Planning Strategies of Wind Corridor Forests Utilizing the Properties of Cold Air

land

Article Planning Strategies of Wind Corridor Forests Utilizing the Properties of Cold Air

Uk-Je Sung 1, Jeong-Hee Eum 1,* , Jeong-Min Son 1 and Jeong-Hak Oh 2

1 Department of Landscape Architecture, Kyungpook National University, 80 Daehak-ro, Buk-gu, Daegu 41566, Korea; [email protected] (U.-J.S.); [email protected] (J.-M.S.) 2 Division of Urban Forests, National Institute of Forest Science, 57 Hoegi-ro, Dongdaemun-gu, Seoul 02455, Korea; [email protected] * Correspondence: [email protected]; Tel.: +82-53-950-5780

Abstract: A wind corridor forest is defined as an urban forest for utilizing the functions of a wind corridor that allow “cool and fresh air (cold air)” generated in forests at night to flow to urban development areas. This study aims to provide planning strategies for implementing a wind corridor forest by analyzing current conditions in Haengbok City (HBC region), Sejong, South Korea. The HBC region had many wind-generating forests (WGF), wind-spreading forests (WSF), and wind- connecting forests (WCF), and secured the connections among the target areas of each wind corridor forest. Despite the favorable conditions for a wind corridor forest, cold air flow showed that there are regions with unfavorable wind conditions in the HBC region. In order to strengthen the functions of a wind corridor forests in the HBC region, four zones were distinguished according to the functional characteristics. Additionally, the planning strategies of a wind corridor forests suitable for each zone were provided, and the strategies for establishing a wind corridor forest were proposed. The results of this study can be used as the fundamental data for establishing guidelines for a wind corridor forest and utilized as resources for selecting regions suitable for a wind corridor forest. Citation: Sung, U.-J.; Eum, J.-H.; Son, J.-M.; Oh, J.-H. Planning Strategies of Keywords: urban forest; ventilation corridor; forest planning; KALM; climate change Wind Corridor Forests Utilizing the Properties of Cold Air. Land 2021, 10, 607. https://doi.org/10.3390/ land10060607 1. Introduction The outbreak of urban heat islands (UHIs) due to urbanization and population growth Academic Editor: Sara A. Gagné is threatening the heat conditions in urban areas [1–3]. UHI represents the temperature in urban dense areas being higher than in suburban areas [4]. The worsening of the urban Received: 22 March 2021 thermal environment poses a biological threat to human health due to heat stress [5]. In Accepted: 4 June 2021 Published: 7 June 2021 South Korea, the number of heat-related illnesses caused by heatwaves reached the highest record of 4526 in 2018. Since the urban thermal environment continues to deteriorate, that

Publisher’s Note: MDPI stays neutral number is expected to increase further in the future. In addition to the problem of the with regard to jurisdictional claims in urban thermal environment, the increase in traffic and energy use in dense cities has led to published maps and institutional affil- an increase in fine dust emissions, resulting in atmospheric environmental problems [6]. iations. Fine dust is a harmful substance that causes lung and heart diseases when exposed for a long time, possibly leading to death [7,8]. In 2019, the annual average concentration of 3 PM2.5 in South Korea was 24.8 µg/m , the worst among OECD countries [9]. It has become an increasingly important task to improve these urban environmental problems and manage urban climate [10]. The Korea Forest Service aims to improve the Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. thermal and atmospheric environment in urban areas by creating urban forests utilizing This article is an open access article the functions of wind corridors (i.e., urban ventilation corridors or cold air paths). Wind distributed under the terms and corridors play a role in improving air pollution and the thermal environment as urban conditions of the Creative Commons ventilation corridors allow “cool and fresh air (cold air, commonly used in English in the Attribution (CC BY) license (https:// term ‘Kaltluft’ originated in German)” generated in forests at night to flow into urban creativecommons.org/licenses/by/ development areas where climate conditions are unfavorable [11]. Wind corridors operate 4.0/). by areas that create cold air, areas that cold air flows into, and the areas that connect the

Land 2021, 10, 607. https://doi.org/10.3390/land10060607 https://www.mdpi.com/journal/land Land 2021, 10, 607 2 of 17

two [12]. Hence, wind corridor forests are classified into three categories according to the role of urban forests in and outside the city that create the wind corridor. Wind-generating forests (WGF) are urban forests that create cold air, mainly located on the city outskirts. Wind-spreading forests (WSF) are greenery areas that create thermal winds caused by differences in temperature between built-up and greenery areas, such as urban parks. Wind-connecting forests (WCF) are linear greenery areas that connect WGFs and WSFs. The main examples of WCF include greenery areas along the river and trees along the street, which function as a corridor that transports cold air through the topographical structure of rivers and roads. Previous studies on wind corridors have focused on the relationship between urban structures and wind corridors and to improve climatic environment in terms of urban planning. Ren et al. [13] assessed wind corridor plans in order to improve the air exchange in different cities in China, investigated urban wind corridors, and applied them to urban master plans. Wong et al. [14] assessed city-level wind corridors using the concept of “building frontal area index” and identified major buildings that limit the flow of wind corridors through the analysis of the target area. Furthermore, they predicted wind corridors in the city and devised ways to identify the areas that benefit from wind corridors and the areas that supply cold air. Grunwald et al. [15] identified different types of wind corridors through the analysis of the data on the cooling rate using the cold-air analysis model of KLAM_21(Kaltluftabflussmodell_21), and analyzed wind corridors of the target area using the tool “Particle Track” (ArcGIS Software, ver.10.6). Furthermore, Grunwald et al. [16] devised a methodology to analyze wind corridors through regression analysis using the data from cold-air analysis of KLAM_21, land-use, topography, and buildings. Unlike previous studies on wind corridors that analyzed their relationship with urban structures, this study focused on the relationship between wind corridors, urban forests, and greenery areas. Studies on wind corridor forests started in 2019 in South Korea. Eum et al. [17] explored the characteristics of wind corridor forests and classified urban forests and greenery areas based on three types of WGF, WSF and WCF in Daegu Metropolitan City. Jang et al. [18] identified meteorological, topographical, and land-use conditions of cold air generation and its transport to establish principles of creating wind corridor forests. Son et al. [19] analyzed the current conditions of wind corridor forests in a medium-sized city, Pyeongtaek in South Korea, and proposed strategies for establishing wind corridors. Although previous studies have established the basis for creating wind corridor forests targeting several cities in South Korea, there were limitations in providing planning strategies of wind corridor forests. This is because the existing cities in need of wind corridor effects are already under development at high density, so it is insufficient to implement an efficient planning of wind corridor forests. In order to provide practical planning strategies for implementing wind corridor forests, Sejong Special Self-Governing City (Sejong) in South Korea was selected, where spaces have been secured to create urban forests and greenery. This study analyzes the current status of wind corridor forests and the cold air conditions. Based on them, it provides planning strategies for wind corridor forests according to the specific urban conditions of Sejong.

