanism of convection [15]. A few studies are available about investigating the cooling

potential of pavement albedo increased. Takahashi [16] found that if high albedo pavements were

applied at the central area of Tokyo, the mean Ta might be reduced by 0.15 C and even down to 0.6

C by simulation. However, it might result in negative consequence on human thermal comfort.

Through an experiment, Taleghani et al. [17] concluded that increasing the albedo of pavement from

0.37 to 0.91 resulted in 1.3 C decrease of Ta, but 2.9 C rise of the mean radiant temperature Tmrt. In

fact, Tmrt is an important factor to determine thermal comfort [18]. The cooling benefit in reducing Ta

but negative impact on Tmrt would make an uncertain condition of thermal comfort. Indeed, besides

albedo, the coverage ratio of high albedo materials [15] may be another important factor for heat

mitigation.

Water body, a natural heat sink, has been used as a design tool by urban planners and architects to

regulate the urban temperature. The associated cooling benefit originates from the enhanced

evaporation of the water body during daytime and the high heat capacity [19]. Using the 5-year

4

climate data and the results of three measurement campaigns carried out in Bilbao, Acero et al. [20]

verified that rivers crossing urban areas were helpful in lowering air temperatures. Moreover, with

the aid of advanced statistics of weather data and geographical information system (GIS), Radhi et al.

[21] found that lack of water could result in 2-3 C increase of Ta in the city under the hot and arid

climate. Taleghani et al. [22] evaluated water installations as heat buffer in Netherland, using a water

pool embedded on 65% of the ground inside a courtyard. The simulated results by ENVI-met showed

that Tmrt was reduced in a range of 18 C to 21 C inside the courtyard with different orientations.

However, Steeneveld et al. [23] had an opposite view. Through analysis of weather observations,

they concluded that the water body increased 95% of the daily maximum UHI and might exert

negative influence on human thermal comfort.

Through the literature review, it is found that HMSs based on landscape design were generally

evaluated in terms of combating heat islands, not commonly from the human-biometeorological

perspective. The human perception of heat and condition of thermal comfort are governed by the

integral effect of thermo-physiological process, which are combined in the human heat balance. To

date, many integrative thermal indices for thermal comfort assessment have been developed, and the

Physiological Equivalent Temperature PET has been widely used to evaluate the effect of urban

design on outdoor thermal comfort [24-28]. Since only Ta is generally applied to evaluate the cooling

benefit in the previous studies, PET should be introduced to evaluate the human-biometeorological

aspect. On the other hand, the cooling potential of HMS is generally evaluated in a singular

landscape design, like vegetation, albedo or water body only. An integrated implementation of

various designs for HMS has not been considered. Among those three common measures of HMS,

comparison of their effectiveness has not been carried out. As a result, this study would focus on the

deficiencies as identified in the literature review, and the primary research objective is to consolidate

and characterize the cooling potentials of various HMSs for landscape from both aspects of

combating heat island and alleviating human heat stress. Possible parameters governing the heat

mitigation potential of the strategies will be explored. Design insights and recommendations will be

proposed for heat mitigation of UHI and/or human heat stress for open space.

5

As the first part of this study, this paper mainly deals with the investigation of thermal performance

of different HMSs in landscape of real urban environment, and the evaluation of the model

performance. Hence, the field measurement and modelling were presented.

2. Methodology

The methods used in this study consist of field measurement and numerical simulation. Field

measurement was designed to collect the microclimate conditions and used for model validation. An

open space simulation model was then established to conduct parametric study of different HMSs,

which will be presented in Part II of this study.

2.1 Field measurement

Hong Kong is located in the southeast coast of China, with latitude of 22º15’ N and longitude of

114º10’ E. It has a monsoon-influenced humid-subtropical climate. City University of Hong Kong

(CityU) was selected as the field measurement site, since such an institutional campus can be

regarded as a small city due to its relatively large coverage. Its diverse landscape features and

complex microclimates offer a valuable opportunity to investigate the multiple HMSs of landscape

design. The observation points distribute around the spaces in various landscape forms among

building blocks. The selected measurement locations are presented in Figure 1, and the

corresponding landscape environments are described in Table 1.

