Impacts of inland water area changes on the local climate of Wuhan,China

Abstract

A distinct feature of Wuhan is that almost a quarter of the total area of this city is covered with water, leading to its unique hot and humid climate characteristics in summer. However, according to records, water area in built-up zone of Wuhan has been reduced by 130.5 km² from 1965 to 2008, while the annual average air temperature has been increased by more than 3℃. To investigate the quantitative connection between the water area reduction and air temperature increase, three scenarios were simulated in a summer; to evaluate the impact of water reduction on the local thermal environment in different water areas; and to study the impact of water reduction on the urban heat island (UHI) phenomenon. Meso-scale meteorological models of Weather Research and Forecasting model were applied in this study for quantitative assessment and prediction. With the predictions, this study reveals that the decreased water area could affect air temperature, wind velocity and wind flow direction, energy balance and the UHI intensity. The simulations show that areas with significant wind velocity, wind direction and air temperature differences are distributed among the downwind zones. Moreover, the areas with high UHI intensity are wider and farther from the boundary of urban areas because of the reduction of water areas.

Keywords

Land-use alteration Meso-scale metrological model Thermal environment Urban climate change Urban heat island Water areas

Introduction

Many factors could contribute to the development of the urban heat island (UHI) phenomenon. Some are related to natural factors such as weather and location.¹^–³ Other factors are related to human activity, such as the reduction of vegetation and water bodies, urban geometry and materials, and anthropogenic heat emissions.⁴ Change of anthropogenic heat release is one of the major reasons leading to the alteration of local climate. Land-use alteration could also affect the local climate, especially when the underlying layer changes from water to land. Lopes et al.,⁵ for instance, examined the impact of land surfaces and aerodynamic roughness and concluded that an increase in roughness would cause a 40% reduction of the wind speed. In addition, changes in albedo and roughness length could also greatly affect the local thermal environment. Jusuf et al.⁶ showed the various impacts of land use on urban ambient temperature.

This study estimated the impact of alteration of land use on the local climate and thermal environment without considering anthropogenic heat release. The domain of interest was Wuhan, which is located in the geographic centre of China. Details of Wuhan are summarized in Table 1. This study focused on the area with the greatest land-use alteration from water to land in Wuhan, which is in the core of the city described as the built-up zone of Wuhan. According to records, water areas have been reduced from 518.25 km² to 387.75 km² in 43 years (from 1965 to 2008). Figure 1 shows the annual average air temperature of Wuhan from 1965 to 2008.⁸ A trend of 3℃ to 5℃ air temperature increase over 43 years has been noted, which could be related to the alteration of land use during that period. A similar average air temperature increase has also been found in some cities in Japan. From 1946 to 1976, Fujibe⁹ found a 1℃ to 1.5℃ warming over the Kanto plain. Hiroyuki et al.¹⁰ carried out numerical simulations of the Kanto plain. Warming due to land-use alteration was found over the Tokyo metropolitan area, and the intensity of the daytime heat island effect in that area was estimated to be 3℃ to 4℃. Before three-dimensional meso-scale models were used to estimate the effect of land use on local climate, most of the studies on land-use alteration were based on observation and analysis. However, observational and analytical studies would demand a large amount of data and would provide limited scope for forecasting.

Table 1.

Details of Wuhan.

Location	113°41′–115°05′ East
	29°58′–31°22′ North
Population	9,100,000 (2010)
Total area	8494.41 km²
Built-up area	660 km²
Water area	2205.06 km²
Number of water bodies in Wuhan	170
Number of water bodies in built-up area of Wuhan	41

Numerical meso-scale simulations can offer predictions. Kirtsaeng et al.¹¹ validated the Weather Research and Forecasting (WRF) model (A new generation of meso-scale simulation model for WRF) performance during heavy rain by using the rainfall data observed at the tropical rainfall measuring missions and verified the rainfall simulation results with the Thai meteorological department. Božnar et al.¹² presented an evaluation of the use of the WRF model as the source of wind profile information and tested WRF’s one day short-term forecasts, comparing these data to the associated data recorded by meteorological stations. The results showed an inadequate consistency with ground-level meteorological stations. Ruiz et al.¹³ tested the WRF model in South America using different configurations to identify which one would give the best estimates of observed surface variables. All of these prior studies showed that the WRF model is reliable for meteorological meso-scale simulations and would provide confidence in the model as a prediction tool to estimate the change of local climate and thermal environments due to land-use alterations.

