• ADVANCES IN ATMOSPHERIC SCIENCES, 2017, 34(12): 1404-1414
    doi: 10.1007/s00376-017-6333-x.
    Effects of Wind Fences on the Wind Environment around Jang Bogo Antarctic Research Station
    Jang-Woon WANG1, Jae-Jin KIM1,, Wonsik CHOI1, Da-Som MUN1, Jung-Eun KANG1, Hataek KWON2, Jin-Soo KIM3, Kyung-Soo HAN3


    This study investigated the flow characteristics altered by Jang Bogo Antarctic Research Station using computational fluid dynamics (CFD) modeling. The topography and buildings around Jang Bogo Station were constructed with computer-aided-design data in the CFD model domain. We simulated 16 cases with different inflow directions, and compared the flow characteristics with and without Jang Bogo Station for each inflow direction. The wind data recorded by the site's automatic weather station (AWS) were used for comparison. Wind rose analysis showed that the wind speed and direction after the construction of Jang Bogo Station were quite different from those before construction. We also investigated how virtual wind fences would modify the flow patterns, changing the distance of the fence from the station as well as the porosity of the fence. For westerly inflows, when the AWS was downwind of Jang Bogo Station, the decrease in wind speed was maximized (-81% for west-northwesterly). The wind speed reduction was also greater as the distance of the fence was closer to Jang Bogo Station. With the same distance, the fence with medium porosity (25%-33%) maximized the wind speed reduction. These results suggest that the location and material of the wind fence should be selected carefully, or AWS data should be interpreted cautiously, for particular prevailing wind directions.

    Key words: Jang Bogo Antarctic Research Station; CFD model; observation environment; wind fence; porosity;
    摘要: 本研究运用计算流体力学(CFD)模式探讨韩国南极张保皋科考站的建造对周边气流特征的改变. 在CFD模式中,我们运用计算机辅助制图(CAD)技术构建张保皋科考站周边的地形和建筑物特征.通过设定16 种不同输入气流(盛行风)方向模拟方案, 对每一方向的盛行风,对比研究张保皋科考站建造前后的风场特征.张保皋科考站的自动气象站风场数据用于模拟结果的对比分析.风玫瑰图分析揭示了张保皋科考站建造前后的风向和风速均有显著差异.通过改变虚拟的防风栅栏与张保皋科考站之间的距离以及防风栅栏的孔隙度,进一步研究防风栅栏对气流特征的影响.在盛行西风条件下,自动气象站处于张保皋科考站下风区,科考站的建造使得下风区风速达到最大程度的减小(西到西北风降低约81%).防风栅栏与张保皋科考站距离越近,风速的减小越明显.当防风栅栏与张保皋科考站距离固定,中等孔隙度(25%-33%)的防风栅栏对风速减小的作用最显著.该研究表明科考站周边防风栅栏的位置和材料结构需要慎重选择, 自动气象站风场数据的分析需谨慎,尤其需要结合盛行风的方向.
    (翻译: 徐希燕)
    关键词: 南极张保皋科考站 ; 计算流体力学(CFD)模式 ; 观测环境 ; 防风栅栏 ; 孔隙度
    1. Introduction

    Jang Bogo Antarctic Research Station (hereafter, JB Station) is located near Terra Nova Bay, Northern Victoria Land, Antarctica. Low-pressure systems are common in this region due to the influences of the Ross Sea and Transantarctic Mountains. At JB Station, westerly winds dominate, and daily mean wind speeds range from 0.5 m s-1 to 38.6 m s-1. Because the station is located at the bottom of a mountain slope, strong winds sweeping down the slope (i.e., katabatic winds) are common (Ma, 1992; Nylen et al., 2004; Yu et al., 2007). The maximum wind speed recorded at Terra Nova Bay is 45 m s-1 (Bromwich, 1989), and similarly strong winds could pose a risk to researchers at the station. A survey of visitors to the Antarctic King Sejong Station revealed that one of the most threatening elements at the station was the rapid change in wind speed and direction (Weber et al., 2016). Therefore, it is necessary to establish a plan to protect residents from wind and ensure a secure living environment at the station.

    By reducing wind speed, windbreaks (i.e., wind fences) are effective at reducing the risks caused by strong winds. Most studies on the efficacy of windbreaks have been performed using wind tunnel experiments (Judd et al., 1996; Lee and Kim, 1999; You and Kim, 2009), and have shown that wind speed reductions depend primarily on the shape and porosity of the windbreak (Dong et al., 2007). In addition, (Zhang et al., 2010) and (Cheng et al., 2016) demonstrated that windbreaks can reduce dust-scattering by weakening near-surface wind speeds. (Cheng et al., 2016) reported that windproof walls could reduce wind speeds by up to 80% in the downwind region. However, most studies have focused on the effects of wind fences installed on flat surfaces, and there are few studies concerning wind fences constructed on inclined planes, such as mountain slopes.

    Computational fluid dynamics (CFD) modeling is a useful tool for evaluating the effects of windbreaks on wind speeds in relatively small areas, such as at JB Station. Several recent studies have evaluated the observation environments around weather stations using CFD models (e.g., Tominaga et al., 2004; Stathopoulos, 2006).

    In the present study, we selected a 1-km2 area around JB Station as the model domain to quantify the effects of the construction of JB Station on the area around the site's automatic weather system (AWS) by comparing data collected before and after the construction of JB Station. In addition, we performed numerical experiments to investigate the effects on wind behavior of installing wind fences around JB Station, and assessed the effects of various wind fence porosities and distances from the station. The goal of this study was to determine the optimal porosity and location of the wind fence to reduce wind speeds around JB Station, while minimally impacting wind patterns around the AWS.

    2. Methods
    2.1. Numerical model

    We used the same CFD model in this study as that described in (Kim and Baik, 2010) —— a model that has been extensively validated against wind-tunnel measurement results (Kim, 2007; Kim and Baik, 2010). The model considers a three-dimensional, non-hydrostatic, non-rotating, and incompressible airflow. For the turbulence parameterization, it uses the k-ε turbulence closure scheme based on renormalization group (RNG) theory. Reynolds-averaged Navier-Stokes equations are solved numerically in a staggered grid system using the Semi-Implicit Method for the Pressure-Linked Equation algorithm and finite volume method. To consider the turbulence effects near solid wall boundaries, the model uses wall functions for the momentum, turbulence kinetic energy (TKE), and TKE dissipation rate equations suggested by (Versteeg and Malalasekera, 1995).

    2.2. Experimental setup

    We used a 1-km2 area around JB Station near Terra Nova Bay, Northern Victoria Land, Antarctica (74°37.4'S, 164°13.7'E) as the target area (Fig. 1). The sea is located to the south and east of the station, and 500-700-m high mountains are to the north and west of the station. JB Station was constructed on a slope facing the sea. The red dot in Fig. 1 indicates the location of the AWS to the east of JB Station. The main building, with a height of 15 m, is the tallest building at the station.

    The grid system is similar to the Arakawa C-grid system; except, at the boundaries of the numerical domain, each velocity component (U, V, W) is defined at the center of each face of a control volume, and scalar quantities such as TKE and its dissipation rate are defined at the center of the control volume. At the boundaries, both the velocity components and scalar quantities are defined at the boundary edges of the grid cells. For details about the grid system and numerical method, see (Baik et al., 2003). The model used numerical domain sizes of 1000 m, 1000 m and 250 m for the x-, y- and z-axes, respectively. A uniform grid system was used and the grid intervals for the x-, y- and z-axes were 5 m, 5 m and 2.5 m, respectively. The grid intervals used in this study were relatively coarse compared to those in simulating flows for a single street canyon and/or a single obstacle. Given the much larger numerical domain for this application, we needed to use larger grid sizes due to the limitation of computation. Many previous studies focusing on neighborhood-scale flows in urban areas (horizontal domain size of ∼ 1 km) have used similar or larger grid sizes (∼ 10 m) (Baik et al., 2009; Gousseau et al., 2011; Gowardhan et al., 2011; Hertwig et al., 2012). Thus, the grid intervals in this study were regarded as optimal for resolving both the surrounding mountainous topography and the buildings at JB station.