6

L1 Garden 1 L2 Garden 2

L3 Garden 3 L4 Entrance 1

L5 Entrance 2 L6 Square CityU campus plan

L7 Entrance 3 L8 Entrance 4 L9 Garden 4 L10 Swimming Pool

Figure 1. Locations of field measurement in CityU campus.

Table 1. Major landscape environments of the field measurement locations.

Code Landscape environment Surface material Key element

L1 Dense greenery area Tall vegetation

Green L2 Average dense greenery area Tall vegetation

L3 Sparse greenery area Lawn

L4 High density hard surface area (open) Concrete texture

Grey

L5 High density hard surface area (open) Concrete pavement

L6 High density hard surface area (open) Granite pavement

L7 High density hard surface area (open) Interlocking cement brick

L8 High density hard surface area (enclosed) Concrete texture

L9 Water pond + dense greenery area Concrete with pebbles, water

and vegetation Blue

L10 Swimming pool Ceramic tile, water

7

The field measurements were carried out from 10:00 to 22:00 local time in August 2015. All the

measurements lasted for three days, and one day data were selected for analysis. Such day was

selected according to weather type over a 24-h period (i.e. Cloudy or Clear Days), data accuracy and

reliability. As a typical summer sunny day, the data collected on 5th

August was applied in this study.

Basic meteorological data are described in Figure 2, the hourly data of air temperature and relative

humidity were derived from the Kowloon City weather station (nearby CityU), whist the hourly solar

radiation, wind speed and direction data were retrieved from Hong Kong Observatory (HKO). It is

observed that the weather on a typical summer sunny day can be characterized by high temperature,

strong solar radiation and light wind conditions.

50

55

60

65

70

75

80

85

90

25

26

27

28

29

30

31

32

33

34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Rel

ativ

e hum

idity (

%)

Air

tem

per

ature

(ºC

)

Local time (h)

(a)

Temp RH

Figure 2. (a) Hourly air temperature and relative humidity, (b) Hourly global solar radiation, wind speed and

direction on 5th August.

8

2.1.1 Collection of meteorological data

For the meteorological data, dry bulb air temperature Ta, wet bulb air temperature Twa, globe

temperature Tg and wind speed V were collected. In the field measurement, there were two kinds of

arrangement: fixed and mobile measurements. Fixed measurement would be arranged as far as

possible, but for those locations with frequently travelling people, mobile measurement was adopted.

The equipment set-ups for the fixed and mobile measurements are shown in Figure 3, and the

technical specifications are described in Table 2. All the fixed and mobile measurement stations were

configured and calibrated as per user manuals. Moreover, all sets of equipment were placed at the

same location of outdoor environment to check their consistency and validity before use.

The measurement duration for fixed testing locations (L1, L6, L9 and L10) lasted for 12 hours, and

all data were collected at an interval of 1 s. The other six locations were covered by mobile

measurements with three identical sets of equipment. It was carried out 15 min per hour in each

location. Then it took 10 min for shifting equipment between the measurement points and 5 min for

stabilizing equipment before measurement. In such measurement period, the data of Twa and Tg were

recorded at 1-min interval, the others were kept at 1-s interval. Those instruments were installed at a

height of 1.5 m above ground, which is the representative pedestrian level under study.

9

Figure 3. The equipment set-ups for fixed and mobile measurements.

Table 2. Technical specifications of the equipment used for fixed and mobile measurements.

Location Parameter Equipment Range Accuracy

L1, L6, L9, L10

(Fixed)

Ta SWA03 10-40 °C ±0.5 C (10-40 C)

Casella Heat Stress Monitor 10-60 ˚C ±1 C

Twa Ditto 5-40˚C ±0.5 C

Tg Ditto 20-120˚C

±0.5 C (20-50 C)

±1.0 C (50-120 C)

V SWA03 0.05-3 m/s ±0.04 m/s (10-34 C)

BABUC M 0-50 m/s ±4%

L2, L3, L4, L5,

L7, L8

(Mobile)

Ta SWA03 10-40 °C ±0.5 °C (10-40 C)

Twa Psychrometer -5-50 °C ±2%

Tg TP875.1 Globe Thermometer -30-120 ˚C ±0.5 C

V SWA03 0.05-3 m/s ±0.04 m/s (10-34 C)

BABUC M 0-50 m/s ±4%

Tmrt was determined by the recorded data of Ta, V and Tg according to Eq. (1) of the globe

temperature method [29]:

𝑇𝑚𝑟𝑡 = [(𝑇𝑔 + 273.15)4 +1.1×108𝑉0.6

𝜀𝐷0.4× (𝑇𝑔 − 𝑇𝑎)]

0.25

− 273.15 (1)

where,

D = globe diameter (mm) (D is 44 mm and 150 mm for fixed and mobile measurement,

respectively.)