A number of studies have focused on the influence of land-use alterations on the local climate and environment, and many realistic meso-scale meteorological simulations of UHI phenomena were carried out by Miao et al.,¹⁴ Trusilova et al.,¹⁵ Freitas et al.,¹⁶ Zhang et al.,^17,18 Taha and Meier,¹⁹ Taha,²⁰ and Erell and Williamson.²¹ Trusilova et al.¹⁵ noted the effects of alteration of land use on urban landscapes, which can significantly contribute to changes of the near-surface temperature and precipitation at local and regional scales. Trusilova et al.¹⁵ also indicated that the alteration of land use to urban landscapes may cause a reduction in the diurnal temperature range by more than −1.2℃ in summer and more than −0.7℃ in winter. Similar studies have also been carried out in China, such as by Miao et al.¹⁴ who evaluated the UHI in Beijing.

Wuhan is the capital of Hubei province in Central China. It is situated on the east of Jianghan Plain, and the city centre lies at the confluence of two rivers (Yangtze River and Hanshui River). The city is surrounded by abundant water bodies; therefore, it has been described as a city of hundred lakes. Water area in the city is almost a quarter of the total area of Wuhan. In recent years, due to the recent rapid industrialization and urban development, vast water areas have been disappearing in the built-up areas of Wuhan. The effect of reduction in water area is rarely studied, not to mention, the study of the impact of reduction of water area on the thermal environment and UHI phenomenon in Wuhan.

The main purpose of this study is to numerically estimate the impact of water reduction on local thermal environment, taking into consideration of air temperature, humidity and wind speeds. Because UHI has become a serious issue both in China and globally, the main focus of this study was on the UHI phenomenon of Wuhan. Ichinose et al.²² indicated the UHI of Tokyo primarily formed because of the decrease of latent heat release. To find the causes of UHI in Wuhan, this study aimed to ascertain the impact of sensible, latent and ground heat fluxes between ground boundary surface and atmosphere before and after land-use alteration.

Numerical model and study area

Numerical model

The numerical model used in the present study is the Advanced Research WRF dynamical solver together with other components of the WRF system that are compatible with that solver and used in producing a simulation. The basic features of the model that were adapted to the present study are summarized in Table 2. Further discussion of surface layer physics can be found in Monin and Obukhov,²⁴ details of microphysics can be found in Thompson et al.,²⁶ and an introduction of long-wave and short-wave radiation physics can be found in Mlawer et al.²⁷ The following is a brief description of the Noah land surface model (LSM) applied in this study.²⁹

Table 2.

Schemes selected in WRF.

Boundary layer physics	Mellor–Yamada–Janjic (Eta) TKE scheme²³
Surface layer physics	Monin–Obukhov (Janjic Eta) scheme²⁴
Surface physics	Unified Noah land-surface model²⁵
Microphysics	Thompson graupel scheme (2-moment scheme in V3.1)²⁶
Long-wave radiation physics	RRTM (rapid and accurate radiative transfer)²⁷
Short-wave radiation physics	Goddard shortwave scheme²⁸

TKE: Turbulence Kinetic Energy.

The ground heat flux of the model thermodynamics is calculated by using equation (1) as follows

C (Θ) \frac{\partial T}{\partial t} = \frac{\partial}{\partial z} [K_{t} (Θ) \frac{\partial T}{\partial z}]

(1)

where T, C, K_t and

Θ

are the soil temperature of (K), volumetric heat capacity (J m⁻³ K⁻¹) and thermal conductivity (W m⁻¹ K⁻¹) that are formulated (equations (2) to (4)) as functions of volumetric soil water content (fraction of unit soil volume occupied by water), respectively.

C = Θ C_{water} + (1 - Θ_{s}) C_{soil} + (Θ_{s} - Θ) C_{air}

(2)

where the volumetric heat capacities are

C_{water} = 4.2 \times 106 J m - 3 K - 1, C_{soil} = 1.26 \times 106 J m - 3 K - 1, C_{air} = 1004 J m - 3 K - 1, and

< ? TeX Θ_{s} ? >

is maximum soil moisture

K_{t} (Θ) = {420 \exp [- (2.7 + P_{f})], P_{f} \leq 5.1 0.1744, P_{f} > 5.1

(3)

and

P_{f} = \log [Ψ_{s} (Θ_{s} / Θ) b]

(4)

where

Ψ_{s}

and b are the saturated soil potential and curve-fitting parameters that depend on the soil type.