    Fig.1. Satellite image of the area around JB Station (from Google Earth, www.earth.google.com).

    Using the building construction algorithm suggested in (Baik et al., 2009), and geographic information system tools, we constructed three-dimensional configurations representative of the surface boundary conditions before and after the construction of JB Station (hereafter referred to as JB-before and JB-after cases, respectively) for the CFD model (Figs. 2a and b). Referring to the Cost Action 732 guideline that the horizontal resolution in street canyons should be ≤ 2 m and the vertical domain size should be larger than six times the building height (Eichhorn, 2004), we conducted additional simulations for an idealized street canyon to determine the grid-interval dependency. The results indicated little difference with grid interval (not shown).

    To investigate the effects of the construction of JB Station on the surface wind environment, numerical simulations were performed for 16 inflow directions (northerly to north-northwesterly at intervals of 22.5°) for both the JB-before and JB-after cases. In addition, we examined the effects of installing wind fences to the north and west of JB Station (Fig. 2c). Wind fences with porosities of 0%, 25%, 33%, 50%, 67% and 75%, and distances from the nearest building of 2H, 4H, 6H and 8H (where the wind fence height, H, was 10 m) were considered. Wind fences were explicitly reproduced by heaping up the fences, as in constructing buildings and topography. We assumed that wind fences were vertical-pole in type because this made it easy to allocate porosity systematically (Fig. 2c). We acknowledge that assumed wind-fence shapes are idealized and that there is numerical limitation involved in not taking the real wind-fence shapes into account in the CFD model. To address this issue, we are currently developing a method that implicitly represents the wind-fence effects by adding additional drag terms to the governing equation set.

    Fig.2. Computational configurations (a) before and (b) after the construction of JB Station, and (c) after the construction of wind fences. Boxes in (d) indicate the wind-fence shapes with different porosities considered in this study.

    The wind-fence analysis was performed assuming westerly inflows, based on the predominance of strong katabatic winds at the station. The lengths of the wind fences to the west and north were each 200 m and the length of the wind fence connecting the two wind fences was 64 m. To obtain results sufficiently adjusted by external forcing (e.g., topography and buildings), we integrated the CFD model up to 3600 s with 0.5-s intervals. In our experience, this configuration is satisfactory in neighborhood-scale simulations.

    Based on observations at the Antarctic (Mitsuhashi, 1982), we assumed log profiles at the inflow boundaries. Also, to consider winds blowing from the mountain slopes, the vertical profiles of the inflows were described as follows: \begin{eqnarray} U(z)&=&\dfrac{u_*}{\kappa}\ln\left(\dfrac{z}{z_0}\right)\cos\theta ;\ \ (1)\\[0.5mm] V(z)&=&\dfrac{u_*}{\kappa}\ln\left(\dfrac{z}{z_0}\right)\sin\theta ;\ \ (2)\\[0.5mm] W(z)&=&\dfrac{u_*}{\kappa}\ln\left(\dfrac{z}{z_0}\right)\tan\alpha ; \ \ (3)\end{eqnarray} where \(u_*,\kappa,z_0,\theta\) and α denote friction velocity, the von Kármán constant (0.4), roughness length (0.05), wind direction, and mountain slope (-6°), respectively. For the TKE and TKE dissipation rates, we used the vertical profiles suggested in (Castro and Apsley, 1997): \begin{eqnarray} k(z)&=&\dfrac{u_*^2}{C_\mu^{1/2}}\left(1-\dfrac{z}{\delta}\right)^2 ;\ \ (4)\\[0.5mm] \varepsilon(z)&=&\dfrac{C_\mu^{3/4}k^{3/2}}{\kappa z} .\ \ (5) \end{eqnarray} Here, δ is the boundary layer depth (1000 m) and Cμ is the empirical constant for the RNG \(\kappa-\varepsilon\) turbulence closure scheme (0.0845) (Yakhot et al., 1992). Zero gradient conditions are applied at the outflow boundaries. At the solid wall boundaries, the same boundary conditions as those in (Kim and Baik, 2010) are applied.

    Fig.3. Wind roses (a) before and (b) after the construction of JB Station.

    3. Results and discussion
    3.1. Analysis of observed surface winds

    We examined the hourly average wind data from the AWS near JB Station over about four years (8 February 2010 to 15 February 2015), excluding the period when JB Station was under construction (1 November 2011 to 31 December 2012). To investigate the effects of the presence of JB Station on the AWS observations, we analyzed wind roses for JB-before (8 February 2010 to 31 October 2012) and JB-after (1 January 2013 to 15 February 2015) cases (Fig. 3). The main wind direction differed minimally between the JB-before and JB-after cases. The occurrence frequencies of westerly, west-northwesterly, northerly, and north-northwesterly inflows in the JB-before cases were 21.7%, 13.9%, 12.7% and 12.0%, respectively. In the JB-after cases, westerly (18.9%), north-northeasterly (14.4%), northerly (11.2%), and west-northwesterly (10.4%) flows were the most prevalent wind directions at the AWS. After the construction of JB Station, the annual mean wind speed decreased by ∼ 1.5 m s-1 (4.7 m s-1 to 3.2 m s-1), while wind speed in summer (December, January and February) and winter (June, July and August) decreased by 1.4 m s-1 (4.6 m s-1 to 3.2 m s-1) and 2.1 m s-1 (5.1 m s-1 to 3.0 m s-1), respectively.

    Figure 4 shows the annual mean wind speed for each wind direction in the JB-before and JB-after cases. Westerly flows were the strongest among the 16 wind directions, decreasing by ∼ 35% (8.1 m s-1 to 5.3 m s-1) in the JB-after case. West-northwesterly and west-southwesterly flows were relatively strong, and decreased by ∼ 35% (6.9 m s-1 to 4.5 m s-1) and 27% (6.0 m s-1 to 4.4 m s-1), respectively, in the JB-after case. Easterly flows had a mean wind speed of ∼ 2.0 m s-1. Relatively strong winds were predominantly westerly, due to the strong katabatic winds from the cold slope of the mountain to the west. The maximum observed wind speed was 35.9 m s-1 in a northwesterly direction. The hourly average wind speed analysis identified 136 extremely strong wind events (≥ 20 m s-1), which were mainly westerly. Even considering climatological variations in wind speed in Antarctica, the analysis of the wind data from the AWS revealed that the construction of JB Station markedly affected the surrounding wind environment, and induced a meaningful decrease in near-surface wind speeds. Therefore, the AWS should be moved to a more suitable location to observe surface winds and maintain the climatological continuity of data unaffected by artificial changes to the geographical features around JB Station.

    Fig.4. Wind speeds observed at the AWS before and after the construction of JB Station.

    3.2. Analysis of the flow characteristics in the JB-before and JB-after cases

    To analyze the flow changes induced by the construction of JB Station, numerical simulations were performed for the 16 inflow directions in the JB-before and JB-after cases. In addition, detailed flow characteristics were described for the three inflow directions (westerly, north-northeasterly, and northerly) with strong katabatic winds and relatively high frequency occurrences in the wind rose analysis.