2. Materials and Methods 2.1. Study Area Sejong City was selected as the target area of this study. Sejong is located at 127◦300 longitude and 36◦300 latitude, bordered by Cheongju City to the east, Gongju City to the west, Cheonan City to the north, and other cities, such as Daejeon Metropolitan City, to the south (Figure1). The total geographic area is 465.2 km 2, and the administrative districts consist of 10 Dongs (neighborhood), 9 Myeons (townships), and 1 Eup (town). Sejong is a multifunctional administrative city with a new city center “Haengbok City (HBC region)” and a former city center, Jochiwon region. The new city center, the HBC region, was chosen as the target area for detailed analysis. The HBC region consists of 10 Dongs and 1 Myeon, Land 2021, 10, x FOR PEER REVIEW 3 of 18 Land 2021, 10, x FOR PEER REVIEW 3 of 18

2 thethe south south (Figure (Figure 1). 1). The The total total geographic geographic area area is is465.2 465.2 km km2, and, and the the administrative administrative dis- dis- trictstricts consist consist of of 10 10 Dongs Dongs (neighborhood), (neighborhood), 9 Myeons9 Myeons (townships), (townships), and and 1 Eup1 Eup (town). (town). Land 2021 10 , , 607 SejongSejong is is a amultifunctional multifunctional administrative administrative city city with with a newa new city city center center “Haengbok “Haengbok City3 ofCity 17 (HBC(HBC region)” region)” and and a aformer former city city center, center, Jochiwon Jochiwon region. region. The The new new city city center, center, the the HBC HBC region,region, was was chosen chosen as as the the target target area area for for deta detailediled analysis. analysis. The The HBC HBC region region consists consists of 10of 10 surrounded by green corridors along the mountainous areas at a height of 251 m above sea DongsDongs and and 1 1Myeon, Myeon, surrounded surrounded by by green green corridors corridors along along the the mountainous mountainous areas areas at aat a level, such as Mt. Wonsu, Mt. Janggun, and Mt. Bihak. Geumgan River, ﬂowing from east heightheight of of 251 251 m m above above sea sea level, level, such such as as Mt Mt. Wonsu,. Wonsu, Mt. Mt. Janggun, Janggun, and and Mt. Mt. Bihak. Bihak. Geumgan Geumgan to west, is located in the south of the HBC region (Figure1). River,River, flowing flowing from from east east to to west, west, is islocated located in inthe the south south of ofthe the HBC HBC region region (Figure (Figure 1). 1).

(a(a) ) (b(b) ) (c)( c) (d)( d)

FigureFigureFigure 1. 1. Maps MapsMaps of ofof study studystudy area: area: ( a( ()aa location)) location location of of ofSejong Sejong Sejong City; City; City; (b ()b (studyb) )study study area area area (HBC (HBC (HBC region); region); region); (c) (landc) ( cland) landcover cover cover of HBCof ofHBC HBCregion; region; region; (d) (d) (topographydtopography) topography of of HBC ofHBC HBC region. region. region.

FigureFigureFigure 2 2shows shows the the wind wind environment environment analyzed analyzed analyzed with with with the the the combined combined combined data data data on on onwind wind wind velocityvelocityvelocity and andand wind windwind direction direction direction over over over four four four years years years (2016 (2016 (2016 to to to2019) 2019) 2019) using using using the the the observational observational observational data data data fromfromfrom five ﬁve five automatic automatic automatic weather weather weather stations stations stations (AWS) (AWS) (AWS) in in inSejong. Sejong. Sejong. The The The prevailing prevailing prevailing wind wind wind direction direction direction was was was differentdifferentdifferent depending dependingdepending on onon the the the location location location of of of observ observation observationation stations. stations. stations. The TheThe important important important point point point is is thatis that that thethethe wind wind wind rose rose rose map map map showed showed showed high high high rates rates rates of of ofcalm calm calm winds winds winds (when (when (when the the the wind wind wind speed speed speed is is loweris lower lower than than than 0.30.30.3 m/s), m/s), m/s), ranging rangingranging from fromfrom 16.9% 16.9%16.9% to toto 34.1%. 34.1%.34.1%. Th TheThe averagee averageaverage wind windwind speed speedspeed measured measuredmeasured across across across the the the five ﬁve five stationsstationsstations was was was lower lower lower than than than 1.1 1.1 1.1 to to to 2 2 m/s.2 m/s. m/s. This ThisThis means meansmeans that thatthat Sejong SejongSejong is is anis an anarea areaarea with with with an an anunfavorable unfavorable unfavorable windwindwind environment environment environment in in in general, general, general, and and and it it isit is isrequir required requireded to to toimprove improve improve the the the wind wind wind conditions conditions conditions to to achieveto achieve achieve aa afavorable favorable favorable thermal thermal thermal and and and atmospheric atmospheric atmospheric environment. environment. environment.

Figure 2. Wind rose map in study area. Figure 2. Wind rose map in study area. Figure 2. Wind rose map in study area.

