Advanced
Seasonal effectiveness of a Korean traditional deciduous windbreak in reducing wind speed
Seasonal effectiveness of a Korean traditional deciduous windbreak in reducing wind speed
Journal of Ecology and Environment. 2014. May, 37(2): 91-97
Copyright © 2014, The Ecological Society of Korea
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial Licens (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • Received : February 18, 2014
  • Accepted : May 13, 2014
  • Published : May 27, 2014
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Insu Koh
Gund Institute for Ecological Economics, The University of Vermont, Burlington, Vermont 05403, USA.
Chan-Ryul Park
Divison of Forest Ecology, Korea Forest Research Institute, Seoul 130-712, Korea
Wanmo Kang
Divison of Forest Ecology, Korea Forest Research Institute, Seoul 130-712, Korea
Dowon Lee
Graduate School of Environmental Studies, Seoul National University, Seoul 151-742, Korea
leedw@snu.ac.kr

Abstract
Little is known about how the increased porosity of a deciduous windbreak, which results from loss of leaves, influences wind speed reduction. We hypothesized that, with loss of foliage, the wind speed reduction effectiveness of a deciduous windbreak decreases on near leeward side but not on further leeward side and that wind speed recovers faster in the full foliage season than in other seasons. During summer, autumn, and winter (full, medium, and non- foliage season, respectively), we observed wind speed and direction around a deciduous windbreak in a traditional Korean village on windward and near and further leeward sides (at –8H, 2H, and 6H; H = 20 m, a windbreak height). We used a linear mixed effects model to determine that the relative wind speed reduction at 2H significantly decreased from 83% to 48% ( F 2,111.97 = 73.6, P < 0.0001) with the loss of foliage. However, the relative wind speed reduction at 6H significantly increased from 26% to 43% ( F 2,98.54 = 18.5, P < 0.0001). Consequently, wind speed recovery rate between 2H and 6H in summer was two times higher than in autumn and ten times higher than in winter ( F 2,102.93 = 223.1, P < 0.0001). These results indicate that deciduous windbreaks with full foliage seem to induce large turbulence and increase wind speed recovery rate on leeward side. Our study suggests that further research is needed to find the optimal foliage density of a deciduous windbreak for maximizing windbreak effectiveness regardless of seasonal foliage changes.
Keywords
INTRODUCTION
Windbreaks or shelterbelts are used to reduce wind speed and alter microclimatic and soil conditions in agricultural landscapes throughout the world ( Brandle et al. 2004 ). The primary effect of windbreaks is reduction of wind speed, which leads to secondary effects such as enhancement of the microclimate for crop yield and protection of livestock, homes and soils from strong winds. The effectiveness of a windbreak depends not only on external structural characteristics such as height, width, length and orientation, but also on internal characteristics such as density (the amount of solid material) and porosity (the amount of open space). Of these characteristics, windbreak height and porosity are the two primary factors that determine a windbreak’s effectiveness in reducing wind speed ( McNaughton 1988 , Brandle et al. 2004 ).
Several studies have tried to examine the relationship between the height and porosity of windbreaks and their effectiveness on wind speed reduction ( Heisler and Deoriwalle 1988 , McNaughton 1988 , Zhang et al. 1995a ). The windbreak height (H) is important to wind speed reduction because it determines horizontal sheltered area. This sheltered area extends for a distance of 5H in the up-wind of the windbreak (windward) and a distance of 30H in the down-wind (leeward) ( Brandle et al. 2004 ). Within this range, there exist two distinct zones called ‘quiet’ and ‘wake’ ( McNaughton 1988 ). In the quiet zone, which is immediately behind a windbreak regardless of internal structures, turbulence is reduced and eddy size is small. Beyond that, in the wake zone, turbulence increases with eddy size and wind speed returns to upwind scale ( McNaughton 1988 ). The windbreak porosity, which is defined as the ratio or percentage of pore space to the space occupied by tree stems, braches, twigs, and leaves ( Loeffler et al. 1992 , Zhu et al. 2003 ), mediates the effectiveness of wind speed reduction and the extent of the two different zones by influencing the rate of airflow passing through the windbreak pores ( McNaughton 1988 , Brandle et al. 2004 ). Therefore, although height is important, porosity is considered the most important factor in determining horizontal patterns of wind speed reduction such as the position of minimum wind speed and the recovery rate of wind speed ( Heisler and Dewalle 1988 , Cleugh 1998 , Brandle et al. 2004 ).