    3.2.1. Wind speed changes at the surface at the station and at the height of the AWS

    We analyzed the effects of the construction of JB Station on near-surface (z=1.25 m) airflow at the station and wind speed and direction at the AWS (z=5 m). Several changes in wind speed were observed at the AWS after the construction of the station (Fig. 5a). Of the westerly winds, westerly, west-southwesterly, west-northwesterly and northwesterly wind speeds at the AWS markedly decreased. The maximum decrease (81.4%) was simulated in the west-southwesterly flow. Conversely, easterly wind speeds changed minimally, with a maximum decrease of 7.2% in the easterly flow. This was likely because the AWS is located east of JB Station, and is downwind of the station only for westerly inflows (not shown). These results are consistent with observations (Fig. 4) at JB Station (JB-before and JB-after cases).

    Fig.5. (a) Percentage wind-speed changes simulated after the construction of JB station, at the AWS (z=5 m) and at the station itself (z=1.25 m) (rectangle in Fig. 1). (b) Comparison of wind directions at the AWS (z=5 m) before and after construction of JB station. In (a), the percentage wind-speed changes at JB Station are averaged over the area indicated by the red rectangle in Fig. 1.

    Fig.6. Horizontal wind vectors and fields of the vertical wind component near the surface (z=1.25 m) (a) before and (b) after the construction of JB Station, and (c) the difference in horizontal wind speed before and after the construction in the westerly case. The wind fields in (b) and (c) are taken from the area in the red-dashed rectangle in (a).

    Wind speeds increased slightly after the construction of the station for five inflow directions: north-northwesterly (3.8%), northerly (0.4%), north-northeasterly (0.4%), southerly (2.1%), and south-southwesterly (5.0%). These increases initially occurred between the station buildings, due to channeling effects (Wang and Takle, 1996; Kim and Kim, 2009), which then enhanced the flows at the AWS. For the other inflow directions, the average surface wind speed around JB Station decreased by 22.9%. From the average wind speeds in the JB-before and JB-after cases, west-northwesterly (-36.9%) and westerly (-34.3%) winds showed particularly substantial reductions in wind speed. These reductions were less prominent for northeasterly (-15.3%) and southerly (-16.9%) winds (Fig. 5a). Wind direction was not notably affected (average change: 4.2°) (Fig. 5b).

    Fig.7. As in Fig. 6 but for the north-northeasterly case.

    Fig.8. As in Fig. 6 but for the northerly case.

    Fig.9. Wind vectors and percentage changes in horizontal wind speed near the surface (z=1.25 m) after the construction of the wind fences for different porosities and distances from JB station in the westerly case. The porosities of the wind fences are 0% [(a) and (d)], 50% [(b) and (e)], and 75% [(c) and (f)], and the distances from JB station are 2H [(a) to (c)] and 8H [(d) to (f)].

    Fig.10. Wind vectors and percentage changes in horizontal wind speed in the vertical plane of y=622.5 m for different porosities and distances from JB station in the westerly case. The porosities of the wind fences are 0% [(a) and (d)], 50% [(b) and (e)], and 75% [(c) and (f)], and the distances from the JB station are 2H [(a) to (c)] and 8H [(d) to (f)].

    3.2.2. Westerly winds (270°)

    The terrain had many overall effects on westerly flows. In the valley located to the west of JB Station, air flowed down and up the west and east slopes of the valley, respectively, and downward flows appeared along the east slope of the valley (Fig. 6a). Flows around JB Station had horizontal wind speeds of 3.5 m s-1 to 4.0 m s-1, and wind speed decreased to the east of JB Station along the coast, although wind speeds were restored farther off the coast. In the JB-after case, the flows around JB Station were more complex due to flow distortions caused by the buildings (Fig. 6b). In the JB-before case, westerly winds flowed to the AWS unobstructed. However, the AWS is located to the east of JB Station, and flows were diverted by the buildings in the JB-after case, which were simulated as changes in wind direction at the AWS (Figs. 6a and b). Moreover, variations in wind speed around JB Station appeared between the JB-before and JB-after cases. On the east side of the buildings, wind speed mainly decreased due to secondary circulations, such as a recirculation zone; however, wind speed between buildings increased slightly due to channeling effects (Fig. 6c). In the JB-after case, the average near-surface (z=1.25 m) wind speed at JB Station decreased by 34.3% compared to the JB-before case, while wind speed at the AWS (z=5.0 m) decreased by 53.3%. The sudden reduction in wind speed between the two cases indicates that the AWS is located within the recirculation zone.

    3.2.3. North-northeasterly winds (22.5°)

    At JB Station, air flows from the ocean are weakened rapidly upon reaching land due to friction, and air flows up the slope on the east side of JB Station. Meanwhile, air flows down the eastern slope and up the northwestern slope of the valley located to the west of JB Station (Fig. 7a). In the JB-after case, the north-northeasterly inflows were distorted by the presence of the buildings (Fig. 7b). In contrast to the westerly winds, much smaller changes occurred in north-northeasterly wind speed (0.02 m s-1) and direction (2.8°) after the construction of the station, due to the upwind location of the AWS relative to the station. The average near-surface (z=1.25 m) wind speed decreased by 24.6% compared with the JB-before case. However, wind speeds to the west of the station increased because the flows were blocked by the maintenance building and intercepted by flows from the northwest direction (Fig. 7c).

    3.2.4. Northerly winds (0°)

    Northerly winds were predominantly downward flows that formed along the downhill slope, although upward flows appeared occasionally along the uphill slope near the station. Flows weakened rapidly near the southern coastline by the station (Fig. 8a). In the JB-after case, the flows were northeasterly to the east of the station and northwesterly to the west of the station, which converged downwind of the buildings (Fig. 8b). Wind speeds increased slightly in between the upper-atmosphere observatory and the main building, as well as the maintenance and main buildings, due to channeling effects (Fig. 8c). The average near-surface (z=1.25 m) wind speeds decreased by 17.3% compared with the JB-before case, while the wind speeds at the AWS (z=5.0 m) increased by 0.4%. Moreover, wind speeds increased to the southwest of the main building compared with the JB-before case, as the northwesterly flows were blocked by the maintenance building.

    3.2.5.Effects of wind fences on the wind environment around JB Station

    Strong katabatic winds are relatively common at JB Station, and our analysis confirmed that the construction of the station has increased wind speeds in some directions around the station. Since wind is the biggest threat to the crews at this Polar station, installation of wind fences may reduce damage to crews and facilities caused by strong winds. Therefore, we investigated the effects of wind fences with various construction parameters on wind speeds at JB Station and at the AWS. For the analysis, wind fences were positioned on the west and north sides of JB Station to target the strongest and most frequent westerly and northerly winds, based on AWS data analysis. We analyzed the effects of fence porosity (0%, 25%, 33%, 50%, 67% and 75%) and distance between the fences and JB Station (2H, 4H, 6H and 8H, where H is the height of the wind fences). Westerly winds were used as the inflow direction, as they are the most frequent and strongest winds at the station.

    Figure 9 shows the surface wind vector field after the construction of the wind fences and the difference between the period after and before construction of the wind fences. Various wind-fence porosities (25%, 33%, 50%, 67% and 75%) were compared with the pre-wind-fence conditions, and the wind fences appeared to reduce wind speed without substantially changing wind direction (Figs. 9b, c, e and f). Lower wind-fence porosities were associated with larger reductions in wind speed in the area between the wind fences and the station, but increases in wind speed were simulated to the east of the station (Figs. 9 and 10). Reductions in wind speeds were significantly larger for the wind fences with a porosity of 0% than those with porosities of 25%, 33%, 50%, 67% and 75% in the windward direction of the fence compared with the conditions before the installation of the wind fence. However, this resulted in the formation of a recirculation area between the wind fences and the main building, which had a flow direction opposite to that of the inflow. In addition, the wind fence inhibited the increases in wind speeds observed in the JB-after case; however, strong winds occurred to the east of the station (Figs. 10a and d).