2.2. Methods In order to identify the target areas to create the wind corridor forests, the current conditions were analyzed for HBC region. The analysis consists of two parts. The ﬁrst part, analysis of wind corridor forests was conducted to understand the current status Land 2021, 10, x FOR PEER REVIEW 5 of 18

Land 2021, 10, 607 4 of 17 images were selected under the condition of less than 10% cloud volume during spring, summer, and autumn so that vegetation vitality is sufficiently displayed. As a result, the ofvegetation urban forests index and was the calculated greenery using areas the located Landsat within TM 8 the image city. on After 8 June the 2020, analysis, where urban the forestsvivid image are classified appeared. into Then, three the types vegetation according index to their within functions the analysis for generating, range was spreading catego- andrized connecting into three coldgrades air using as well Natural as its effects.Break methods As the second of ArcGIS part, ver. characteristics 10.6. of cold air flow representing important factors in wind corridor forest planning were analyzed to find 𝑁𝐷𝑉𝐼 = (1) places that need to be strengthened or supplemented with wind corridor. Based on the analysis of the first part, planning strategies are established for preserving the urban forests In the case of WCF, the analysis data on ventilation networks (Korea Research Insti- that perform the wind corridor and for selecting the areas that need to be complemented tute for Human Settlements, 2018) was used to identify whether WCF can be utilized ad- by wind corridor forests. The analysis was conducted throughout Sejong (analysis range: ditionally. Since the ventilation networks were calculated considering the nocturnal cold 45 km × 48 km, resolution: 30 m), and then a detailed analysis for HBC region (analysis air drainage based on topography and urban structures, areas on the ventilation networks range: 1.4 km × 1.2 km, resolution: 30 m) was carried out. The following presents the were considered as available areas for wind corridor forests. The ventilation network was details of the analysis methods used in this study. analyzed using the digital elevation model (DEM) provided by the National Geographic 2.2.1.Information Classification Institute of based Wind on Corridor ArcGIS Forest hydrological modeling. Based on the 1-km grid, the height of buildings was estimated, and the plan area index, façade area index, zero- In order to understand the distribution of urban forests and greenery areas, and to plane displacement, and roughness length were measured. The higher degree of ventila- classify them according to three types of wind corridor forest, the analysis schemes used tion network means a corridor with better flow, and two to six degrees out of a total eight by Eum et al. [17] were modified. Figure3 shows the modified schemes tailored to the degrees in the ventilation network was judged to be suitable for WCF of Sejong. Finally, properties of Sejong. the WCFs adjacent to the ventilation networks were selected as the main target sites.

Figure 3. Analysis schemes for wind corridor forests in this study. Figure 3. Analysis schemes for wind corridor forests in this study.

2.2.2.In Characteristics order to analyze of Cold the currentair Flow status of urban forests and greenery areas, the study mainlySimulations used the forestusing typenumerical map formodels 2015 can released be used by theto confirm Korea Forestthe flow Service, of cold and air sup- in plementedcomplicated the terrains urban green[21]. spaceRepresentative data from themodels cadastral for simulating map for 2018 cold by theair Ministryflow are of Land,KLAM_21 Infrastructure, (Kaltluftabflussmodell_21) and Transport, andand theKALM land (Kaltluftabflussmodell). cover map for 2018 by The the MinistryKLAM_21 of Environment,developed by asthe well German as satellite Weather images. Service The (D forestWD) type is a maptwo-dimensional, is a representative mathematical- map across thephysical country simulation and contains model various for calculating property co informationld air flow and on accumulation forest such as in physiognomy, topograph- species,ically structured age group, terrain and density.[22]. The Since KALM the mode forestl typecalculates map wasthe evolution released in of 2015, the cold the mapair wasflow complemented and the height withof cold urban air layer green occurr spaceed data on anda clear land and cover calm maps night while (i.e., comparingnocturnal themdrainage with flow) satellite [23]. images It is specialized to accurately in reﬂectanalyzing the characteristicsthe properties of of the cold rapidly air due developing to land studycover and area. topographical For example, characteristics data corresponding [24,25]. toTo forest determine areas the in the overall land wind cover corridors map and greenof Sejong spaces with such high as rates plazas, of calm parks, winds, green KALM areas, amusementwas selected parks, for this and study. public areas of the urban green space data were included in the forest type map to update the information on urban forests and greenery areas. Using the updated forest type map, the study classiﬁed urban forests and greenery areas into three types of wind corridor forest. First, since the wind-generating forest (WGF) Land 2021, 10, 607 5 of 17

refers to a forest located outside the city center, the urban forest is classified as a WGF, excluding the urban green space planning data that classified parks and green areas located in the city center. In contrast, the wind-spreading forest (WSF) and wind-connecting forest (WCF) were green spaces located in the city center. WSF corresponds to areal greenery spaces such as squares, parks, public areas and facility green areas, while WCF corresponds to linear green spaces such as riverside and roadside trees. In addition, three types of wind corridor forest were graded. WGF was classified into three grades according to the density levels of the forest type map. The higher grade has the better function for generating cold air, which means that it should be preserved instead of developed. WSF was graded using the vegetation index. In order to identify the vegetation index, the normalized difference vegetation index (NDVI), which is the most commonly used vegetation index, was analyzed. NDVI is calculated using Equation (1): NIR is near-infrared spectral reflectance, and RED is visible red spectral reflectance. Green plants hold positive NDVI values because the reflected rate of the NIR is smaller than the RED [20]. The NDVI is strongly correlated with the density and vitality of the vegetation. A value of NDVI closer to 1 means better vitality and density of vegetation. In this study, satellite images were selected under the condition of less than 10% cloud volume during spring, summer, and autumn so that vegetation vitality is sufficiently displayed. As a result, the vegetation index was calculated using the Landsat TM 8 image on 8 June 2020, where the vivid image appeared. Then, the vegetation index within the analysis range was categorized into three grades using Natural Break methods of ArcGIS ver. 10.6.