The prognostic equation of model hydrology is described by equation (5) as follows

\frac{\partial Θ}{\partial t} = \frac{\partial}{\partial z} (D \frac{\partial Θ}{\partial z}) + \frac{\partial K}{\partial z} + F_{Θ}

(5)

where D and K are the soil water diffusivity and hydraulic conductivity, respectively. Both of which are a function of

Θ

F_{Θ}

represents the sources and sinks (i.e. precipitation, evaporation, and runoff) for soil water. The relationships of D,

K (Θ)

and

Ψ (Θ)

are defined by equations (6) to (8)

D = K (Θ) \frac{\partial Ψ}{\partial Θ}

(6)

K (Θ) = K_{s} (Θ / Θ_{s}) 2 b + 3

(7)

Ψ (Θ) = Ψ_{s} / (\frac{Θ}{Θ_{s}}) b

(8)

where K_s is the saturated hydraulic conductivity depending on soil type.

Evaporation within the hydrology model is calculated according to equation (9) as follows

E = E_{dir} + E_{c} + E_{t}

(9)

where E,

E_{dir}

, E_c and E_t are total evaporation, direct evaporation from the top shallow soil layer, evaporation of precipitation intercepted by the canopy, and transpiration via canopy and roots, respectively. The parameters are defined by equations (10) to (14)

E_{dir} = (1 - σ_{f}) β E_{p}

(10)

β = \frac{Θ_{l} - Θ_{w}}{Θ_{ref} - Θ_{w}}

(11)

E_{c} = σ_{f} E_{p} (\frac{W_{c}}{S}) n

(12)

\frac{\partial W_{c}}{\partial t} = σ_{f} P - D - E_{c}

(13)

E_{t} = σ_{f} E_{p} B_{c} [1 - (\frac{W_{c}}{S}) n]

(14)

where E_p,

σ_{f}

Θ_{ref}

Θ_{w}

, W_c, S, D, P and B_c are potential evaporation, green vegetation fraction, field capacity, wilting point, intercepted canopy water content, maximum canopy capacity, the excess precipitation or drip, input total precipitation and function of canopy resistance, respectively, with n = 0.5.

As noted by Chen and Dudhia,²⁹ the Noah LSM has overcome the major weaknesses of the simple LSM, which are: failure to take into account of soil moisture changes in the simulation; not reflecting the impact of recent precipitation; no snow cover prediction; relatively coarse resolution of land use; and not considering the vegetation evapotranspiration and runoff processes in the model simulation. Noah LSM has made significant improvements towards accurately representing land and atmosphere interactions, and this model is now applicable for this study.

Validation

A sensitivity analysis was conducted to validate the WRF model by comparing the simulated and measured air temperature at 2 m above ground level (AGL), which is referred to as T2 in the following text. The measured data were obtained from a meteorological station (30.37°N, 114.08°W) located in a rural area of Wuhan city, as shown in Figure 2.

Figure 3 compares the measured and calculated air temperatures at 2 m AGL (T2) on 26 July 2008. The data vary during the period from 0000 local solar time (LST) to 0800 LST, with a maximum difference of 1.56℃ at 0500 LST. Differences in T2 were caused exclusively by anthropological heat in the simulations because the measured data were especially high at night. For the other periods, remarkable agreements were found, with a minimum difference of 0.15℃ (at 1100 LST).

In addition, a mobile measurement was conducted from 13 to 16 August 2011. The route of the mobile measurement is marked in a red line in Figure 2. Figure 4 shows the comparisons between the measured and calculated T2 on 15 August 2011. Figure 4(a) and (b) shows the results at 1200 LST and 2100 LST, respectively. Along the route of the mobile measurement, although some of the points show that T2 differences were higher than 1℃, most of the points were less than 1℃. Remarkable agreements were found in the increasing and decreasing trends of T2 from the starting point to the terminal point along the route. Major differences found in Figure 4(b) were near the starting point close to the river side 2100 LST. The measured results were higher than the simulated results, with differences being primarily the result of anthropogenic heat released from vehicles, which was difficult to properly simulate in the urban canopy model, not to mention in the slab model.

These results provide confidence in the model as a predictive tool that can be used to estimate the change of local climates and thermal environments due to land-use alteration.