    In the cases with distance to the station of 2H, wind fences with porosities of 0%, 25%, 33%, 50%, 67% and 75% decreased the average near-surface (z=1.25 m) wind speeds by 15.3%, 17.6%, 17.4%, 16.0%, 10.6% and 7.7%, respectively. As the distance to the station became larger, the average near-surface wind speeds decreased (Fig. 11a). For the fixed distance to the station, the efficiency of wind fences for wind-speed reduction was maximized at porosities of 25% (2H) or 33% (4H-8H). Wind-speed changes at the AWS (z=5.0 m) with distance and porosity showed a non-monotonic variation (Fig. 11b). The installation of wind fences with distance of 2H and porosities of 0%, 25%, 33%, 50%, 67% and 75%, decreased (changed) the wind speeds (directions) by 7.6% (3.5°), 5.8% (2.8°), 5.9% (3.1°), 8.5% (3.2°), 7.5% (2.6°) and 4.7% (2.2°), respectively (Fig. 11b). In the case of a distance of 4H, the wind-speed variation to porosities was similar to that in the cases of 2H. However, in the cases with distance of 8H, wind fences with porosities of 0%, 25%, 33% and 50% increased the wind speeds by about 6.9%, 7.1%, 4.4% and 1.9%, respectively; whereas, wind fences with porosities of 67% and 75% decreased the wind speed by about 2.3% and 1.5%, respectively. Wind direction at the AWS (z =5.0 m) was changed by about 1.3°, 0.5°, 0.7°, 1.0°, 1.2° and 0.9° by the installation of wind fences, with respective porosities of 0%, 25%, 33%, 50%, 67% and 75%. A previous study (Martin, 1995) reported that a reduction of wind speed by vertical wind fences on flat terrain was maximized for porosities of 40%-60%. However, in this study, the lower porosities (25%-33%), apart from the no-porosity cases, resulted in a larger decrease in wind speed around JB Station. This discrepancy was caused by the fact that wind fences were installed on a slightly inclined slope in this study. Analysis of the rates of change in near-surface wind speeds around JB Station showed that the maximum rates of change (increase or decrease) in wind speeds decreased monotonically as wind-fence porosity and distance to JB Station increased (Fig. 12).

    Fig.11. Percentage changes in horizontal wind speed with wind-fence porosity and distance from JB station for (a) at the station itself (z=1.25 m) and (b) at the AWS (z=5.0 m).

    Fig.12. Box plots for the percentage changes in near-surface (z=1.25 m) wind speeds around JB Station in the cases with a distance to JB station of (a) 2H, (b) 4H, (c) 6H, and (d) 8H. The upper and lower black circles indicate the outliers; the bars above and below the boxes indicate the upper and lower extremes, respectively; and the upper, middle and lower segments of boxes indicate the upper quartiles, medians and lower quartiles, respectively.

    4. Summary and conclusions

    In this study, we analyzed the wind environment at JB Station in Terra Nova Bay, Antarctic, including the effects of the construction of JB Station on wind flow and AWS observations, and the effects of installing wind fences on wind behavior around JB Station. To investigate the effects of the construction of the station on wind at the AWS, we analyzed wind roses for the periods before (8 February 2010 to 31 October 2012) and after (1 January 2013 to 15 February 2015) construction of the station. To investigate the effects of JB Station on surface wind, numerical simulations were performed for 16 inflow directions (from northerly to north-northwesterly, at intervals of 22.5°) for both the JB-before and JB-after cases. In the after-JB case, changes in wind speed and wind direction were observed around JB Station and at the AWS. Westerly inflows along the mountain slopes had the highest wind speeds and occurrence frequencies, and the station had the greatest effects on westerly inflows, as the AWS is located to the east of JB Station. According to these results, the current location of the AWS near JB Station is unsuitable for research on katabatic winds. Moreover, if the AWS is transferred to the east or north side of JB Station, more accurate AWS data are expected due to smaller effects of the station.

    Furthermore, we investigated the effects of installing wind fences on wind speed around JB Station. We assessed distances between the wind fences and the station of 2H, 4H, 6H and 8H, and wind fence porosities of 0%, 25%, 33%, 50%, 67% and 75%. Based on the results of the numerical simulation, reductions in near-surface (z=1.25 m) wind speeds increased as the distance between the station and the wind fences decreased (except when porosity = 0%), and wind speed reduction was greatest at porosities of 25%-33%. However, wind fences with a porosity of 0% inhibited the increased flows around the buildings, but increased flows to the east of the station, which greatly affected winds at the AWS. Considering the numerical experiments from this study and results from previous studies, installing wind fences with porosities of 25%-33% would offer the best performance in promoting safety and reducing structural damage at JB Station.

    Acknowledgements. The authors would like to thank the anonymous reviewers for their valuable comments and suggestions. This study was funded by a Korea Polar Research Institute project (PE16250). Hateak KWON is financially supported by PE17010 of Korea Polar Research Institute.