= globe emissivity ( is 0.95.)

10

2.1.2 Albedo measurement

As albedo is a key thermophysical property, measurement was carried out to determine its values of

the different landscape surfaces within the campus. Later, these measurement results of albedo would

be adopted in model development. The albedo values of main types of landscape surfaces within the

campus are tested and summarized in Table 3. Testing of albedo was conducted in accordance with

ASTM E1918 [30] by using dual Kipp-Zonen CMP6 pyranometers as shown in Figure 4. The

measurement was carried out on a clear sunny day (6th

August) from 12:00 to 14:00. The outputs of

the pyranometers were recorded on two Fluke 77 Multimeters.

Table 3. Main surface materials under assessment in CityU campus.

Tall vegetation Lawn Concrete with pebbles

Concrete texture Concrete pavement Granite pavement

Interlocking cement brick Asphalt concrete Ceramic tile

11

Figure 4. Dual pyranometers used for on-site measurement of surface albedo.

2.2 Numerical simulation

In order to have an in-depth study of the heat mitigation potentials of various HMSs, landscape

models were built with the microclimate simulation tool ENVI-met. ENVI-met is a

three-dimensional Computational Fluid Dynamics (CFD) modeling program that simulates

surface-plant-air interactions in urban environments [31]. In this study, the up-to-date ENVI-met V4

was used, with significant improvements to the previous versions. The primary merit is able to

simulate various meteorological conditions by forcing the model with a user-specified weather

profile, so that temporal variations can be accounted for. This means that hourly Ta and RH data

measured in a meteorological station, which is adjacent to the simulation site, can be included.

Therefore, more realistic simulation results can be achieved [32,33]. ENVI-met was originally

developed for temperate climatic conditions, thus the simulation in hot-humid environments should

be validated. As such, a campus model would be built and validated by field measurement data in

this study.

2.2.1 Model configuration

Figure 5 illustrates the model of the CityU campus and the surrounding environment. The model

12

domain was constructed based on the campus map. Whist the surrounding area was built with

reference to the information obtained from the Lands Department of Hong Kong [34]. The land use

and ground cover were modeled within a radius of 15-17.5 m around each measurement location to

represent the environmental characteristics within the area [35]. The main ground surface materials

are listed in Table 3. Default material properties in ENVI-met were adopted, but the corresponding

albedo values were obtained by measurement. The region of 150-m radius from the centre of the

campus was established to consider the territorial urbanization effect according to Oke [36]. The

boundary conditions and initial settings are presented in Table 4.

.

Figure 5. Model of CityU campus and surrounding environment.

13

Table 4. Boundary conditions and initial settings of campus modeling in ENVI-met V4.

Type Item Unit Value/input

Model domain Size of grid cells (Δx, ∆y, ∆z) m 3, 3, 0.6 (vertical grid with a

telescoping factor of 20%

above the height of 3 m)

Number of grid cells (Δx, ∆y, ∆z) - 230, 175, 30

Nesting grids - 5

Model rotation out of grid north ºN -136

Soil profiles in nesting grids - Concrete pavement

Meteorological

inputs

Air temperature ºC Hourly data from weather

station nearby CityU

Relative humidity % Ditto

Wind speed at 10 m m/s 1.24

Wind direction ºN 90 (East)

Roughness length at reference point m 0.1

Cloud cover octans 1

Specific humidity at 2,500 m g/kg 7

Solar adjust factor - 0.93

Soil data Initial Ta and RH of soil upper layer (1-20 cm) K

%

301.7

40

Initial Ta and RH of soil middle layer (20-50 cm) K

%

302.9

45

Initial Ta and RH of soil deep layer (50-200 cm) K

%

302

50

The physical properties of building and surface materials were mainly derived from the local

building design guidelines [37]. The vegetation elements fell into several categories: lawns, tall grass,

shrubs and trees. The tree or shrub cover areas were specified based on the virtually projected foliage

area on the horizontal floor. For the vegetation species, default values of LAI were used for lawn, tall

grass and shrubs. Trees of the measurement site were mainly Ficus Benjamina, Delonix Regia and

Maleleuca Leucadendron, which were regarded as high-, medium- and low-canopy-density trees,

respectively in this study. The corresponding LAI values mainly referred to [38,39].