Description of Wuhan city and domain calculation

Wuhan is located in the geographic centre of China and is the capital city of Hubei province. The Yangtze River (Figure 5) is the longest river in Asia and runs through Wuhan. The central region of Wuhan includes the entire built-up zone within a radius of 30 km, which consists of many rivers and lakes. Water areas inside the built-up zone of Wuhan have been reduced due to urban development from 1965 to 2008 by 130.5 km². The water area maps of the built-up zone from 1965 and 2008⁸ are shown in Figure 5(a) and (b), respectively, and are used to calculate the domains shown in Figure 6.

Domain 3 includes the entire built-up zone of Wuhan and is centred at 114.30°E, 30.50°N. Resolution of the horizontal grid of domain 3 is 0.5 km. The coarse grid values at the boundaries of the nested area have provided boundary conditions for the nested grid. The vertical direction was divided into 35 layers from the ground surface up to an altitude of approximately 20 km in the model. Further details of computational domains and grid arrangements are listed in Table 3.

Table 3.

Computational domains and grid arrangements.

	Domain (X × Y × Z), km	Grid number	Grid size (km)
D1	225 × 225 × approx. 20	50 × 50 × 35	4.5
D2	150 × 150 × approx. 20	100 × 100 × 35	1.5
D3	50 × 50 × approx. 20	100 × 100 × 35	0.5

The original data for land use in Wuhan were obtained from United States Geological Survey (USGS) land-use and land-cover dataset. The geological data applied in the simulation model were modified based on the USGS dataset. There are eight categories of land use in the Wuhan area: urban, dry land crop (Dryland C.P.), irrigation crop (Irrg. C.P.), crop/grass mosaic (C/G. Mosaic), crop/wood mosaic (C/W. Mosaic), grassland, shrub land and water. Table 4 shows the land-use categories used in this study and the values of the radiation and surface parameters for each of the 10 categories. In this study, anthropogenic heat was not taken into account so that the focus was on the impact of land-use alteration on local climate and thermal environments.

Table 4.

Surface parameters of USGS land-use and land-cover data.

Land use	Albedo (%)	Moisture avail. (%)	Emissivity (% at 9 µm)	Roughness length (cm)	Thermal inertia (cal cm⁻² k⁻¹ s^−1/2)
Urban	15	10	88	80	0.03
Dryland C.P.	17	30	98.5	15	0.04
Irrg. C.P.	18	50	98.5	15	0.04
C/G. Mosaic	18	25	99	14	0.04
C/W. Mosaic	16	35	98.5	20	0.04
Grassland	19	15	98.5	12	0.03
Shrubland	22	10	88	10	0.03
Water bodies	8	100	98	0.01	0.06

To quantitatively evaluate the impact of land-use alterations on local climate and thermal environments, three scenarios were designed: land use in 1965; land use in 2008; and an ideal scenario with no water areas within the built-up zone of Wuhan. To clarify the effect of urbanization and land-use alteration, the same meteorological dataset was applied to the three scenarios. In addition, only water areas were changed within the scenarios. Table 5 lists the names and details of water areas remained in domain 3 in each of the three scenarios. Simulations for the thermal environment in 1965, 2008 and the no-water scenario are referred to as scenarios ‘1965’, ‘2008’ and ‘No-water’, respectively. Scenario ‘No-water’ presumes that all the water areas inside the built-up zone of Wuhan vanished, except for the Yangtze River. The land-use distribution of the three scenarios is given in Figure 7.

Table 5.

Water area changes by land-use alteration in the three scenarios.

Scenario	Water area (km²)
‘1965’	518.25
‘2008’	387.75
‘No-water’	229

Initial condition

The numerical integration for the three scenarios started at 2000 LST with a 1° × 1° horizontal resolution of meteorological data from the National Centers for Environmental Prediction Final Operational Global Analysis data for the period from 23 to 27 July of 2008 under clear-sky weather conditions. This period was selected to guarantee the necessary conditions for the appearance of a clear UHI and included the hottest days of the year in 2008.

Results and discussion

Simulation of summer period for 2008

The hottest summer period of 2008 was simulated using the 2008 land-use data. The simulated surface winds and T2 at 1600 LST and 0600 LST for scenario ‘2008’ are shown in Figure 8(a) and (b), respectively. Figure 8 shows the relationship between water area and urban landscape in Wuhan. In contrast to Figure 8(a), Figure 8(b) shows an apparent warming effect of water area. Therefore, water areas indicated a better ‘cold source’ under the conditions in Figure 8(a) than under the conditions in Figure 8(b). A slight cooling effect can be found along the water bank.