Many simulation and experimental studies have shown that a medium dense (i.e., medium porosity) windbreak generally shows better windbreak effectiveness than a high dense (i.e., low porosity) windbreak ( Brandle et al. 2004 ). Such studies have suggested that although the low porosity windbreaks reduce wind speed near leeward side more than medium porosity windbreaks, low porosity windbreaks also induce more turbulence leeward side, slightly faster recovery rate of wind speed, and narrower quiet zone when compared to medium porosity windbreaks with the largest sheltered distances ( Wang and Takle 1996 , Vigiak et al. 2003 ). Most of these results are based on the studies of coniferous windbreaks that maintain a similar foliage density regardless of season. Consequently, to maximize windbreak effectiveness, windbreak design for coniferous trees has focused on the spatial arrangement of tree plantation ( Skidmore and Hagen 1970 , Hagen et al. 1981 , Heisler and Dewalle 1988 , Zhang et al. 1995b , Wu et al. 2013 ). Although fewer studies have focused on deciduous windbreaks, researchers have recognized that the amount of foliage, which determines deciduous windbreak density, varies with seasonal phonological stages ( Finch 1988 , Loeffler et al. 1992 ). Understanding the effects of the phenology stages of deciduous windbreaks on wind speed reduction and sheltered distances is important knowledge that can contribute to designing and managing windbreaks for optimal effectiveness ( Finch 1988 , Zhou et al. 2004 ) by minimizing the seasonal variability.
There is, however, a knowledge gap on how the seasonal phenology stages of deciduous windbreaks influence wind speed reduction. Generally, the loss of foliage of a deciduous windbreak results in the decrease of windbreak effectiveness. For example, a recent study of a deciduous windbreak showed that wind speed reduction on a further leeward side as well as a near leeward side simply decreased with decreasing foliage ( Středová et al. 2012 ). However, it is also possible that, when compared to medium or non-foliage windbreaks, a full foliage deciduous windbreak would induce more turbulance, thus further increasing wind speed recovery rate on leeward side, much like very dense coniferous windbreaks. With such effects, the deciduous windbreak with full folage is not likely to show similar windbreak effectiveness on a further leeward side and a near leeward side. Therefore, further research is needed to clarify whether medium or nonfoliage of deciduous windbreaks has more influence in reducing turbulence than a full foliage one. In this paper, we hypothesized that the wind speed reduction effectiveness of a deciduous windbreak decreases near leeward side but not on further leeward side with loss of foliage. To highlight the increase of turbulence on leeward side by high dense deciduous windbreaks, we also hypothesized that wind speed recovers faster in the full foliage season than in other seasons.
MATERIALS AND METHODS
- Study site and field measurements
To test our hypotheses, we selected a deciduous windbreak in the middle of an apple orchard in Sachon-ri, Uiseong-gun, Korea. This orchard was established by villagers about 600 years ago to protect the village from the western winds ( Fig. 1 ). The windbreak is one type of Korean traditional village grove called a bibosoop from the perspective of fengshui (literally meaning: wind and water, which is translated as geomancy in English) ( Park et al. 2006 , Lee et al. 2007 , Koh et al. 2010 ). Because the windbreak has been preserved as a natural monument in Korea, it has been strictly excluded from artificial managements; thus, its foliage becomes highly dense in summer, medium dense in fall, and non-foliage in winter. Its structure is as follows: 800 m long in North and South orientation, 40 m wide on average, and 20 m high (hereafter, H = 20 m). The total tree density of DBH (diameter breast height) in the maeulsoop area is 46 m/ha: which is composed of Quercus aliena (17.9 m/ha), Q. acutissima (9.4 m/ha), Sophora japonica (6.9 m/ha), Cornus walteri (4.6 m/ha), Celtis sinensis (2.9 m/ha), Zelkova serrata (2.8 m/ ha), and Hemiptelea davidii (1.4 m/ha) ( Choi et al. 2011 ).