    [1] Baik J.-J., J.-J. Kim, and H. J. S. Fernando, 2003: A CFD model for simulating urban flow and dispersion. J. Appl. Meteor., 42, 1636- 1648, doi: 10.1175/1520-0450(2003)042<1636: ACMFSU>2.0.CO;2.
    A three-dimensional computational fluid dynamics (CFD) model is developed to simulate urban flow and dispersion, to understand fluid dynamical processes therein, and to provide practical solutions to some emerging problems of urban air pollution. The governing equations are the Reynolds-averaged equations of momentum, mass continuity, heat, and other scalar (here, passive pollutant) under the Boussinesq approximation. The Reynolds stresses and turbulent fluxes are parameterized using the eddy diffusivity approach. The turbulent diffusivities of momentum, heat, and pollutant concentration are calculated using the prognostic equations of turbulent kinetic energy and its dissipation rate. The set of governing equations is solved numerically on a staggered, nonuniform grid system using a finite-volume method with the semi-implicit method for pressure-linked equation (SIMPLE) algorithm. The CFD model is tested for three different building configurations: infinitely long canyon, long canyon of finite length, and orthogonally intersecting canyons. In each case, the CFD model is shown to simulate urban street-canyon flow and pollutant dispersion well.
    DOI:10.1175/1520-0450(2003)0422.0.CO;2      URL     [Cited within:1]
    [2] Baik J.-J., S.-B. Park, and J.-J. Kim, 2009: Urban flow and dispersion simulation using a CFD model coupled to a mesoscale model.Journal of Applied Meteorology and Climatology,48,1667-1681,doi: 10.1175/2009JAMC2066.1.
    ABSTRACT Flow and pollutant dispersion in a densely built-up area of Seoul, Korea, are numerically examined using a computational fluid dynamics (CFD) model coupled to a mesoscale model [fifth-generation Pennsylvania State University-National Center for Atmospheric Research Mesoscale Model (MM5)]. The CFD model used is a Reynolds-averaged Navier-Stokes equations model with the renormalization group k 2 turbulence model. A one-way nesting method is employed in this study. MM5-simulated data are linearly interpolated in time and space to provide time-dependent boundary conditions for the CFD model integration. In the MM5 simulation, four one-way nested computational domains are considered, and the innermost domain with a horizontal grid size of 1 km covers the Seoul metropolitan area and its adjacent areas, including a part of the Yellow Sea. The NCEP final analysis data are used as initial and boundary conditions for MM5. MM5 is integrated for 48 h starting from 0300 LST 1 June 2004 and the coupled CFD-MM5 model is integrated for 24 h starting from 0300 LST 2 June 2004. During the two-day period, a high-pressure system was dominant over the Korean peninsula, with clear conditions and weak synoptic winds. MM5 simulates local circulations characterized by sea breezes and mountain/valley winds. MM5-simulated synoptic weather and near-surface temperatures and winds are well matched with the observed ones. Results from the coupled CFD-MM5 model simulation show that the flow in the presence of real building clusters can change significantly as the ambient wind speed and direction change. Diurnally varying local circulations mainly cause changes in ambient wind speed and direction in the present simulation. Some characteristic flows-such as the double-eddy circulation, channeling flow, and vertical recirculation vortex-are simulated. Pollutant dispersion pattern and the degree of lateral pollutant dispersion are shown to be complicated in the presence of real building clusters and under varying ambient wind speed and direction. This study suggests that because of the sensitive dependency of urban flow and pollutant dispersion on variations in ambient wind, time-dependent boundary conditions should be used to better simulate or predict them when the ambient wind varies over the period of CFD model simulation.
    DOI:10.1175/2009JAMC2066.1      URL     [Cited within:2]
    [3] Bromwich D. H., 1989: An extraordinary katabatic wind regime at terra nova bay, Antarctica. Mon. Wea. Rev., 117, 688-695, doi: 10.1175/1520-0493(1989)117<0688:AEKWRA>2. 0.CO;2.
    Abstract Three years of automatic weather station observations for the months of February to April show that intense katabatic winds persistently blow across the western shore of Terra Nova Bay. The data demonstrate that the anomalously strong katabatic winds of Adelie Land are not unique, and thus strongly support the proposition that most of the cold boundary layer air from the ice sheet crosses the coastline in a small number of narrow zones. Furthermore the observations prove that katabatic winds can routinely blow for substantial distances across flat terrain in marked contrast to the abrupt dissipation previously monitored just offshore from East Antarctica. Winter wind conditions onset suddenly in mid-February and are characterized by negligible directional variations and by speeds mostly ranging between 10 and 30 m s 1 . Katabatic winds at Terra Nova Bay both affect and are affected by the regional atmospheric circulation. This katabatic airflow is a time-averaged source of cold boundary layer air for the western Ross Sea. Maximum thermal contrast with the regional temperature field occurs between January and June. Temperature observations suggest that the katabatic winds at Inexpressible Island am primarily of the boratype throughout the year. Strong southerly geostrophic winds over the western Ross Sea appear to suppress the katabatic outflow during winter while weak zonal pressure gradients coincide with intensified katabatic drainage. This relationship is suggested to arise because clouds modulate the radiative production of cold surface air over the interior of the ice sheet.
    DOI:10.1175/1520-0493(1989)1172.0.CO;2      URL     [Cited within:1]
    [4] Castro I. P., D. D. Apsley, 1997: Flow and dispersion over topography: A comparison between numerical and laboratory data for two-dimensional flows.Atmos. Environ.,31,839-850,doi: 10.1016/S1352-2310(96)00248-8.
    Computations of the low and dispersion over two-dimensional hills of various slope and submerged in a neutrally stable boundary layer are described. The results are compared with those of corresponding laboratory experiments undertaken by the U.S. Environmental Protection Agency (Khurshudyan er al. 1981, Report EPA-600/4-81-067). It is shown that a suitably modified k turbulence model generally produces reasonable agreement for the mean flow behaviour, but somewhat lower values for the turbulent kinetic energy and the lateral plume spread. This latter deficiency can be offset by normalising ground-level concentration values by the maximum value obtained in the absence of the hill. The resulting terrain amplification factors are in good agreement with the laboratory data. For a hill slope large enough to generate steady separation. the impact of the recirculating flow on concentration levels is also well predicted although. for somewhat lower slopes when separation is intermittent, results are less satisfactory.
    DOI:10.1016/S1352-2310(96)00248-8      URL     [Cited within:1]
    [5] Cheng J.-J., J.-Q. Lei, S.-Y. Li, and H.-F. Wang, 2016: Disturbance of the inclined inserting-type sand fence to wind-sand flow fields and its sand control characteristics.Aeolian Research,21,139-150,doi: 10.1016/j.aeolia.2016.04.008.
    The inclined inserting-type sand fence is a novel sand retaining wall adopted along the Lanxin High-Speed Railway II in Xinjiang for controlling and blocking sand movement. To verify the effectiveness of the new fence structure for sand prevention, a wind tunnel test was used for flow field test simulation of the sand fence. The results indicate that the inclined inserting-type sand fence was able to deflect the flow of the sand and was able to easily form an upward slant acceleration zone on the leeward side of the sand fence. As shown by the percentage change in sand collection rates on the windward side and the leeward side of the sand fence, the sand flux per unit area at 4-m height in the slant upward direction increased on the leeward side of the inclined inserting-type sand fence. By comparing the flow fields, this site is an acceleration zone, which also reaffirms the correspondence of wind搒and flow fields with the spatial distribution characteristic of the wind-carried sand motion. The field sand collection data indicates that under the effects of the inclined inserting-type sand fence, the sandy air currents passing in front and behind the sand fence not only changed in quality, but the grain composition and particle size also significantly changed, suggesting that the inclined inserting-type sand fence has a sorting and filtering effect on the sandy air currents that passed through. The fence retained coarse particulates on the windward side and fine particulates within the shade of the wind on the leeward side.
    DOI:10.1016/j.aeolia.2016.04.008      URL     [Cited within:2]