Figure 9 illustrates the average diurnal T2, surface boundary temperature (TSK), relative humidity at 2 m AGL (Q2) and upward moisture flux for the entire area of the built-up zone. As shown in Figure 9(a), the peak value of TSK appears at 1400 LST and was 43.86℃. Compared to TSK, the T2 peak appeared 1 h later and has a value of 37.88℃. The average diurnal T2 range for the entire area of the built-up zone was 11℃. Figure 9(b) shows the average diurnal Q2 range for the entire area of the built-up zone which was 5 g/kg.

Figures shown in this section give a general idea of the local climate and thermal environment in the hottest summer period of 2008.

Impact of land-use alteration on local thermal environment

Figures 10 to 13 show the changes in heat budget, humidity and temperature caused by land-use alteration (water area reduction) in the built-up zone of Wuhan. To clarify the impact of water area changes on the urban thermal environment, only the area without changes in land use and cover was evaluated. To quantify the effect, one solution was to remove the comparisons between cases with water area changes. As previously mentioned, there are three scenarios: the control case of ‘2008’, land-use data of ‘1965’ and ‘No-water’, in which no water is left inside the built-up zone of Wuhan (except for the Yangtze River).

The model simulations of the heat budget at land surfaces for scenarios ‘1965’, ‘2008’ and ‘No-water’ are shown in Figure 10(a) to (c), respectively. Rn is the value of the net heat budget of the land surface. The land surface temperature increases with a positive Rn and decreases with a negative Rn. Water is very different from land, with an albedo three times smaller and a specific heat capacity three times larger. Therefore, water can absorb and store much more heat than land. The heat flux components of Rn, such as sensible heat flux, latent heat flux and ground heat flux, were calculated and presented in Figure 10(a) to (c) for the built-up zone of Wuhan including inland water area for scenarios ‘1965’, ‘2008’ and ‘No-water’, respectively. In the case with a larger water area, the upward latent heat flux is larger before sunrise and after sunset and smaller around midday, which is most likely because of the change of evaporation caused by the land-use alteration. The amount of water, heat loss due to evaporation, water vapour saturation deficit, wind velocity and turbulent diffusion can all affect the latent heat-related evaporation. Before sunrise and after sunset, the heat supplied, water vapour saturation deficit and turbulent diffusion are comparatively small; therefore, the main reason for the differences among the three scenarios is the amount of water in the built-up zone of Wuhan, and the ‘1965’ scenario has an advantage during that period. Around midday, evaporation in ‘1965’ was obviously smaller than the other two scenarios (shown in Figure 11). During that period, solar radiation was the main heat supply and worked as the primary force to promote evaporation; however, the supersaturated water vapour returned to the water body and blocks the evaporation during humid conditions.

Because water can absorb and store much more heat than land, it should also be capable of reducing the heat burden to the atmosphere because of increased heat gains and decreased heat losses compared to land. As expected, our calculated results of the water surface temperature varied around the initial value of 27.6℃ in the model of WRF, with the bias in a range of ±0.3℃. Compared to the land surface, the water surface showed less of an effect on the atmosphere due to heating up the air because of its lower surface temperature. For scenario ‘1965’, the wider water increased the relative humidity, especially around the midday when the heating of the atmosphere is strongest; however, it had the least amount of evaporation. Because less evaporation may reduce the latent heat loss, people sometimes feel warmer when conditions are cold and wet than when conditions are hot and dry. In addition, according to Figure 10, the diurnal range of humidity is smaller with larger water areas, which is the same for the diurnal range of T2 shown in Figure 11. Because of the smaller differences in thermal environments between day and night in the scenario with a larger water area, the environment is more comfortable in ‘1965’ than in ‘2008’ and ‘No-water’. Figure 12 shows that with the water area decreasing from ‘1965’ to ‘2008’, the T2 was significantly higher after sunset which illustrates that water areas had an effect on the local thermal environment during that period. Furthermore, a reduction of water areas weakens the cooling effect of water after sunset.