About the weather data inputs, the ENVI-met default value of specific humidity at 2,500 m was

adopted. The roughness length was kept default since the field measurements were done in urban

areas. The input 24-h Ta and RH (used to force the model) were those depicted in Figure 2(a). Since

14

the temporal variations of wind speed and direction could not be considered, it was necessary to use

the adjusted median value of the hourly data between 10:00 and 22:00 for input [27]. Since

ENVI-met required wind speed at 10 m, the data obtained from HKO (42 m above ground level)

were adjusted according to the power law for wind profile as recommended by [40,41].

𝑊𝑆10 = 𝑊𝑆ℎ(10/ℎ)0.35 (2)

where,

WSh = the wind speed (m/s) at the height of h

The incoming solar radiation was calculated by ENVI-met based on latitude, longitude, date, time

and cloud cover. Then the solar factor was derived by comparing the solar radiation at 13:00 from

simulation and that from HKO. The cloud cover value at 13:00 and initial soil temperature profiles

were also obtained from HKO [42]. Due to lack of information of soil moisture, estimation was made

for the initial soil moisture profile. As recommended by ENVI-met, the simulation should be started

before sunrise, and the total running hour should be longer than 6 h to overcome the influence of the

initialization. Therefore the simulation run started at 6:00 on 5th

August and ended at 8:00 on 6th

August, totally 26 hours.

2.2.2 Model validation

The parameters Ta, RH, V, Tmrt, and PET were used in validation of the developed landscape model.

ENVI-met V4 would give a good approximation of Tmrt for each grid point [43]. PET was calculated

using RayMan model [44], which follows the method described by Höppe [24]. It was estimated by

Ta, RH, V, Tmrt, human clothing and activity level in the model. The clothing and activity level were

assumed to be 0.9 clo and 80 W in RayMan model in this study, respectively.

To validate the simulated and measured results, besides the commonly used regression analysis, the

other statistical suites namely Mean Bias Error (MBE), Mean Absolute Error (MAE), and Root Mean

Square Error (RMSE) were also adopted [45]. R2

describes the proportion of the variance in

15

measured data explained by the model, and typically values greater than 0.5 are considered

acceptable [46]. A positive value of MBE indicates that the simulation results are higher than the

experimental data and vice versa. MAE measures how far predicted values are away from the

observed values. RMSE provides insights into the average difference between observation and

prediction. The hourly values of Ta, RH, Tmrt, and PET at a height of 1.5 m above ground in the 10

measurement locations were examined in the period of 10:00-22:00. Thus, the validation of each

parameter was based on 130 pairs of the observed and simulated values.

3. Results and discussion

3.1 Albedo of different surface materials

Within the campus, the albedo values of main types of landscape surfaces were tested and

summarized in Table 5. The surface material of concrete pavement, granite pavement and ceramic

tile displayed relative higher albedo values, whilst the interlocking cement brick and asphalt concrete

showed relative lower albedo values. Materials with low albedo values represented lower reflecting

power of a surface. Through solar radiation absorption, the surface temperature was more easily

increased to a point higher than the adjacent air temperature. It then intensifies the heat island effect

through convection.

16

Table 5. The measured values of albedo () for main surface materials in CityU campus.

3.2 Thermal performance of HMSs in landscape

Figure 6 depicts the thermal environment changed throughout the whole typical summer day (5th

August) in different locations. In Figure 6(a), the peak temperature difference between L4 (open

concrete surface area) and L1 (densely greenery area) could be more than 4.5 ºC at 14:00. This

showed that the difference of Ta within the landscape of a campus scale would be large, showing that

local heat islands existed. The high density hard surface areas created the hottest thermal

environment, followed by blue and green areas. As shown in Figure 6(b), in general, the fluctuation

trends of RH correlated well with the Ta, but demonstrated opposite trends of diurnal variation.