Wind speed and direction sensors (HOBO S-WCAM003; Onset computer, Bourne, MA, USA) were installed at three different locations at a height of 4.5 m, which is approximately 1.5 m above the orchard of 3 m high apple trees. At –8H in windward side, wind speed and direction were measured from June 2007 to April 2008. At 2H and 6H on leeward side, wind speed and direction were measured from August 2007 to April 2008 ( Fig. 1 ). The data were collected every 3 seconds and logged by HOBO H21-002 Micro Station (Onset computer) in hourly averaged values. The hourly averaged data were categorized into three seasons each with two months period to keep similar foliage density across each season, which was high foliage (summer: between August and September 2007), medium foliage (autumn: between October and November 2007), and non-foliage (winter: between January and February 2008). This hourly averaged data was used to analyze the wind speed reduction by the windbreak.
PPT Slide
Lager Image
Study site description. (a) The location of study site (white box) in Sachon-ri, Uiseong-gun, Korea, (b) an aerial photograph of the three measurement locations of wind speed and direction in windward side (–8H, H = 20 m) and in leeward side (2H and 6H) of the deciduous windbreak on (captured 2010 from Daum map, http://map.daum.net), and (c) a photo of the automatic weather station installation at 2H (captured on 02 June 2007 by D. Lee).
- Relative wind speed reduction and wind speed recovery rate
We used relative wind speed of the approaching wind speed at –8H (U xH / U –8H ) to measure a relative wind speed reduction at two leeward locations ( Zhang et al. 1995b ). Therefore, the relative wind speed reduction was determined by (U –8H – U xH ) / U –8H (here, xH means 2H and 6H). The wind speed recovery rate between 2H and 6H was determined by the difference between the relative wind speeds at 2H and 6H (i.e., U 6H / U –8H – U 2H / U –8H ).
The relative wind speed reduction and wind speed recovery rates for each season (summer, autumn, and winter) were calculated using the hourly averaged wind speed data. To reduce the variability of approaching wind direction and speed, we identified the times when the approaching wind at –8H was blowing from the west (i.e., between 247.5 and 292.5 longitudinal degrees) and the wind speed was over 1 m/s. For the identified times when the hourly averaged western wind speed data at –8H was over 1.0 m/s, hourly averaged wind speed data at 2H and 6H were used for further analysis of wind speed reduction.
- Statistical analysis
A linear mixed effects model was used to test our hypothesis on the effects of seasonal foliage changes on relative wind speed reduction and wind speed recovery rate between 2H and 6H on the leeward side. In the mixed effects model, we determined season to be a fixed effect. To reduce the pseudo replication of hourly averaged wind data collected on the same date that may be potentially correlated, we determined date of wind measurements as a random effect. Pair-wise comparisons of significant differences of least squared means among seasons (i.e., the level of foliage) were investigated using Student’s t test. In all cases, the degree of freedom of the models was estimated by Satterthwaite’s approximation method in order to account for the unequal number wind speed sampling data sets from the three seasons ( Littell et al. 1996 ). The linear mixed effects model and the pair-wise comparisons were performed using the functions of ‘lmer’ in the R package ‘lme4’ (http://cran.r-project.org/web/packages/ lme4) and the function of ‘difflsmeans’ in the R package ‘lmeTest’ (http://www.cran.r-project.org/web/packages/ lmerTest), respectively, which produce the same data frame as SAS software gives in PROC MIXED statement.
RESULTS
The prevailing wind was blowing from the west in the study site during the observation period ( Fig. 2 ). For example, between 10:00 and 18:00 at –8H, western winds occupied 30% (mean ± SD: 1.06 ± 0.52 m/s, maximum 2.60 m/s) in the summer, 60% (1.40 ± 0.95 m/s, maximum 4.08 m/s) in the autumn, and 67.5% (2.75 ± 1.55 m/s, maximum 6.12 m/s) in the winter. For western wind at – 8H, the number of times that the collected hourly averaged wind speed data had a speed beyond 1.0 m/s was 25 (in 10 days) in the summer, 180 (in 28 days) in the autumn, and 431 (in 31 days) in the winter.
PPT Slide
Lager Image
Wind speed and direction at –8H (H = 20 m) in the windward side during June 2007 to April 2008. Each circle indicates the frequency of wind direction during the observed period. Asterisk mark (*) indicates the wind frequency of the wind speed below 0.5 m/s (calms) during the observation period.