To determine the distribution of the areas greatly affected by water in the built-up zone of Wuhan, Figure 13 illustrates the differences of T2, which were higher than 0.5℃ between ‘1965’ and ‘2008’ and between ‘No-water’ and ‘2008’. Although the areas with T2 were higher than 0.5℃ and varied hourly with the wind direction, they were mostly located at the downwind side of the areas with obviously reduced water areas. The influence of water area reduction was wide spread in Wuhan because of the wind. At nightfall (before 2000 LST), the area with T2 differences higher than 0.5℃ consists of almost 1/5 of the built-up zone.

Because a cooling process always takes place in the evening, the cooling effect would be weakened if the water area is reduced. Therefore, as shown in Figure 13, the warming effect in the evening is much clearer when water areas were reduced. Accordingly, additional heat would accumulate in the built-up zone of Wuhan, making the thermal environment increasingly hot and thus causing a stronger effect of the UHI phenomenon.

Impact of land-use alterations on local wind systems

In this section, estimates are given of the impact of land-use alterations for scenarios ‘1965’, ‘2008’ and ‘No-water’ on local wind systems. To clarify the effects of urbanization and land-use alteration, the same meteorological dataset was applied in the three scenarios. Furthermore, only water areas were changed among scenarios. Wind velocity at 10 m AGL for the built-up zone at 1400 LST for scenarios ‘1965’, ‘2008’ and ‘No-water’ is shown in Figure 14(a) and (b). To compare the different scenarios, ‘1965’ and ‘2008’ are shown in Figure 14(a); ‘No-water’ and ‘2008’ are shown in Figure 14(b).

Wind velocity increased significantly when the water areas were reduced in scenarios of ‘1965’ to ‘2008’ to ‘No-water’. The higher wind velocity might be caused by the higher air temperatures. The areas with significant differences are highlighted inside the contour lines. The areas are in the downwind direction of 1400 LST. As mentioned earlier, the decrease of water area would cause the air temperature to rise. Cool air coming from rural area, passes through the built-up area, would become heated up, and would gain more kinetic energy. Therefore, wind velocity is higher in the cases of less water area and significant velocity differences are located in the downwind direction. In addition, the wind direction has slightly changed towards the east because of the reduction of water area. Altered wind velocity and direction are caused by the alteration of air temperature and its distribution. The areas with significant differences in wind velocity and direction are distributed among the downwind zones, which also show differences in air temperature and distribution.

Figure 15 shows the average wind velocity and direction over the course of a day for the entire area of the built-up zone for the three scenarios. During the period from 1400 LST to 1700 LST, wind velocity increased with a reduction in water area. Conversely, during the period from 0000 LST to 0600 LST, the wind velocity decreased along with water area. Within the hottest period of the day, the T2 would be highest in scenario with ‘No-water’ and would cause the largest local pressure differences. Hence, the wind velocity was also the highest in ‘No-water’ area. During the predawn hours, the wind velocity is primarily determined by roughness length because of small temperature differences between the three scenarios. The scenario of ‘1965’ had the largest water area and had the lowest roughness length.

Impact of land-use alterations on UHI

To confirm the hypothesis that water area reduction has a stronger effect on the UHI phenomenon, we calculated the UHI intensity for the three scenarios, which is shown in Figure 16. UHI intensity is defined as the T2 difference between the urban and rural landscapes. In this study, UHI intensity was calculated by the average T2 of the urban and rural landscapes in domain 3 (T2 of the water areas was not included). The UHI phenomenon is significant in the evening, and water area reductions have made it stronger, with a difference of less than 0.6℃ in the UHI intensity for the three scenarios. Generally, the UHI problem was underestimated by using the average values of UHI intensity, which ignored the influence of anthropogenic heat release. From Figure 16, evident of UHI intensity differences among cases appeared around midday, and after 2200 LST. It shows that the cooling effect of water happened during these two periods. Therefore, reduction of water area would cause air temperature to rise and heat to accumulate. Hence, the reduction of water area would affect the urban outdoor thermal comfort and cause more serious UHI phenomenon.

Figure 17 shows the areas with UHI intensity higher than 3℃ for the three scenarios. In Figure 17, the areas with UHI intensity were higher than 3℃ when there was a reduction in water areas. For areas with a high UHI intensity close to the boundary of urban landscapes, the UHI phenomenon tends to be mitigated by a penetration of cool air from suburban environments. However, with the reduction of water areas, areas with high UHI intensity are wider and farther from the boundary of urban landscapes; therefore, more difficult for the cool air from suburban environments to permeate deeply into the urban area when there is a reduction in water areas. Therefore, the heat generated in the urban areas could accumulate gradually, and the UHI phenomenon could become stronger due to the reduction in water areas.