NIR − RED NDVI = (1) NIR + RED In the case of WCF, the analysis data on ventilation networks (Korea Research In- stitute for Human Settlements, 2018) was used to identify whether WCF can be utilized additionally. Since the ventilation networks were calculated considering the nocturnal cold air drainage based on topography and urban structures, areas on the ventilation networks were considered as available areas for wind corridor forests. The ventilation network was analyzed using the digital elevation model (DEM) provided by the National Geographic Information Institute based on ArcGIS hydrological modeling. Based on the 1-km grid, the height of buildings was estimated, and the plan area index, façade area index, zero-plane displacement, and roughness length were measured. The higher degree of ventilation network means a corridor with better ﬂow, and two to six degrees out of a total eight degrees in the ventilation network was judged to be suitable for WCF of Sejong. Finally, the WCFs adjacent to the ventilation networks were selected as the main target sites.

2.2.2. Characteristics of Cold air Flow Simulations using numerical models can be used to confirm the flow of cold air in complicated terrains [21]. Representative models for simulating cold air flow are KLAM_21 (Kaltluftabflussmodell_21) and KALM (Kaltluftabflussmodell). The KLAM_21 developed by the German Weather Service (DWD) is a two-dimensional, mathematical-physical simulation model for calculating cold air flow and accumulation in topographically structured terrain [22]. The KALM model calculates the evolution of the cold air flow and the height of cold air layer occurred on a clear and calm night (i.e., nocturnal drainage flow) [23]. It is specialized in analyzing the properties of cold air due to land cover and topographical characteristics [24,25]. To determine the overall wind corridors of Sejong with high rates of calm winds, KALM was selected for this study. The KALM model is based on the shallow water equations, the vertically integrated form of fluid mechanics [23]. The main equation is shown as follows (Equation (2)):

∂u ∂t + u·grad u = −ggradHe − cu|u| ∂h (2) ∂t + div(uh) = q Land 2021, 10, 607 6 of 17

where u = (u, v) is the horizontal velocity vector of cold air, h is the height of the cold air layer, H is the height of the top of cold air layer, c is a friction coefﬁcient, and q is the rate of generation of cold air. ge = g∆T/T0 is the amount of reduced gravity with T0 and ∆T representing the ambient temperature and the deﬁcit of a layer-averaged temperature. The layer-averaged horizontal velocity is expressed by u and v components. The KALM model provides value of the parameters for seven types of land use. Table1 shows the land cover types categorized for KALM simulation and their parameters used in this study.

Table 1. Land use types applied for the KALM model and their parameters used for this study.

Zero-Plane Temperature Cold Air Production Land Cover Type Roughness Length (m) Displacement Difference (k) Rate (m3/m2h) Height (m) Suburb 4.0 0.0 0.441 2.0 Forest 8.0 30.0 1.070 5.0 Open land 10.0 15.0 0.100 0.0 City center 2.0 0.0 0.714 5.0 Water 4.0 0.0 0.001 0.0 Trafﬁc areas, airports, etc. 4.0 0.0 0.010 1.0 Industrial areas 4.0 0.0 0.22 5.0

As major input data for analysis, the land cover map released in 2018 (the scale of 1:25,000) and the digital map by the Ministry of Environment were used. The 22 categories in the land cover map were reclassified into seven land use types (suburb, forest, open land, city center, water, traffic area, and industrial area) to apply to the KALM model. In addition, four parameters for each land use type were applied. For temperature difference and zero- plane displacement height, the values used in the KALM model were applied, and the values for surface roughness length and cold air production rate were taken from Eum [26] and Eum et al. [10], considering the structural properties of the cities in South Korea. The analysis of characteristics of cold air flow consists of an overall analysis for Sejong (analysis range: 45 km × 48 km, resolution: 30 m) and a detailed one for the HBC region (analysis range: 1.4 km × 1.2 km, resolution: 30 m). As the results, simulation data on cold air flow and the height of cold air layer over six hours (360 min) were extracted. Then, the cold air speed and the height in each administrative area (unit: Dong) were compared. Since this study did not include the validation of cold air flow by KALM simulation, the previous studies comparing observation values and model results by KALM were reviewed. Son et al. [2,27] analyzed the characteristics of cold air using KALM in mountainous regions of South Korea and compared the observation data from several meteorological stations with the simulation results. The studies confirmed that the air temperature decreased significantly at night in the region where the cold air flow was analyzed to be favorable, and that the KALM model was affordable for simulating cold air flow in mountainous regions such as Sejong and the HBC region.

3. Results 3.1. Wind Corridor Forests The distribution of wind-generating forests (WGFs) in Sejong is shown in Figure4. The total area of WGFs is 241 km2, accounting for 52% of Sejong City’s total area. With Mt. Bokdusan and Mt. Nogobong in the east, Mt. Janggunsan in the west, Mt. Yuegaesan and Mt. Bihaksan in the south, and Mt. Gorisan and Mt. Geumseongsan in the north, the overall distribution of WGFs in Sejong is revealed to be even. The area of WGF in HBC region is 54.82 km2, and when graded according to the density of the forest type map, the area ratio of Grade 1 was the highest at 68.82%, followed by Grade 2 (30.41%) and Grade 3 (0.77%). WGFs surround the outskirts of the HBC region, and those in the north play a major role as they are adjacent to residential and commercial areas. The major WGFs are Mt. Guksabong, Mt. Ohsan, and Mt. Wonsusan. However, Land 2021, 10, x FOR PEER REVIEW 7 of 18

3. Results 3.1. Wind Corridor Forests The distribution of wind-generating forests (WGFs) in Sejong is shown in Figure 4. The total area of WGFs is 241 km2, accounting for 52% of Sejong City’s total area. With Mt. Bokdusan and Mt. Nogobong in the east, Mt. Janggunsan in the west, Mt. Yuegaesan and Mt. Bihaksan in the south, and Mt. Gorisan and Mt. Geumseongsan in the north, the overall distribution of WGFs in Sejong is revealed to be even. The area of WGF in HBC region is 54.82 km2, and when graded according to the density of the forest type map, the area ratio of Grade 1 was the highest at 68.82%, followed Land 2021, 10, 607 7 of 17 by Grade 2 (30.41%) and Grade 3 (0.77%). WGFs surround the outskirts of the HBC region, and those in the north play a major role as they are adjacent to residential and commercial areas.the forests The major in the WGFs north are are Mt. fragmented Guksabong, due Mt. to urbanOhsan, development; and Mt. Wonsusan. thus, it However, is necessary the to forestsstrengthen in the the north function are fragmented to produce colddue airto throughurban development; forest management thus, it and is necessary restoration. to strengthen the function to produce cold air through forest management and restoration.