Diverse microclimates all tended towards humid conditions. In Figure 6(c), the Tmrt difference could

be even substantial, it reached about 32 ºC between L6 (open granite surface area) and L1 (densely

greenery area) at 14:00. Large Tmrt difference would result in great variance of thermal comfort. In

fact, the dense greenery area would help intercept large amount of radiant energy, and it contributed

much to the reduction of Tmrt. People in open high density hard surface environments would

experience substantial heat stress, followed by blue and green areas.

Surface type

Tall vegetation 0.20

Lawn 0.30

Concrete with pebbles 0.20

Concrete texture 0.30

Concrete pavement 0.36

Granite pavement 0.36

Interlocking cement brick 0.15

Asphalt concrete 0.10

Ceramic tile 0.40

17

28

29

30

31

32

33

34

35

36

37

38

10 11 12 13 14 15 16 17 18 19 20 21 22

Air

Tem

per

atur

e (º

C)

Local time (h)

(a) L1 L2 L3 L4 L5

L6 L7 L8 L9 L10

40

50

60

70

80

90

100

10 11 12 13 14 15 16 17 18 19 20 21 22

Rel

ativ

e H

umid

ity (

%)

Local time (h)

(b) L1 L2 L3 L4 L5

L6 L7 L8 L9 L10

20

25

30

35

40

45

50

55

60

65

70

10 11 12 13 14 15 16 17 18 19 20 21 22

Mea

n R

adia

nt

Tem

per

ature

(ºC

)

Local time (h)

(c)L1 L2 L3 L4 L5

L6 L7 L8 L9 L10

Figure 6. On-site field measurement results of diurnal profiles of (a) air temperature, (b) relative humidity, and

(c) mean radiant temperature at 1.5 m above ground in different locations.

18

3.3 Evaluation of the model performance

Figure 7 presents the relationship between the simulated and measured values, and Table 6

summarizes the quantative evaluation results of model performance. From both Figure 7 and Table 6,

good relationship can be observed between the modelled and measured data for Ta, RH, Tmrt and PET.

R2

value indicated a strong correlation between the simulated and the experimentally determined Ta

(R2=0.89), RH (R

2=0.77), Tmrt (R

2=0.88), and PET (R

2=0.89). It slightly underestimated Ta and RH,

with MBE of -0.04 ºC and -0.97%. But the discrepancies maintained in a small range, the calculated

MAE for Ta and RH were 0.56 ºC and 3.93%, whist the RMSE were constrined at 0.63 ºC and 4.49%,

respectively. Compared to the previous studies [11,27,32,33], these results had better agreement

since ENVI-met V4 enables to force climate variables (Ta, RH) along the simulation. It takes into

account the atmospheric temporal variations and consequently represents the profile change better

along the day. Although V did not have a high R2, it was only slightly underestimated as reflected by

MBE (-0.11 m/s). Actually the difference was very small, with MAE and RMSE of 0.21 and 0.26 m/s,

respectively, compared to [47,48]. RMSE was 28% of the mean observed V according to [49]. In fact,

ENVI-met V4 is unlikely to incorporate varing winds in simulation, i.e. the temporal variations of

wind speed and wind direction could not be considered.

On the other hand, Tmrt was overestimated, with MBE of 1.9 ºC, MAE of 5.07 ºC and RMSE of 5.7

ºC. In addition, PET was also overestimated, with MBE of 1.15 ºC, MAE of 2.7 ºC and RMSE of

3.04 ºC. The accuracy of PET is apparently determined by Ta, RH, V and Tmrt. In complex urban

environment, the current results regarding Tmrt and PET were believed to be reasonable with

reference to [11,27]. Despite some discrepancies between the measured and simulated results, the

campus model was capable of reproducing the main features of temporal and spatial distributions of

the intra-urban microclimate.