Conclusions

Predicting the impact of land-use alteration on thermal environment and climate is possible by applying the meso-scale meteorological models as demonstrated by this study. The WRF model is one of the meso-scale meteorological models used in this study; this model has taken into account of the major weakness of the simple LSM and has been validated as a much improved model for this study.

With the validation, remarkable agreements were found with a minimum T2 difference of 0.15℃ between the simulations and measurements (at 1100 LST). Although there are specific points where T2 differences were higher than 1℃, most of the T2 differences were relatively small. Significant T2 differences are primarily due to the anthropogenic heat release which was not considered. Because anthropogenic heat release is not the issue, we primarily discussed the way the applied model would fit the goals of this study and its reliability as a prediction tool. However, studying the urban canopy model with the inclusion of anthropogenic heat release will be considered as a future area of research.

With a reduction in water area, air temperature would be significantly increased after sunset, thus affecting the local thermal environment. The cooling effect of water would mainly happen in the midday and after sunset. Wind velocity and direction would be altered because of changes in air temperature and distribution. The areas with obvious changes in wind velocity, wind direction are mostly located at the downwind side of the areas with evident reductions in water areas. The influence of water area reduction was wide spread in Wuhan because of the changes in wind flow.

It also became clear that land-use alteration (water area reduction) would reduce the relative humidity, leading to a reduced evaporation and reduced latent heat loss, with urban residents feeling warmer when the weather is cold and wet than when it is hot and dry. T2 is higher with smaller water areas, the humidity would be lower and the latent heat flux would become higher. Further research is necessary to investigate whether scenario of ‘2008’ is more comfortable than ‘1965’ and if scenario of ‘No-water’ is more comfortable than ‘2008’.

Our study has illustrated that the reduction in water area has a definite strong effect on the UHI phenomenon. With the reduction of water areas, the areas with high UHI intensity are wider and farther from the boundary of urban landscapes, to allow the cool air from suburban areas to permeate deeply into the urban areas to cool the environment. Therefore, the heat generated in urban areas would accumulate gradually, and the UHI phenomenon would become stronger due to the reduction of water areas.

Authors’ contribution

Xuefan Zhou contributed to the project implementation and writing of the paper, Ryozo Ooka contributed to the general supervision of the project, Hong Chen was the project leader and contributed to the general review of the project, Yoichi Kawamoto contributed to the technical support and Hideki Kikumoto contributed to the detailed supervision of the project.

Declaration of conflicting interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Footnotes

Funding

This work was supported by National Nature Science Foundation of China (Grant No. 50978110), Specialized Research Fund for the Doctoral Program of Higher Education (Grant No. 20100142110040) and China Scholarship Council (File No. 2010616031), National Science and Technology Support Program (Grant No. 2011BAJ03B03).

References

Sato

Murakami

Ooka

Yoshida

. Analysis of regional characteristics of the atmospheric heat balance in the Tokyo metropolitan area in summer. J Wind Eng Ind Aerodyn 2008; 96: 1640–1654.

Radhi

Fikry

Sharples

. Impacts of urbanisation on the thermal behaviour of new built up environments: a scoping study of the urban heat island in Bahrain. Landsc Urban Plann 2013; 113: 47–61.

CWF

Deng

. The influence of synoptic and atmospheric boundary layer on PM₁₀ and urban heat island. Indoor Built Environ 2013; 22(5): 796–907.

CWF

Jin

Dong

. Regression analysis of the relationship between urban heat island effect and urban canopy characteristics in a mountainous city, Chongqing. Indoor Built Environ 2012; 21(6): 821–836.

Lopes

Saraiva

Alcoforado

. Urban boundary layer wind speed reduction in summer due to urban growth and environmental consequences in Lisbon. Environ Model Softw 2011; 26: 241–243.

Jusuf

Wong

Hagen

Anggoro

Hong

. The influence of land use on the urban heat island in Singapore. Habitat Int 2007; 31: 232–242.

Wuhan Meteorological Service. Meteorological data of Wuhan from 1965 to 2008, http://www.whqx.gov.cn/bmgk.jsp (2008, accessed November 2010).