(a) (b)

FigureFigure 4. 4. DistributionDistribution of of wind-gen wind-generatingerating forest forest (WGF): (WGF): (a (a) )Sejong; Sejong; (b (b) )HBC HBC region. region.

Wind-spreadingWind-spreading forests forests (WSFs) (WSFs) in in Sejong Sejong were were mainly mainly parks parks in in the the urbanization urbanization area area andand green green spaces, spaces, mostly mostly concentrated concentrated in in the the center center of of the the HBC HBC region. region. As As the the HBC HBC region region isis a arecently recently built built planned planned city, city, WSFs WSFs were were evenly evenly distributed. distributed. WhenWhen the the vegetation vegetation index index was was analyzed analyzed using using satellite satellite images images (Figure (Figure 5),5), the the forest forest parksparks with with a a large area, suchsuch asas Duruteul Duruteul Neighborhood Neighborhood Park Park and and Gounteul Gounteul Neighborhood Neighbor- hoodPark, Park, had thehad highestthe highest vegetation vegetation index. index. However, However, small small parks parks located located in residential in residential and andcommercial commercial areas areas showed showed a lowa low vegetation vegetation index, index, because because treestrees inin the city city have have lower lower vitalityvitality due due to to the the unfavorable unfavorable growing growing envi environment.ronment. When When WSFs WSFs in the in HBC the HBC region region are classifiedare classiﬁed into three into threegrades grades according according to the tovegetation the vegetation index, index,the average the average value of value WSFs of wasWSFs 0.3, was equivalent 0.3, equivalent to grade to 2. grade Six of 2. the Six ma ofjor the parks major and parks urban and forest urban in forest the HBC in the region HBC wereregion categorized were categorized as grade as1, gradewhich 1,has which a high has vegetation a high vegetation index with index the value with theof 0.4 value or more,of 0.4 including or more, includingDuruteul Neighborhood Duruteul Neighborhood Park, Mt. Goihwasan, Park, Mt. Goihwasan, and Goeunteul and Neighbor- Goeunteul hoodNeighborhood Park (Table Park 2). (Table2). Wind-connecting forests (WCFs) of Sejong were well-established even in the city center, but there are some sections where WCFs are cut off. The major WCFs of the HBC region are Geumgang River, Jaechun Stream, Bangchukchun Stream, and roadside green areas with linear and wide forms (Figure6). LandLand 20212021, ,1010, ,x 607 FOR PEER REVIEW 8 8of of 18 17

(a) (b)

FigureFigure 5. Distribution 5. Distribution of wind-spr of wind-spreadingeading forest forest (WSF): (WSF): (a) (Sejong;a) Sejong; (b) (HBCb) HBC region region (expressed (expressed with with vegetation vegetation index). index).

Table 2. Grades and vegetation index of major wind-spreading forests. Table 2. Grades and vegetation index of major wind-spreading forests. Target Areas Grade Target Areas Grade Name Vegetation Index Name Vegetation Index Duruteul Neighborhood Park 0.49 DuruteulMt. Neighborhood Goihwasan Park 0.49 0.47 Mt. Goihwasan 0.47 Grade 1 (0.38–0.62) Gounteul Neighborhood Park 0.45 Grade 1 (0.38–0.62) Gounteul Neighborhood Park 0.45 Hansolteul Neighborhood Neighborhood Park Park 0.43 0.43 Byeolteul Neighborhood Neighborhood Park Park 0.42 0.42 LilyLily of the ValleyValley Neighborhood Neighborhood Park Park 0.42 0.42 Gadeukteul Neighborhood Neighborhood Park Park 0.370.37 GradeGrade 2 (0.19–0.37) 2 (0.19–0.37) Choryeo Choryeo Historical Historical Park Park 0.34 0.34 JaechunteulJaechunteul Neighborhood Neighborhood Park Park 0.230.23 GradeGrade 3 (0.00–0.18) 3 (0.00–0.18) Sejong Sejong Central Central Park Park 0.2 0.2

Wind-connectingThe ventilation network forests that(WCFs) is critical of Sejong in selecting were well-established areas for utilizing even WCF in the was city formed center,along butthe there river are such some as sections Geumgang where River, WCFs Jaechun are cut Stream, off. The and major Bangchukchun WCFs of the Stream. HBC regionAlthough are Geumgang there are WCFs River, located Jaechun on Stream the ventilation, Bangchukchun network, Stream, there areand also roadside some green WCFs areasthat arewith disconnected. linear and wide forms (Figure 6). TheAccording ventilation to the network analysis that of the is overallcritical conditions in selecting of windareas corridor for utilizing forests WCF(Figure was7 ), formedthe HBC along region the had river many such wind-generating, as Geumgang wind-spreading,River, Jaechun Stream, and wind-connecting and Bangchukchun forests Stream.and secured Although the connections there are WCFs among located the target on the areas ventilation of each ne windtwork, corridor there forest.are also Further- some WCFsmore, that ventilation are disconnected. networks were connected to the city center along the river. The vegetation indices were low in most of WSF and WCF; the average vegetation index of HBC region was 0.42, but the indices for WSF and WCF were 0.3 and 0.24, respectively, which were lower than the average. The index for WGF was 0.49 on average. WGF with a large proportion of 34.2% within the analysis range showed a high vegetation index, indicating its role in producing cold air in the HBC region. Although WSF and WCF occupy a small area in the HBC region, they are adjacent to residential and commercial areas. Hence, it Land 2021, 10, 607 9 of 17

Land 2021, 10, x FOR PEER REVIEW 9 of 18 is necessary to improve the vegetation index of these two types in order to increase the positive effect of cold air on the HBC region.