The relative wind speed reduction on the near leeward side (2H) significantly decreased with the loss of foliage ( F 2, 111.97 = 73.6, P < 0.0001). The mean values of the measured relative wind speed reduction at 2H in the summer, autumn, and winter were 82.75%, 67.29%, and 47.83%, respectively (all-paired Student’s t test, P < 0.0001; see Fig. 3a ). However, the relative wind speed reduction on further leeward side (6H) significantly increased with the loss of foliage ( F 2, 98.54 = 18.5, P < 0.0001). The mean values of the measured relative wind speed reduction at 6h in the same three seasons were 24.63%, 36.24%, and 43.30%, respectively ( Fig. 3b ). Consequently, the estimated least squared mean of the wind speed recovery rate between 2H and 6H in summer was significantly high (60.01%), but the wind speed recovery rate was decreased in autumn (32.13%) and winter (5.97%) ( F 2, 102.93 = 223.1, P < 0.0001; all-paired Student’s t test, P < 0.0001; see Fig. 3c ).
PPT Slide
Lager Image
Relative wind speed reduction and wind speed recovery rate in the leeward side of the maeulsoop. (a) Relative wind speed reduction at 2H, (b) relative wind speed reduction at 6H (H = 20 m), and (c) wind speed recovery rate between 2H and 6H for three different seasons (foliage density). Different letters on bars indicate significant differences of least square means using Student’s t test.
DISCUSSION
Our study found that the seasonal foliage changes of a deciduous windbreak influenced the spatial pattern of wind speed reduction. The deciduous windbreak with full foliage showed more effect on wind speed reduction than one with medium or non- foliage on the near leeward side (2H). However, the effect of the full foliage windbreak on the wind sped reduction was less than a medium or non- foliage one on the further leeward side (6H) because the full foliage windbreak showed a much greater wind speed recovery rate. Our results indicate that the deciduous windbreak with full foliage seems to induce large turbulence and the deciduous windbreak with non-foliage seems to induce small turbulence on leeward side. This suggests that the amount of tree trunk and branches of the deciduous windbreak in winter had windbreak porosity more appropriate to maintaining the wind speed reduction effect on a further leeward side.
Although further research is need to examine the relationship between foliage density and porosity and the spatial pattern of wind speed reduction in further areas (0H to 30H), the wind speed reduction pattern can be conceptually derived from the current results ( Fig. 4 ). If these wind speed reduction patterns occur, overall sheltered distance can increase in winter more than in summer. Therefore, it is essential to find an optimal porosity that maximizes the windbreak effectiveness of deciduous trees which vary seasonally. Finding an optimal porosity will help villagers to get the most substantial benefit from the windbreak.
PPT Slide
Lager Image
The conceptual spatial patterns of relative wind speed between –5H and 30H (H = 20 m) according to seasons. The observed relative wind speed at 2H and 6H in each season is illustrated as box plots. Village with a line indicates the spatial location of village in the leeward side of the windbreak.
These results suggest that foliage density management may be especially important for deciduous windbreaks. In many guidelines for designing windbreaks, the arrangement and the width of tree planting have received considerable attention for both coniferous and deciduous tree types ( Fewin and Helwig 1988 , Finch 1988 , Bitog et al. 2012 ), and our study suggests that maintaining proper porosity by pruning (or less likely thinning) may be needed for deciduous windbreaks, which may be the only acceptable trees in certain soil types and are sometimes preferred by farmers in agricultural landscapes ( Tuskan and Laughlin 1991 ). A wind tunnel experiment also supported pruning treatment, showing that the maximum windbreak effectiveness in an optimal degree of porosity was enhanced by a trimming treatment ( Vollsinger et al. 2005 ).