19

y = 0.83 x + 5.18 R² = 0.89

26

28

30

32

34

36

38

26 28 30 32 34 36 38

Ta-

Env

i-m

et (

ºC)

Ta-Measurement (ºC)

(a)

y = 0.55 x + 29.81 R² = 0.77

40

45

50

55

60

65

70

75

80

85

90

40 50 60 70 80 90

RH

-Envi-

met

(ºC

)

RH-Measurement (ºC)

(b)

y = 1.24 x - 7.24 R² = 0.88

20

30

40

50

60

70

80

20 30 40 50 60 70 80

Tm

rt-E

nvi-

met

(ºC

)

Tmrt-Measurement (ºC)

(c)

y = 1.22 x - 6.52 R² = 0.89

25

30

35

40

45

50

55

60

25 30 35 40 45 50 55 60

PE

T-E

nvi-

met

(ºC

)

PET-Measurement (ºC)

(d)

Figure 7. Regression analysis between the simulated and measured values of (a) Ta, (b) RH, (c) Tmrt and (d)

PET at 1.5 m above ground during 10:00-22:00 on 5th August.

Table 6. Quantative evaluation of model performance.

Index Ta (ºC) RH (%) V (m/s) Tmrt (ºC) PET (ºC)

R² 0.89 0.77 0.57 0.88 0.89

MBE -0.04 -0.97 -0.11 1.90 1.15

MAE 0.56 3.93 0.21 5.07 2.70

RMSE 0.63 4.49 0.26 5.70 3.04

4. Conclusions

On a whole, a series of locations representing various environmental conditions were selected to

conduct field measurements. They were dominated by distinct landscape features, which mainly

including ground level greening, pavement materials with varied albedos and water pools. The

albedo of different surface materials in CityU campus was measured.

20

Various environments all tended towards humid conditions. But the differences of Ta within the

landscape of a campus scale were large, as well as the Tmrt. The peak temperature difference between

L4 (open concrete surface area) and L1 (densely greenery area) reached more than 4.5 ºC at 14:00.

The Tmrt difference could be even substantial, which was about 32 ºC between L6 (open granite

surface area) and L1 (densely greenery area) at 14:00. The “hot” and “cool” locations in CityU

campus were identified, and the existence of local heat islands. The grey areas had the highest

possibility in uncomfortable thermal environment generation. In contrast, green and blue areas

demonstrated certain heat mitigation potential. Different landscape features with varied dominating

HMSs demonstrated distinct thermal performances. The relationship between microclimate and

diverse HMSs in landscape was verified.

With the measured albedo values of different ground surfaces and the collected microclimate

conditions, the landscape model built with the simulation tool ENVI-met was validated. Hence, the

settings of such landscape model would be passed on, and used to establish an open space model for

parametric analysis and prediction of mitigation potentials of different HMSs to be covered in Part II

of this study.

Acknowledgement

This work described in this paper is fully supported by a grant from the Research Grants Council of

the Hong Kong Special Administrative Region, China (Project No. CityU 123812).

References

[1] Santamouris, M. (2007). Heat island research in Europe: the state of the art. Advances in

Building Energy Research, 1(1), 123-150.

[2] Ng, E. (2009). Wind and heat environment in densely built urban areas in Hong Kong. Global

Environmental Research, 13(2), 169-178.

[3] Khalizah, S., Hanita, N. and Idilfitri, S. (2012). Modification of urban temperature in

21

hot-humid climate through landscape design approach: a review. Social and behavioural

sciences, 68, 439-450.

[4] Shahidan, M. F., Jones, P. J., Gwilliam, J. and Salleh, E. (2012). An evaluation of outdoor and

building environment cooling achieved through combination modification of trees with ground

materials. Building and Environment, 58, 245-257.

[5] Santamouris, M., Gaitani, N., Spanou, A., Saliari, M., Giannopoulou, K., Vasilakopoulou, K.

and Kardomateas, T. (2012). Using cool paving materials to improve microclimate of urban

areas-Design realization and results of the flisvos project. Building and Environment, 53,

128-136.

[6] Gago, E.J., Roldan, J., Pacheco-Torres, R. and Ordóñez, J. (2013). The city and urban heat

islands: a review of strategies to mitigate adverse effects. Renewable and Sustainable Energy

Reviews, 25, 749-758.

[7] Ng, E., Chen, L., Wang, Y. and Yuan, C. (2012). A study of the cooling effects of greening in a

high-density city: An experience from Hong Kong. Building and Environment, 47, 256–271.

[8] Srivanit, M. and Hokao, K. (2013). Evaluating the cooling effects of greening for improving

the outdoor thermal environment at an institutional campus in the summer. Building and

Environment, 66, 158-172.