Wuhan Land Resources and Planning Bureau. 武汉市土地利用变更情况 [Land use alteration data of Wuhan], http://www.wpl.gov.cn/pt-337-2.html (2006, accessed 10 April 2011).

Fujibe

. Long-term falling trends of pressure over the Kanto Plain as evidence of increasing heat content in the lower atmosphere in the daytime of the warm season. J Meteor Soc Jpn 1994; 72: 785–792.

10.

Hiroyuki

Fujio

Hiromaru

Masafumi

. The effects of land-use alteration on the sea breeze and daytime heat island in the Tokyo metropolitan area. J Meteor Soc Jpn 2000; 78(4): 405–420.

11.

Kirtsaeng

Kreasuwun

Chantara

. Weather Research and Forecasting (WRF) model performance for a simulation of the 5 November 2009 heavy rainfall over southeast of Thailand. Chiang Mai J Sci 2012; 39(3): 511–523.

12.

Božnar

Mlakar

Grašič

. Short-term fine resolution WRF forecast data validation in complex terrain in Slovenia. Int J Environ Pollut 2012; 50(1–4): 12–21.

13.

Ruiz

Saulo

Nogues-Paegle

. WRF model sensitivity to choice of parameterization over South America: validation against surface variables. Mon Weather Rev 2010; 138(8): 3342–3355.

14.

Miao

Chen

LeMone

Tewari

Wang

. An observational and modeling study of characteristics of urban heat island and boundary layer structures in Beijing. J Appl Meteor Climatol 2009; 48(3): 484–501.

15.

Trusilova

Jung

Churkina

Karstens

Heimann

Claussen

. Urbanization impacts on the climate in Europe: numerical experiments by the PSU-NCAR Mesoscale Model (MM5). J Appl Meteor Climatol 2008; 47(5): 1442–1455.

16.

Freitas

Rozoff

Cotton

Dias

PLS

. Interactions of an urban heat island and sea-breeze circulations during winter over the metropolitan area of Sao Paulo, Brazil. Bound-Layer Meteor 2007; 122(1): 43–65.

17.

Zhang

Gao

Wang

Chen

. Modeling the impact of urbanization on the local and regional climate in Yangtze River Delta, China. Theor Appl Climatol 2010; 102: 331–342.

18.

Zhang

Zhu

. Urban heat island and boundary layer structures under hot weather synoptic conditions: a case study of Suzhou City, China. Adv Atmos Sci 2011; 28(4): 855–865.

19.

Taha H and Meier A. Mitigation of urban heat islands: meteorology, energy, and air quality impacts. In: Proceedings of the international symposium on monitoring and management of urban heat island, CREST/JST, Keio University, Fujisawa, Japan, November 19–20 1997; pp.124–163.

20.

Taha

. Meso-urban meteorological and photochemical modeling of heat island mitigation. Atmos Environ 2008; 42: 8795–8809.

21.

Erell

Williamson

. Intra-urban differences in canopy layer air temperature at a mid-latitude city. Int J Climatol 2007; 27(9): 1243–1255.

22.

Ichinose

Shimodozono

Hanaki

. Impact of anthropogenic heat on urban climate in Tokyo. Atmos Environ 1999; 33: 3897–3909.

23.

Mellor

Yamada

. Development of a turbulence closure model for geophysical fluid problems. Rev Geophys 1982; 20(4): 851–875.

24.

Monin

Obukhov

. Basic laws of turbulent mixing in the surface layer of the atmosphere. Tr Akad Nauk SSSR Geophiz 1954; 24(151): 163–187.

25.

Mitchell

Lin

Rogers

Grunmann

Koren

Gayno

Tarpley

. Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta model. J Geophys Res 2003; 108(D22): 8851–8851.

26.

Thompson

Rasmussen

Manning

. Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: description and sensitivity analysis. Mon Weather Rev 2004; 132: 519–542.

27.

Mlawer

Taubman

Brown

Iacono

Clough

. Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J Geophys Res 1997; 102(D14): 16663–16682.

28.

Mielikainen

Huang

Goldberg

. GPU acceleration of the updated goddard shortwave radiation scheme in the weather research and forecasting (WRF) model. Appl Earth Observ Remote Sens 2012; 5(2): 555–562.

29.

Chen

Dudhia

. Coupling an advanced land surface-hydrology model with the Penn State-NCAR MM5 modeling system. Part I: model implementation and sensitivity. Mon Weather Rev 2001; 129: 569–585.