(a) (b) Figure 6. Distribution of wind-connecting forest (WCF) and ventilation network: (a) Sejong; (b) major areas available for Land 2021Figure, 10, 6.x FORDistribution PEER REVIEW of wind-connecting forest (WCF) and ventilation network: (a) Sejong; (b) major areas available for10 of 18 utilizing WCF in HBC region (expressed with satellite image). utilizing WCF in HBC region (expressed with satellite image). According to the analysis of the overall conditions of wind corridor forests (Figure 7), the HBC region had many wind-generating, wind-spreading, and wind-connecting forests and secured the connections among the target areas of each wind corridor forest. Furthermore, ventilation networks were connected to the city center along the river. The vegetation indices were low in most of WSF and WCF; the average vegetation index of HBC region was 0.42, but the indices for WSF and WCF were 0.3 and 0.24, respectively, which were lower than the average. The index for WGF was 0.49 on average. WGF with a large proportion of 34.2% within the analysis range showed a high vegetation index, indicating its role in producing cold air in the HBC region. Although WSF and WCF occupy a small area in the HBC region, they are adjacent to residential and commercial areas. Hence, it is necessary to improve the vegetation index of these two types in order to increase the positive effect of cold air on the HBC region.

(a) (b)

FigureFigure 7.7. TheThe overalloverall conditionsconditions ofof windwind corridorcorridor forestsforests inin HBCHBC region:region: ((aa)) windwind corridorcorridor forests;forests; ((bb)) vegetationvegetation indexindex ofof wind corridor forests. wind corridor forests. 3.2. Cold Air Flow The cold air produced at night was analyzed using the KALM model to understand the overall properties of the wind corridor in Sejong (Figures 8 and 9). The time series analysis of the average speed of cold air flow showed a high value (0.48 m/s) at the initial time (60 min) when cold air was produced, but the value decreased (0.44 m/s) until 180 min. The highest value of 0.49 m/s appeared from 240 to 360 min (Figure 10a). This indicates that the cold air flow was strong in the initial period when the cold air was created in WGFs, but then it gradually stabilized. The wind speed of cold air around Geumgang River became stronger over time. At 360 min after the simulation began, the highest value was 2.15 m/s, appearing at Sejong Lake Park. Regarding the height of the cold-air layer, it was confirmed that the cold-air layers were piling up through valleys in major areas of WGF, such as Mt. Guksabong, Mt. Ohsan, and Mt. Wonsusan, at the initial time of cold air production (60 min). Then, the cold-air layers increased over time and were widely distributed throughout most of HBC region. When comparing the height of cold-air layer by time, the highest and average heights 360 min after the simulation began were 182 m and 115.07 m, respectively. In particular, the generated cold air was piled up high at Sejong Lake Park located in the southeastern part of HBC region (Figure 10b).

(a) (b) (c) Land 2021, 10, x FOR PEER REVIEW 10 of 18

(a) (b) Figure 7. The overall conditions of wind corridor forests in HBC region: (a) wind corridor forests; (b) vegetation index of wind corridor forests.

3.2. Cold Air Flow The cold air produced at night was analyzed using the KALM model to understand the overall properties of the wind corridor in Sejong (Figures 8 and 9). The time series analysis of the average speed of cold air flow showed a high value (0.48 m/s) at the initial time (60 min) when cold air was produced, but the value decreased (0.44 m/s) until 180 Land 2021, 10, 607 min. The highest value of 0.49 m/s appeared from 240 to 360 min (Figure 10a). This10 indi- of 17 cates that the cold air flow was strong in the initial period when the cold air was created in WGFs, but then it gradually stabilized. The wind speed of cold air around Geumgang 3.2. Cold Air Flow River became stronger over time. At 360 min after the simulation began, the highest value was 2.15The coldm/s, appearing air produced at Sejong at night Lake was Park. analyzed using the KALM model to understand the overallRegarding properties the height of the of windthe cold-air corridor layer, in Sejong it was (Figuresconfirmed8 and that9). the The cold-air time series layers analysiswere piling of the up averagethrough speedvalleys of in cold major air areas ﬂow showedof WGF, a such high as value Mt. Guksabong, (0.48 m/s) at Mt. the Ohsan, initial timeand Mt. (60 min)Wonsusan, when cold at the air initial was produced, time of cold but theair valueproduction decreased (60 min). (0.44m/s) Then, until the 180cold-air min. Thelayers highest increased value over of 0.49 time m/s and appeared were widely from distributed 240 to 360 throughout min (Figure most 10a). of This HBC indicates region. that the cold air ﬂow was strong in the initial period when the cold air was created in When comparing the height of cold-air layer by time, the highest and average heights 360 WGFs, but then it gradually stabilized. The wind speed of cold air around Geumgang min after the simulation began were 182 m and 115.07 m, respectively. In particular, the River became stronger over time. At 360 min after the simulation began, the highest value generated cold air was piled up high at Sejong Lake Park located in the southeastern part was 2.15 m/s, appearing at Sejong Lake Park. of HBC region (Figure 10b).

Land 2021, 10, x FOR PEER REVIEW 11 of 18

(a) (b) (c)

(d) (e) (f)

FigureFigure 8. 8. TheThe speed speed of of cold cold air air flow ﬂow by by minute minute after after the the start start of of cold cold air air genera generationtion in in HBC HBC region: region: ( (aa)) 60 60 min; min; ( (bb)) 120 120 min; min; ((cc)) 180 180 min; min; ( (dd)) 240 240 min; min; ( (ee)) 300 300 min;( min;(ff)) 360 360 min. min. Regarding the height of the cold-air layer, it was conﬁrmed that the cold-air layers were piling up through valleys in major areas of WGF, such as Mt. Guksabong, Mt. Ohsan, and Mt. Wonsusan, at the initial time of cold air production (60 min). Then, the cold-air layers increased over time and were widely distributed throughout most of HBC region. When comparing the height of cold-air layer by time, the highest and average heights 360 min after the simulation began were 182 m and 115.07 m, respectively. In particular, the generated cold air was piled up high at Sejong Lake Park located in the southeastern part of HBC region (Figure 10b).