[9] Wong, N.H., Jusuf, S.K., Win, A.A.L., Thu, H.K., Negara, T.S. and Xuchao, W. (2007).

Environmental study of the impact of greenery in an institutional campus in the tropics.

Building and Environment, 42, 2949-2970

[10] Yahia, M.W. and Johansson, E. (2014). Landscape interventions in improving thermal comfort

in the hot dry city of Damascus, Syria-The example of residential spaces with detached

buildings. Landscape and Urban Planning, 125, 1-16.

[11] Lee, H., Mayer, H. and Chen, L. (2016). Contribution of trees and grasslands to the mitigation

of human heat stress in a residential district of Freiburg, Southwest Germany. Building and

Environment, 148, 37-50.

[12] Givoni B. (1989). Urban design in different climates. World Meteorological Organization,

Geneva, Switzerland.

[13] Fahmy, M., Sharples, S. and Yahiya, M. (2010). LAI based trees selection for mid latitude

urban developments: a microclimate study in Cairo, Egypt. Building and Environment, 45,

345-357.

[14] Kotzen B. (2003). An investigation of shade under six different tree species of the Negev desert

towards their potential use for enhancing micro-climatic conditions in landscape architectural

development. Journal of Arid Environments, 55(2), 231-274.

[15] Santamouris, M. (2013). Using cool pavements as a mitigation strategy to fight urban heat

island-A review of the actual developments. Renewable and Sustainable Energy Reviews 26,

224-240.

[16] Takahashi, T. (2011). Challenge for cool city Tokyo, Osaka. GSEP workshop, Washington D.C.,

September, 12-13

[17] Taleghani, M, Sailor, D.J., Tenpierik,, M. and Van den Dobbelsteen, A. (2014). Thermal

assessment of heat mitigation strategies: the case of Portland State University, Oregon, USA.

Building and Environment, 73, 138-150.

[18] Ali-Toudert, F. (2005) Dependence of outdoor thermal comfort on street design in hot and dry

climate. Ph.D. thesis, Freiburg University, Freiburg.

[19] Coutts, A.M., Tapper, N.J., Beringer, J., Loughnan, M. and Demuzere, M. (2013). Watering our

cities: The capacity for Water Sensitive Urban Design to support urban cooling and improve

human thermal comfort in the Australian context. Physical Geography, 37, 2–28.

22

[20] Acero, J.A., Arrizabalaga, J., Kupski, S. and Katzschner, L. (2012). Urban heat island in a

coastal urban area in northern Spain. Theoretical and Applied Climatology, 113(1), 137-154.

[21] Radhi, H., Fikry, F. and Sharples, S. (2013). Impacts of urbanisation on the thermal behaviour

of new built up environments: a scoping study of the urban heat island in Bahrain. Landscape

and Urban Planning, 113, 47–61.

[22] Taleghani, M., Tenpierik, M., Dobbelsteen, A. and Sailor, D.J. (2014). Heat in courtyards: a

validated and calibrated parametric study of heat mitigation strategies for urban courtyards in

the Netherlands. Solar Energy, 103, 108-124.

[23] Steeneveld, G.J., Koopmans, S., Heusinkveld, B.G. and Theeuwes, N.E. (2014). Refreshing the

role of open water surfaces on mitigating the maximum urban heat island effect. Landscape

and Urban Planning, 121, 92-96.

[24] Höppe, P. (1999). The physiological equivalent temperature-A universal index for the

biometeorological assessment of the thermal environment. International Journal of

Biometeorology, 43(2), 71–75.

[25] Ng, E. and Cheng, V. (2012). Urban human thermal comfort in hot and humid Hong Kong.

Energy and Buildings, 55, 51-65.

[26] Chen, L., & Ng, E. (2012). Outdoor thermal comfort and outdoor activities: A review of

research in the past decade. Cities, 29(2), 118-125.

[27] Acero, J.A. and Herranz-Pascual, K. (2015) A comparison of thermal comfort conditions in

four urban spaces by means of measurements and modelling techniques. Building and

Environment, 93, 245-257.

[28] Lee, H., Mayer, H. and Chen, L. (2016). Contribution of trees and grasslands to the mitigation

of human heat stress in a residential district of Freiburg, Southwest Germany. Building and

Environment, 148, 37-50.