Pruning can be introduced as one of the methods to maintain proper porosity for enhancing the effect of a deciduous windbreak on wind speed reduction. In addition to enhancing the main effect, maintaining proper porosity is also important for enhancing secondary windbreak effects that mitigate microclimates such as increasing temperature and humidity ( Cleugh 1998 ) , reducing evaporation in paddy field, increasing crop production ( Foereid et al. 2002 ), and enhancing odor dispersion ( Lin et al. 2006 ). Along with these effects, pruning helps to protect trees from possible disease by removing unhealthy branches ( Mika 1986 , Canellas and Montero 2002 ). In addition, it may enhance crop yield by reducing the area under the shade near windbreaks ( Sudmeyer and Speijers 2007 ). It is difficult, however, to find the best aerodynamic porosity related to windbreak effectiveness. Recently developed optical stratification porosity methods for estimating aerodynamic porosity may help determine the appropriate amount of pruning for the dense deciduous windbreak ( Zhu et al. 2003 ). Pruning should also be approached carefully in the aspect of period and the quantity because of potential damage on habitats for diverse wildlife ( Leal et al. 2013 ).
In conclusion, this study found that the reduced foliage density of an old and dense deciduous grove enhanced windbreak effectiveness. Based on this result, we suggest that the management of foliage density through pruning is needed for windbreaks that have characteristics similar to our study windbreak. We also suggest further research on the relationship between characteristics of broadleaves and aerodynamic porosity in order to determine the appropriate amount of pruning for the old, dense deciduous windbreaks.
Acknowledgements
This study was supported by the Environmental Planning Institute of Seoul National University and Korea Forest Research Institute, Korea. We would like to thank a graduate student, Kwanghoon Choi, and the residents and foreman of village Sachon-ri for cooperation in installing and maintaining the Automatic Weather Station. We also especially thank to Dr. Pinkert at Humboldt State University for her great assistance in editing this paper.
References
Bitog JP , Lee IB , Hwang HS , Shin MH , Hong SW , Seo IH , Kwon KS , Mostafa E , Pang ZZ 2012 Numerical simulation study of a tree windbreak Biosyst Eng http://dx.doi.org/10.1016/j.biosystemseng.2011.10.006 111 40 - 48
Brandle JR , Hodges L , Zhou XH 2004 Windbreaks in north american agricultural systems Agroforest Syst 61 65 - 78
Canellas I , Montero G 2002 The influence of cork oak pruning on the yield and growth of cork Ann for Sci http://dx.doi.org/10.1051/forest:2002061 59 753 - 760
Choi J-U , Kim D-Y , Kim M-H , Kang B-H , Jeong M-C , Jo L-W , Kim S-B 2011 Current state of the roadside forest in sachon-ri, uiseong and the perspectives on the name of the natural monuments J Korean Inst Trad Landsc Arch 29 52 - 60
Cleugh HA 1998 Effects of windbreaks on airflow, microclimates and crop yields Agroforest Syst http://dx.doi.org/10.1023/A:1006019805109 41 55 - 84
Fewin RJ , Helwig L 1988 Windbreak renovation in the american great plains Agr Ecosyst Environ http://dx.doi.org/10.1016/0167-8809(88)90046-1 22/23 571 - 582
Finch SJ 1988 Field windbreaks - design criteria Agr Ecosyst Environ http://dx.doi.org/10.1016/0167-8809(88)90020-5 22/23 215 - 228
Foereid B , Bro R , Mogensen VO , Porter JR 2002 Effects of windbreak strips of willow coppice - modelling and field experiment on barley in denmark Agr Ecosyst Environ http://dx.doi.org/10.1016/S0167-8809(02)00007-5 93 25 - 32
Hagen LJ , Skidmore EL , Miller PL , Kipp JE 1981 Simulation of effect of wind barriers on airflow Trans ASAE http://dx.doi.org/10.13031/2013.34381 24 1002 - 1008
Heisler GM , Dewalle DR 1988 Effects of windbreak structure on wind flow Agr Ecosyst Environ http://dx.doi.org/10.1016/0167-8809(88)90007-2 22/23 41 - 69