(a) (b) (c)

(e) (f) (g) Figure 9. The height of cold air layer by minute after the start of cold air generation in HBC region: (a) 60 min; (b) 120 min; (c) 180 min; (d) 240 min; (e) 300 min; (f) 360 min. Land 2021, 10, 607 11 of 17

Land 2021Figure, 10, 9.x FORThe PEER height REVIEW of cold air layer by minute after the start of cold air generation in HBC region: (a) 60 min; (b) 12012 min; of 18

(a) (b)

Figure 10. Comparison of of cold-air cold-air properties properties by by minute minute after after the the start start of of cold cold air air generation: generation: (a ()a changes) changes in incold-air cold-air wind wind speed by time; (b) changes in cold-air layer height by time. speed by time; (b) changes in cold-air layer height by time.

3.3.3.3. The The Comprehensive Comprehensive Analysis Analysis of of Wind Wind Corridor Corridor Forests Forests in in HBC HBC Region Region AccordingAccording to to the the analysis analysis of of wind wind corridor corridor forests forests and and cold cold air air flow, ﬂow, it itwas was judged judged thatthat the the HBC HBC region region has has abundant abundant wind wind corri corridordor forests and favorable conditions availa- available blefor for cold cold air air ﬂow. flow. However, However, there there were were differences differences in in the the current current status status of of each each region, region, so sodetailed detailed analysis analysis is requiredis required for for each each administrative administrative district district to suggestto suggest a strategy a strategy for windfor windcorridor corridor forest forest planning. planning. When comparing vegetation index by administrative districts (Figure 11), Yeongi- myeon (0.43) and Areum-dong (0.42) had high values, while Boram-dong (0.21) and Saerom-dong (0.18) had low values. Except for Yeongi-myeon, the indices in the other districts were lower than the average vegetation index of wind corridor forest. Thus, the planning of additional wind corridor forests is critical in order to increase the vegetation index across all regions. We also compared the vegetation index of three types of wind corridor forest according to the administrative district. In order to complement and strengthen the types with lower values, we analyzed the most critical type for each district. The districts where WGF is important are Yeongi-myeon, Goun-dong, and Areum-dong, and those where WSF in needed in advance are Yeongi-myeon, Dodam-dong, Jongchon- dong, and Boram-dong. WCF should be established first in the rest. When comparing the highest speed of cold air flow by administrative districts (Fig- ure 12), Eojin-dong and Yeongi-myeon had high values, whereas Jongchon-dong and Hansol-dong had low values. The average speed of cold air flow was high in Daepyeong- dong and Goun-dong, and low in Saerom-dong and Hansol-dong. Hansol-dong had the lowest value of both the highest and average speeds. The difference of the average values between Hansol-dong and Daepyeong-dong, which has the highest value of cold air flow, was 0.84 m/s. Land 2021, 10, 607 12 of 17

When comparing vegetation index by administrative districts (Figure 11), Yeongi- myeon (0.43) and Areum-dong (0.42) had high values, while Boram-dong (0.21) and Saerom- dong (0.18) had low values. Except for Yeongi-myeon, the indices in the other districts were lower than the average vegetation index of wind corridor forest. Thus, the planning of additional wind corridor forests is critical in order to increase the vegetation index across all regions. We also compared the vegetation index of three types of wind corridor forest according to the administrative district. In order to complement and strengthen the types with lower values, we analyzed the most critical type for each district. The districts where WGF is important are Yeongi-myeon, Goun-dong, and Areum-dong, and those where WSF Land 2021, 10, x FOR PEER REVIEW 13 of 18 in needed in advance are Yeongi-myeon, Dodam-dong, Jongchon-dong, and Boram-dong. WCF should be established ﬁrst in the rest.

Land 2021, 10, x FOR PEER REVIEW 13 of 18

(a) (b)

FigureFigure 11. 11.Comparison Comparison of of wind wind corridorcorridor forest properties by by administrative administrative districts: districts: (a ()a vegetation) vegetation index index of ofwind wind corridor corridor forests; (b) proportion of three types of wind corridor forest. forests; (b) proportion of three types of wind corridor forest.

When comparing the highest speed of cold air flow by administrative districts (Figure 12), Eojin-dong and Yeongi-myeon had high values, whereas Jongchon-dong and Hansol-dong had low values. The average speed of cold air ﬂow was high in Daepyeong-dong and Goun-dong, and low in Saerom-dong and Hansol-dong. Hansol-dong had the lowest value (ofa) both the highest and average speeds. The difference( ofb) the average values between Figure 11. Comparison of windHansol-dong corridor forest and properties Daepyeong-dong, by administrative which districts: has ( thea) vegetation highest valueindex of of wind cold corridor air ﬂow, was forests; (b) proportion of three0.84 types m/s. of wind corridor forest.

(a) (b) Figure 12. Comparison of speed of cold air flow by administrative districts at 360 min after the start of cold air generation: (a) highest value of cold air flow; (b) average value of cold air flow.