    [6] Dong Z. B., W. Y. Luo, G. Q. Qian, and H. T. Wang, 2007: A wind tunnel simulation of the mean velocity fields behind upright porous fences.Agricultural and Forest Meteorology,146,82-93,doi: 10.1016/j.agrformet.2007.05.009.
    Porosity is the most important parameter that determines the efficiency of wind fences. The present study provided a deeper understanding of mean flow regime behind fences with different porosities at different wind velocities by means of a scaled wind tunnel simulation. Velocities were measured using particle image velocimetry and the mean velocity field was obtained and discussed. The mean velocity fields obtained at different wind velocities were similar. Analyzing the streamline patterns revealed an inherent link between fence porosity and mean airflow characteristics behind the fence. The optimal fence porosity is considered to be the critical porosity above or below which airflow characteristics differ strongly. According to the present study, the optimal porosity is found to be around 0.2 or 0.3, which corresponds to a critical porosity above which bleed flow dominates and below which reversed flow becomes significant. The parameters characterizing the reverse cell behind fences were well correlated with porosity. The velocity profiles revealed seven typical flow regions behind fences, characterized by different velocity gradients. The airflow becomes less complicated and the number of flow regions decreases as fence porosity increases. Some regions, especially the reverse cell and small vortex, disappear when the porosity exceeds a certain value. The flow regions gradually merge as the distance downwind increases, and eventually recover a single velocity profile due to downward transfer of momentum from overlying layers. The recovery distance decreases with increasing fence porosity.
    DOI:10.1016/j.agrformet.2007.05.009      URL     [Cited within:1]
    [7] Eichhorn J., 2004: MISKAM-Handbuch zu Version 4 (with update for Version 6). Available online at http://www.lohmeyer.de/ de/system/files/content/download/software/miskam_6_manual _english.pdf.
    [Cited within:1]
    [8] Gousseau P., B. Blocken, T. Stathopoulos, and G. J. F. van Heijst, 2011: CFD simulation of near-field pollutant dispersion on a high-resolution grid: A case study by LES and RANS for a building group in downtown Montreal.Atmos. Environ.,45,428-438,doi: 10.1016/j.atmosenv.2010.09.065.
    Turbulence modeling and validation by experiments are key issues in the simulation of micro-scale atmospheric dispersion. This study evaluates the performance of two different modeling approaches (RANS standard k- and LES) applied to pollutant dispersion in an actual urban environment: downtown Montreal. The focus of the study is on near-field dispersion, i.e. both on the prediction of pollutant concentrations in the surrounding streets (for pedestrian outdoor air quality) and on building surfaces (for ventilation system inlets and indoor air quality). The high-resolution CFD simulations are performed for neutral atmospheric conditions and are validated by detailed wind-tunnel experiments. A suitable resolution of the computational grid is determined by grid-sensitivity analysis. It is shown that the performance of the standard k- model strongly depends on the turbulent Schmidt number, whose optimum value is case-dependent and a priori unknown. In contrast, LES with the dynamic subgrid-scale model shows a better performance without requiring any parameter input to solve the dispersion equation.
    DOI:10.1016/j.atmosenv.2010.09.065      URL     [Cited within:1]
    [9] Gowardhan A. A., E. R. Pardyjak, I. Senocak, and M. J. Brown, 2011: A CFD-based wind solver for an urban fast response transport and dispersion model.Environmental Fluid Mechanics,11,439-464,doi: 10.1007/s10652-011-9211-6.
    In many cities, ambient air quality is deteriorating leading to concerns about the health of city inhabitants. In urban areas with narrow streets surrounded by clusters of tall buildings, called stree
    DOI:10.1007/s10652-011-9211-6      URL     [Cited within:1]
    [10] Hertwig D., G. C. Efthimiou, J. G. Bartzis, and B. Leitl, 2012: CFD-RANS model validation of turbulent flow in a semi-idealized urban canopy.Journal of Wind Engineering and Industrial Aerodynamics,111,61-72,doi: 10.1016/j.jweia. 2012.09.003.
    Urban flow fields computed by two steady Computational Fluid Dynamics models based on the Reynolds-averaged Navier Stokes equations (CFD-RANS) are compared to validation data measured in a boundary-layer wind-tunnel experiment. The numerical simulations were performed with the research code ADREA and the commercial code STAR-CD. Turbulent flow within and above a 1:225-scale wind-tunnel model representing a novel semi-idealized urban complexity represents the test case. In a systematic study the quality of the numerical predictions of mean wind fields is evaluated with a focus on the identification of model strengths and limitations. State-of-the-art validation metrics for numerical models were used to quantify the agreement between the data sets. Based on detailed spatial identification of locations of good or bad comparison the study showed how unsteady flow effects within street canyons are a major cause for discrepancies between numerical and experimental results.
    DOI:10.1016/j.jweia.2012.09.003      URL     [Cited within:]
    [11] Judd M. J., M. R. Raupach, and J. J. Finnigan, 1996: A wind tunnel study of turbulent flow around single and multiple windbreaks, Part I: Velocity fields. Bound.-Layer Meteor., 80, 127-165, doi: 10.1007/BF00119015.
    This paper describes wind-tunnel experiments on the flow around single and multiple porous windbreaks (height H ), sheltering a model plant canopy (height H /3). The mean wind is normal to the windbreaks, which span the width of the wind tunnel. The incident turbulent flow simulates the adiabatic atmospheric surface layer. Five configurations are examined: single breaks of three solidities (low, medium, high; solidity = 1 - porosity), and medium-solidity multiple breaks of streamwise spacing 12 H and 6 H . The experimental emphases are on the interactions of the windbreak flow with the underlying plant canopy; the effects of solidity; the differences in shelter between single and multiple windbreaks; and the scaling properties of the flow. Principal results are: (1) the "quiet zones" behind each windbreak are smaller in multiple than single arrays, because of the higher turbulence level in the very rough-wall internal boundary layer which develops over the multiple arrays. Nevertheless, the overall shelter effectiveness is higher for multiple arrays than single windbreaks because of the "nonlocal shelter" induced by the array as a whole. (2) The flow approaching the windbreak decelerates above the canopy but accelerates within the canopy, particularly when the windbreak solidity is high. (3) A strong mixing layer forms just downwind of the top of each windbreak, showing some of the turbulence and scaling properties of the classical mixing layer formed between uniform, coflowing streams. (4) No dramatic increase in turbulence levels in the canopy is evident at the point where the deepening mixing layer contacts the canopy (around x/H = 3) but the characteristic inflection in the canopy wind profile is eliminated at this point.
    DOI:10.1007/BF00119015      URL     [Cited within:1]
    [12] Kim J.-J., 2007: The effects of obstacle aspect ratio on surrounding flows. Atmosphere, 17, 381- 391.
    URL     [Cited within:1]
    [13] Kim J.-J., D.-Y. Kim, 2009: Effects of a building's density on flow in urban areas.Adv. Atmos. Sci.,26,45-56,doi: 10.1007/s00376-009-0045-9.
    The effects of a building’s density on urban flows are investigated using a CFD model with the RNG κ - 07 turbulence closure scheme. Twenty-seven cases with different building’s density parameters (e.g., building and street-canyon aspect ratios) are numerically simulated. As the building’s density parameters vary, different flow regimes appear. When the street canyon is relatively narrow and high, two counterrotating vortices in the vertical direction are generated. The wind speed along streets is mainly affected by the building’s length. However, it is very difficult to find or generalize the characteristics of the street-canyon flows in terms of a single building’s density parameter. This is because the complicated flow patterns appear due to the variation of the vortex structure and vortex number. Volume-averaged vorticity magnitude is a very good indicator to reflect the flow characteristics despite the strong dependency of flows on the variation of the building’s density parameters. Multi-linear regression shows that the volume-averaged vorticity magnitude is a strong function of the building’s length and the street-canyon width. The increase in the building’s length decreases the vorticity of the street-canyon flow, while, the increase in the streetcanyon width increases the vorticity.
    DOI:10.1007/s00376-009-0045-9      URL     [Cited within:1]