Koh I , Kim S , Lee D 2010 Effects of bibosoop plantation on wind speed, humidity, and evaporation in a traditional agricultural landscape of korea: Field measurements and modeling Agr Ecosyst Environ http://dx.doi.org/10.1016/j.agee.2009.10.008 135 294 - 303
Leal AI , Correia RA , Palmeirim JM , Granadeiro JP 2013 Does canopy pruning affect foliage-gleaning birds in managed cork oak woodlands? Agroforest Syst http://dx.doi.org/10.1007/s10457-012-9556-y 87 355 - 363
Lee D , Koh I , Park C Y 2007 Ecosystem services of traditional village groves in korea Seoul National University Seoul
Lin XJ , Barrington S , Nicell J , Choiniere D , Vezina A 2006 Influence of windbreaks on livestock odour dispersion plume in the field Agr Ecosyst Environ http://dx.doi.org/10.1016/j.agee.2006.02.014 116 263 - 272
Littell RC , Milliken GA , Stroup WW , Wolfinger RD 1996 SAS system for mixed models SAS Institute Inc North Carolina
Loeffler AE , Gordon AM , Gillespie TJ 1992 Optical porosity and windspeed reduction by coniferous windbreaks in southern ontario Agroforest Syst http://dx.doi.org/10.1007/BF00053117 17 119 - 133
McNaughton KG 1988 Effects of windbreaks on turbulent transport and microclimate Agr Ecosyst Environ http://dx.doi.org/10.1016/0167-8809(88)90006-0 22/23 17 - 39
Mika A 1986 Physiological responses of fruit trees to pruning. In: Horticultural reviews (Janick J, ed) John Wiley & Sons Inc. Hoboken, NJ 337 - 378
Park C-R , Shin J-H , Lee D 2006 Bibosoop: A unique korean biotope for cavity nesting birds J Ecol Field Biol http://dx.doi.org/10.5141/JEFB.2006.29.2.075 29 75 - 84
Skidmore EL , Hagen LJ 1970 Evaporation in sheltered areas as influenced by windbreak porosity Agr Meteorol http://dx.doi.org/10.1016/0002-1571(70)90032-4 7 363 - 374
Středová H , Podhrázská J , Litschmann T , Středa T , Rožnovský J 2012 Aerodynamic parameters of windbreak based on its optical porosity Contrib Geophysics Geodesy 42 213 - 226
Sudmeyer RA , Speijers J 2007 Influence of windbreak orientation, shade and rainfall interception on wheat and lupin growth in the absence of below-ground competition Agroforest Syst http://dx.doi.org/10.1007/s10457-007-9070-9 71 201 - 214
Tuskan GA , Laughlin K 1991 Windbreak species performance and management-practices as reported by montana and north-dakota landowners J Soil Water Conserv 46 225 - 228
Vigiak O , Sterk G , Warren A , Hagen LJ 2003 Spatial modeling of wind speed around windbreaks Catena http://dx.doi.org/10.1016/S0341-8162(03)00018-3 52 273 - 288
Vollsinger S , Mitchell SJ , Byrne KE , Novak MD , Rudnicki M 2005 Wind tunnel measurements of crown streamlining and drag relationships for several hardwood species Can J Forest Res http://dx.doi.org/10.1139/x05-051 35 1238 - 1249
Wang H , Takle ES 1996 On three-dimensionality of shelterbelt structure and its influences on shelter effects Bound-Layer Meteorol http://dx.doi.org/10.1007/BF00120076 79 83 - 105
Wu T , Yu M , Wang G , Wang Z , Duan X , Dong Y , Cheng X 2013 Effects of stand structure on wind speed reduction in a Metasequoia glyptostroboides shelterbelt Agroforest Syst http://dx.doi.org/10.1007/s10457-012-9540-6 87 251 - 257
Zhang H , Brandle JR , Meyer GE , Hodges L 1995 A model to evaluate windbreak protection efficiency Agroforest Syst http://dx.doi.org/10.1007/BF00704868 29 191 - 200
Zhang H , Brandle JR , Meyer GE , Hodges L 1995 The relationship between open windspeed and windspeed reduction in shelter Agroforest Syst http://dx.doi.org/10.1007/BF00711717 32 297 - 311
Zhou XH , Brandle JR , Mize CW , Takle ES 2004 Three-dimensional aerodynamic structure of a tree shelterbelt Definition, characterization and working models Agroforest Syst 63 133 - 147
Zhu JJ , Matsuzaki I , Gonda Y 2003 Optical stratification porosity as a measure of vertical canopy structure in a Japanese coastal forest For Ecol Manage http://dx.doi.org/10.1016/S0378-1127(01)00813-1 173 89 - 104