[29] Thorsson, S., Lindberg, F., Eliasson, I. and Holmer, B. (2007). Different methods for

estimating the mean radiant temperature in an outdoor urban setting. International Journal of

Climatology, 27, 1983-1993.

[30] ASTM E1918-06 (2015). Standard test method for measuring solar reflectance of horizontal

and low-sloped surfaces in the field. American Society for Testing and Materials. West

Conshohocken, PA, USA.

[31] Bruse, M. and Fleer, H. (1998). Simulating surface–plant–air interactions inside urban

environments with a three dimensional numerical model. Environmental Modeling and

Software, 13, 373-384.

[32] Yang, X., Zhao, L., Bruse, M. and Meng, Q. (2013). Evaluation of a microclimate model for

predicting the thermal behaviour of different ground surfaces. 60, 93-104.

[33] Acero, J.A. and Arrizabalaga, J. (2016). Evaluating the performance of ENVI-met model in

diurnal cycles for different meteorological conditions, Theoretical and Applied Climatology,

doi:10.1007/s00704-016-1971-y.

[34] Lands Department of Hong Kong (2007). Government Digital Topographic Map Data, Survey

and Mapping Office.

[35] Giridharan R, Lau SSY, Ganesan S, Givoni B. (2007). Urban design factors influencing urban

heat island intensity in high rise high density environments of Hong Kong. Building and

Environment, 42(10), 3669-3684.

[36] Oke, T.R. (1987). Boundary Layer Climates. Methuen, London.

[37] PB-BEC (2007). Guidelines on Performance based building energy code, 2007 edition.

Electrical & Mechanical Services Department, Hong Kong.

[38] Mohd Fairuz Shahidan (2015) Potential of Individual and Cluster Tree Cooling Effect

23

Performances Through Tree canopy Density Model Evaluation in Improving Urban

Microclimate, Current world environment, 10 (2), 398-413.

[39] Tan, Z., Lau, K.K.L. and Ng, E. (2016). Urban tree design approaches for mitigating daytime

urban heatisland effects in a high-density urban environment. Energy and Buildings, 114,

265-274.

[40] Davenport, A.G. (1960). Rationale for determining design wind velocities, American Society

of Civil Engineers Journal of the Structural Division, 86, 39-68.

[41] Ng, E., Yuan, C., Chen, L., Ren, C. and Fung, J.C. (2011). Improving the wind environment in

high-density cities by understanding urban morphology and surface roughness: a study in Hong

Kong. Landscape and Urban Planning, 101(1), 59-74.

[42] Hong Kong Observatory (2014). Summary of Meteorological and Tidal Observations in Hong

Kong. Website: http://www.hko.gov.hk/publica/smo/smo2014.pdf (last accessed on 6 July

2016).

[43] Ali-Toudert F. and Mayer H. (2006). Numerical study on the effects of aspect ratio and

orientation of an urban street canyon on outdoor thermal comfort in hot and dry climate.

Building and Environment, 41(2), 94-108.

[44] Matzarakis, A., Rutz, F. and Mayer, H. (2010). Modelling radiation fluxes in simple and

complex environments: basics of the RayMan model. International Journal of Biometeorology,

54, 131-139.

[45] Willmott, C.J. (1982). Some comments on the evaluation of model performance. Bulletin of the

American Meteorological Society, 63(11), 1309-1313.

[46] Santhi, C, Arnold, J. G., Williams, J. R., Dugas, W. A., Srinivasan, R., and Hauck, L. M. (2001).

Validation of the SWAT model on a large river basin with point and nonpoint sources. Journal

of the American Water Resources Association, 37(5), 1169-1188.

[47] Krüger E.L., Minella F.O. and Rasia F. (2011). Impact of urban geometry on outdoor thermal

comfort and air quality from field measurements in Curitiba, Brazil. Building and Environment,

46(3), 621-634. [48] Song, B., Park, K. and Jung, S. (2014). Validation of ENVI-met model with in situ

measurements considering spatial characteristics of land use types, Journal of the Korean

Association of Geographic Information Studies,17(2), 156-172.

[49] Huttner, S., (2012). Further development and application of the 3D microclimate simulation

ENVI-met, Ph.D. dissertation. Gutenberg university.