    [14] Kim J.-J., J.-J. Baik, 2010: Effects of street-bottom and building-roof heating on flow in three-dimensional street canyons.Adv. Atmos. Sci.,27,513-527,doi: 10.1007/s00376-009-9095-2.
    Using a computational fluid dynamics model, the effects of street-bottom and building-roof heating on flow in three-dimensional street canyons are investigated. The building and street-canyon aspect ratios are one. In the presence of street-bottom heating, as the street-bottom heating intensity increases, the mean kinetic energy increases in the spanwise street canyon formed by the upwind and downwind buildings but decreases in the lower region of the streamwise street canyon. The increase in momentum due to buoyancy force intensifies mechanically induced flow in the spanwise street canyon. The vorticity in the spanwise street canyon strengthens. The temperature increase is not large because relatively cold above-roof-level air comes into the spanwise street canyon. In the presence of both street-bottom and building-roof heating, the mean kinetic energy rather decreases in the spanwise street canyon. This is caused by the decrease in horizontal flow speed at the roof level, which results in the weakening of the mean flow circulation in the spanwise street canyon. It is found that the vorticity in the spanwise street canyon weakens. The temperature increase is relatively large compared with that in the street-bottom heating case, because relatively warm above-roof-level air comes into the spanwise street canyon.
    DOI:10.1007/s00376-009-9095-2      URL     [Cited within:3]
    [15] Lee S.-J., H.-B. Kim, 1999: Laboratory measurements of velocity and turbulence field behind porous fences.Journal of Wind Engineering and Industrial Aerodynamics,80,311-326,doi: 10.1016/S0167-6105(98)00193-7.
    ABSTRACT Flow characteristics of turbulent wake behind porous fences have been investigated experimentally. The velocity fields were measured using the two-frame PTV method in a circulating water channel. The fence models used in this study have geometric porosity (ε) of 0%, 20%, 40% and 65%, respectively. Each fence model was located in uniform flow whose boundary layer thickness (δ) at the fence location was about 0.1 of the fence height (H). Among the porous fences used in this study, the porous fence with porosity ε=20% shows the maximum reduction of mean streamwise velocity, but it has the highest vertical mean velocity at about x/H=1 location and large turbulence intensity in the near wake region. However, the porous fence with ε=40% has good flow characteristics for abating wind erosion with small turbulent fluctuations and a relatively large reduction in mean velocity. Except for the solid fence (ε=0%), two shear layers develop from the porous fences. As the fence porosity (ε) increases, the height of the shear layer and the streamline curvature decrease. When the porosity (ε) is greater than 40%, there is no re-circulation flow behind the fence due to the strong bleed flow, the Reynolds shear stress is nearly negligible in the entire near-wake region and relatively small turbulent kinetic energies are concentrated in the region just behind the fence (x/H<0.5). When the fence porosity is less than 20%, the Reynolds shear stress and turbulent kinetic energy are strong over the fence and in the shear layer near the reattachment region.
    DOI:10.1016/S0167-6105(98)00193-7      URL     [Cited within:1]
    [16] Ma Y. M., 1992: Preliminary study on vertical velocity caused by katabatic wind in Antarctica and its influence on atmospheric circulation.Adv. Atmos. Sci.,9,247-250,doi: 10.1007/BF02657515.
    DOI:10.1007/BF02657515      URL     [Cited within:1]
    [17] Martin P., 1995: Wind protective fences of PARAWEB compositions. Techtextil-Symposium 1995, Lecture No. 537, 1- 8.
    [Cited within:1]
    [18] Mitsuhashi H., 1982: Measurements of snowdrifts and wind profiles around the huts at Syowa station in Antarctica. Antarctic Record, 75, 37- 56.
    The author joined in the 19th Japanese Antarctic Research Expedition and measured the forms and quantities of snowdrifts accumulated around the existing high floor huts (Observation Hut and Ionosphere Hut) located in the major part of Syowa Station. He also investigated the characteristics of the wind profile near the huts selectively on days when strong wind was blowing by obtaining the roughness length and power index of mean wind profile on the snow-covered ground. The results of measurement are summarized as follows : (1) The snowdrifts around the high floor huts formed a wind-scoop and changed into U type with a sharp ridge line on the lee side. The annual cumulative quantities of snowdrifts, measured in the measuring section on the lee side, were 78.3m^3 and 181.7m^3 around Observation Hut and Ionosphere Hut respectively. (2) The characteristics of the wind profile conformed to the logarithmic law comparatively. The roughness length, Z_0,was in the range from 10^0 to 10^ with 2.210^(m) as the mean value. The power index 伪 ranged from 1/2.9 to 1/7 with 1/4.9 as the mean value.
    URL     [Cited within:1]
    [19] Nylen T. H., A. G. Fountain, and P. T. Doran, 2004: Climatology of katabatic winds in the McMurdo dry valleys, southern Victoria Land, Antarctica. J. Geophys. Res.,109, doi: 10.1029/ 2003JD003937.
    Katabatic winds dramatically affect the climate of the McMurdo dry valleys, Antarctica. Winter wind events can increase local air temperatures by 30℃C. The frequency of katabatic winds largely controls winter (June to August) temperatures, increasing 1℃C per 1% increase in katabatic frequency, and it overwhelms the effect of topographic elevation (lapse rate). Summer katabatic winds are important, but their influence on summer temperature is less. The spatial distribution of katabatic winds varies significantly. Winter events increase by 14% for every 10 km up valley toward the ice sheet, and summer events increase by 3%. The spatial distribution of katabatic frequency seems to be partly controlled by inversions. The relatively slow propagation speed of a katabatic front compared to its wind speed suggests a highly turbulent flow. The apparent wind skip (down-valley stations can be affected before up-valley ones) may be caused by flow deflection in the complex topography and by flow over inversions, which eventually break down. A strong return flow occurs at down-valley stations prior to onset of the katabatic winds and after they dissipate. Although the onset and termination of the katabatic winds are typically abrupt, elevated air temperatures remain for days afterward. We estimate that current frequencies of katabatic winds increase annual average temperatures by 0.7℃ to 2.2℃C, depending on location. Seasonally, they increase (decrease) winter average temperatures (relative humidity) by 0.8℃ to 4.2℃ ( to %) and summer temperatures by 0.1℃ to 0.4℃C (% to %). Long-term changes of dry valley air temperatures cannot be understood without knowledge of changes in katabatic winds.
    DOI:10.1029/2003JD003937      URL     [Cited within:1]
    [20] Stathopoulos T., 2006: Pedestrian level winds and outdoor human comfort.Journal of Wind Engineering and Industrial Aerodynamics,94,769-780,doi: 10.1016/j.jweia.2006.06.011.
    Outdoor human comfort in an urban climate may be affected by a wide range of parameters, including wind speed, air temperature, relative humidity, solar radiation, air quality, human activity, clothing level, age, etc. Several criteria have been developed in the wind engineering community for evaluating only the wind-induced mechanical forces on the human body and the resulting pedestrian comfort and safety. There are significant differences among the criteria used by various countries and institutions to establish threshold values for tolerable and unacceptable wind conditions even if a single parameter, such as the wind speed is used as criterion. These differences range from the speed averaging period (mean or gust) and its probability of exceedance (frequency of occurrence) to the evaluation of its magnitude (experimental or computational). The paper addresses the progress made towards the computational evaluation of pedestrian level winds. All existing criteria for wind and thermal comfort are absolute criteria, which specify the threshold values or comfort ranges for respective weather parameters. The paper will outline an approach towards the establishment of an overall comfort index taking into account, in addition to wind speed, the temperature and relative humidity in the area.
    DOI:10.1016/j.jweia.2006.06.011      URL     [Cited within:1]
    [21] Tominaga Y., A. Mochida, T. Shirasawa, R. Yoshie, H. Kataoka, K. Harimoto, and T. Nozu, 2004: Cross comparisons of CFD results of wind environment at pedestrian level around a high-rise building and within a building complex.Journal of Asian Architecture and Building Engineering,3,63-70,doi: 10.3130/jaabe.3.63.
    ABSTRACT Recently, prediction of the wind environment around a high-rise building using Computational Fluid Dynamics (CFD) has been carried out at the practical design stage. However, very few studies have examined the accuracy of CFD including the velocity distribution at pedestrian level. Thus, a working group for CFD prediction of the wind environment around a building was organized by the Architectural Institute of Japan (AIJ). This group consisted of researchers from several universities and private companies. In the first stage of the project, the working group planned to carry out cross comparison of CFD results of flow around a single high-rise building model placed within the surface boundary layer and flow within a building complex in an actual urban area obtained from various numerical methods. This was done in order to clarify the major factors affecting prediction accuracy. This paper presents the results of this comparison.
    DOI:10.3130/jaabe.3.63      URL     [Cited within:1]
    [22] Versteeg H. K., W. Malalasekera, 1995: An Introduction to Computational Fluid Dynamics: The Finite Volume Method. Longman,Malaysia, 257 pp.
    Versteeg, H K; Malalasekera, W
    DOI:10.1016/j.enganabound.2004.05.002      URL     [Cited within:1]
    [23] Wang H., E. S. Takle, 1996: On shelter efficiency of shelterbelts in oblique wind.Agricultural and Forest Meteorology,81,95-117,doi: 10.1016/0168-1923(95)02311-9.
    ABSTRACT The changes of shelter effects in oblique flows are studied by numerical simulations. The simulated results show that horizontal profiles of wind speed and the location of the minimum wind speed (i.e., maximum wind reduction) move toward the shelterbelt when approach flows depart from the normal, and wind speed may exceed the undisturbed wind speed in the middle lee because of the channeling effect of shelterbelts. With increasing wind incidence angle (IA), the minimum wind speed may decrease or increase, and the rate of decrease in the shelter distance may be faster or slower than cos(IA), both depending on the height of observation and the density and width of shelterbelts. The mean wind speed reduction over 30H (H is shelterbelt height) leeward shows similar characteristics to the shelter distance. The change of shelter effects in oblique flows may result from (i) change of effective shelterbelt density, (ii) different efficiencies in reducing wind speed in directions perpendicular and parallel to the belt, and (iii) change of horizontal wind direction as the flow recovers to the undisturbed direction. The relative importance of each factor, which depends on the height of observation and the density and width of shelterbelts, determines the variation of shelter effects. The simulations produced all the observed qualitative characteristics of shelter effects and for those previous reports giving measured values, the model produced results that are in good agreement.
    DOI:10.1016/0168-1923(95)02311-9      URL     [Cited within:1]
    [24] Weber N. J., M. A. Lazzara, L. K. Keller, and J. J. Cassano, 2016: The extreme wind events in the ross island region of Antarctica.Wea. Forecasting,31,985-1000,doi: 10.1175/WAF-D-15-0125.1.
    Abstract Numerous incidents of structural damage at the U.S. Antarctic Program’s (USAP) McMurdo Station due to extreme wind events (EWEs) have been reported over the past decade. Utilizing nearly 20 yr (~1992–2013) of University of Wisconsin automatic weather station (AWS) data from three different stations in the Ross Island region (Pegasus North, Pegasus South, and Willie Field), statistical analysis shows no significant trends in EWE frequency, intensity, or duration. EWEs more frequently occur during the transition seasons. To assess the dynamical environment of these EWEs, Antarctic Mesoscale Prediction System (AMPS) forecast back trajectories are computed and analyzed in conjunction with several other AMPS fields for the strongest events at McMurdo Station. The synoptic analysis reveals that McMurdo Station EWEs are nearly always associated with strong southerly flow due to an approaching Ross Sea cyclone and an upper-level trough around Cape Adare. A Ross Ice Shelf air stream (RAS) environment is cr...
    DOI:10.1175/WAF-D-15-0125.1      URL     [Cited within:1]
    [25] Yakhot V., S. A. Orszag, S. Thangam, T. B. Gatski, and C. G. Speziale, 1992: Development of turbulence models for shear flows by a double expansion technique.Physics of Fluids A: Fluid Dynamics,4,1510-1520,doi: 10.1063/1.858424.
    DOI:10.1063/1.858424      URL     [Cited within:1]
    [26] You K.-P., Y.-M. Kim, 2009: Effect of protection against wind according to the variation porosity of wind fence.Environmental Geology,56,1193-1203,doi: 10.1007/s00254-008-1219-y.
    This study examines shelter effect against the wind by using wind fence with various porosities and distance. The shelter effect of wind fence was investigated by a wind tunnel test. Flow characteristics of velocities and turbulences behind wind fence were measured using a hot-wire anemometer. This was done by varying the porosity by 0, 20, and 40% of the wind fence. The wind fence distance ranged from 1 H to 9 H . In addition, the overall characterization of the wind fence was investigated by measuring a total of 28 points on the wind fence, which forms a lattice structure on it with 7 points in the lateral direction and 4 points in the vertical direction. The results indicate that the degree of the turbulence is lowered and the velocity of the wind is decreased when porosity of 40% is used at a distance of 4 H –7 H . The effectiveness of the wind fence depends on the porosity and distance. Porosity of 20% proved to be effective for the protection area of 1 H –3 H , while that of 40% was effective for the protection area of 4 H –6 H .
    DOI:10.1007/s00254-008-1219-y      URL     [Cited within:1]
    [27] Yu Y., X. M. Cai, and X. S. Qie, 2007: Influence of topography and large-scale forcing on the occurrence of katabatic flow jumps in Antarctica: Idealized simulations.Adv. Atmos. Sci.,24,819-832,doi: 10.1007/s00376-007-0819-x.
    姝he Regional Atmospheric Modeling System(RAMS),which is a non-hydrostatic numerical model,has been used to investigate the impact of terrain shape and large-scale forcing on the Antarctic surface-wind regime,focusing on their roles in establishing favorable flow conditions for the formation of katabatic flow jumps.A series of quasi-2D numerical simulations were conducted over idealized slopes representing the slopes of Antarctica during austral winter conditions.Results indicate that the steepness and variations of the underlying slope play a role in the evolution of near-surface flows and thus the formation of katabatic flow jumps.However,large-scale forcing has a more noticeable effect on the occurrence of this small-scale phenomenon by establishing essential upstream and downstream flow conditions,including the upstream supercritical flow,the less stably stratified or unstable layer above the cold katabatic layer,as well as the cold-air pool located near the foot of the slope through an interaction with the underlying topography. Thus,the areas with steep and abrupt change in slopes,e.g.near the coastal areas of the eastern Antarctic, are preferred locations for the occurrence of katabatic flow jumps,especially under supporting synoptic conditions.
    DOI:10.1007/s00376-007-0819-x      URL     [Cited within:1]
    [28] Zhang N., J.-H. Kang, and S.-J. Lee, 2010: Wind tunnel observation on the effect of a porous wind fence on shelter of saltating sand particles.Geomorphology,120,224-232,doi: 10.1016/ j.geomorph.2010.03.032.
    A porous wind fence is an artificial barrier widely employed to abate wind erosion. This study investigated the shelter effect of a porous wind fence on saltating sand in a simulated atmospheric boundary layer (ABL). A wind fence with a porosity ε 02=0238.5% was installed on a flat bed of sand collected from a beach (diameter, d 02=02200–30002μm). A high-speed digital camera was used to capture consecutive images of saltating sand particles around the fence at a frame rate of 4000 frames per second (fps). In addition, the particle tracking velocimetry (PTV) method was employed to extract the instantaneous velocity fields of saltating sand particles. From these data, the mean velocity and volume concentration of saltating sand, mass flux, and kinetic energy were evaluated. As a result, the mean velocities decrease dramatically on the leeward side of the fence, and a high-velocity region exists in the shear layer above the fence. The sand mass flux distributions with height around the fence are represented by an exponential function. Both the particle concentration and mass flux decay largely in the leeward region. The present experimental results can provide useful information to understand sand transport through a porous fence and allow the creation of a new control measure of wind erosion of sand particles.
    DOI:10.1016/j.geomorph.2010.03.032      URL     [Cited within:1]
    RichHTML Viewed    
    Abstract Viewed    


    Related articles:
    Key words
    Jang Bogo Antarctic Research Station
    CFD model
    observation environment
    wind fence

    Jang-Woon WANG
    Jae-Jin KIM
    Wonsik CHOI
    Da-Som MUN
    Jung-Eun KANG
    Hataek KWON
    Jin-Soo KIM
    Kyung-Soo HAN