4. Planning Strategies for Wind Corridor Forest Based on the analysis of wind corridor forests and cold air flow in the HBC region, the following presents the strategies to plan for implementing wind corridor forests. Ac- cording to the consideration of the geographical location and each important type of wind (acorridor) forest in each administrative district, four zones are(b distinguished:) zone A, with Goun-dong, Areum-dong, and Jongchon-dong; zone B, with Dodam-dong and Eojin- FigureFigure 12. 12.Comparison Comparison of speeddong; of cold coldzone air air C, flow flowwith by by Hansol-dong administrative administrative anddi districtsstricts Saerom at at360-dong; 360 min min after and after the zone the startstart D, of withcold of cold airDaepyeong-dong, generation: air generation: (a) highest value of cold air flow; (b) average value of cold air flow. (a) highest value of cold air flow;Boram-dong, (b) average and value Sodam-dong of cold air flow.(Figure 13). In zone A, WGFs surround the outskirts, and large areas of WSFs are distributed adjacent to the residential areas. In addition, Goun- 4. Planning Strategies for Wind Corridor Forest dong and Areum-dong are districts with the highest average value of cold air flow. More- over,Based Jongchon-dong on the analysis was ofalso wind favorable. corridor Th forestse wind and corridor cold air forests flow in of the zone HBC A region,should thestrengthen following the presents cold-air the flow strategies by securing to plan the connectionfor implementing with WGFs wind surrounding corridor forests. them Ac-and cordingcomplement to the theconsideration existing WGFs of the and geographical WSFs. In zone location B, there and eachare large important WGFs, type such of aswind Mt. corridorOhsan and forest Mt. in Wonsubong, each administrative in the northeast district, fourof the zones residential are distinguished: area, and Bangchukchun zone A, with Goun-dong,Stream and majorAreum-dong, roads cross and the Jongchon-dong; residential areas. zone Furthermore, B, with Dodam-dong Eojin-dong is and the admin-Eojin- dong;istrative zone district C, with with Hansol-dong the highest and value Saerom of cold-dong; air flow and andzone also D, withranks Daepyeong-dong, on the top tier in Boram-dong,the comparison and of Sodam-dong average values. (Figure For the13). wind In zone corridor A, WGFs forests surround of zone the B, theoutskirts, function and to large areas of WSFs are distributed adjacent to the residential areas. In addition, Goun- dong and Areum-dong are districts with the highest average value of cold air flow. More- over, Jongchon-dong was also favorable. The wind corridor forests of zone A should strengthen the cold-air flow by securing the connection with WGFs surrounding them and complement the existing WGFs and WSFs. In zone B, there are large WGFs, such as Mt. Ohsan and Mt. Wonsubong, in the northeast of the residential area, and Bangchukchun Stream and major roads cross the residential areas. Furthermore, Eojin-dong is the administrative district with the highest value of cold air flow and also ranks on the top tier in the comparison of average values. For the wind corridor forests of zone B, the function to Land 2021, 10, 607 13 of 17

4. Planning Strategies for Wind Corridor Forest Based on the analysis of wind corridor forests and cold air flow in the HBC region, the following presents the strategies to plan for implementing wind corridor forests. According to the consideration of the geographical location and each important type of wind corridor forest in each administrative district, four zones are distinguished: zone A, with Goun- dong, Areum-dong, and Jongchon-dong; zone B, with Dodam-dong and Eojin-dong; zone C, with Hansol-dong and Saerom-dong; and zone D, with Daepyeong-dong, Boram-dong, and Sodam-dong (Figure 13). In zone A, WGFs surround the outskirts, and large areas of WSFs are distributed adjacent to the residential areas. In addition, Goun-dong and Areum- dong are districts with the highest average value of cold air flow. Moreover, Jongchon-dong was also favorable. The wind corridor forests of zone A should strengthen the cold-air flow by securing the connection with WGFs surrounding them and complement the existing

Land 2021, 10, x FOR PEER REVIEWWGFs and WSFs. In zone B, there are large WGFs, such as Mt. Ohsan and Mt. Wonsubong,14 of 18 in the northeast of the residential area, and Bangchukchun Stream and major roads cross the residential areas. Furthermore, Eojin-dong is the administrative district with the highest value of cold air flow and also ranks on the top tier in the comparison of average values. For thegenerate wind corridorcold air forestsshould of be zone strengthened B, the function by complementing to generate cold large air should WGFs be and strengthened the WCFs byof complementingrivers and roads large should WGFs be andenhanced the WCFs in consideration of rivers and of roads the shouldcold-air be flow enhanced from the in considerationnorth. Zone C ofis thethe cold-airzone where flow the from effect the of north. WGFs Zone is weak, C is thebut zoneWSFs where and WCFs the effect play ofan WGFsimportant is weak, role. but Moreover, WSFs and the WCFs districts play in an zone important C have role. a relatively Moreover, low the speed districts of incold zone air Cflow, have especially a relatively in lowHansol-dong, speed of cold where air flow,the va especiallylue is considerably in Hansol-dong, low compared where the to value other isadministrative considerably lowdistricts. compared The wind to other corridor administrative forests of zone districts. C require The wind planning corridor of WSFs forests in ofas zonemany C regions require planningas possible of as WSFs well in as as WCFs many using regions rivers as possible and major as well roads as WCFsconnected using to riverszone A. and Lastly, major zone roads D connected has WGFs to located zone A. in Lastly, the south zone of D the has residential WGFs located area in and the a south large ofarea the of residential WSFs such area as Mt. and Goihwasan. a large area In of WSFsaddition, such the as Geumgang Mt. Goihwasan. River, Inwhich addition, serves the as Geumganga wind corridor, River, whichis adjacent, serves so as the a wind highest corridor, speed is of adjacent, cold air so flow the is highest considered speed to of coldhave airbeen flow high is consideredin Daepyeong-dong. to have been The high average in Daepyeong-dong. value of zone D Theis placed average in the value mid-range. of zone DThe is placedwind corridor in the mid-range. forests of zone The windD should corridor strengthen forests ofthe zone cold-air D should flow by strengthen securing thethe cold-airconnection flow with by securingWGFs and the implement connection WCFs with to WGFs allow and cold implement air that flows WCFs along to allowGeumgang cold airRiver that into flows the along city (Figure Geumgang 14). River into the city (Figure 14).

FigureFigure 13.13. DivisionDivision ofof zoneszones forfor windwind corridorcorridorforest forest implantation implantation in in HBC HBC region region based based on on adminis- admin- trativeistrative districts. districts.

(a) (b) Land 2021, 10, x FOR PEER REVIEW 14 of 18

generate cold air should be strengthened by complementing large WGFs and the WCFs of rivers and roads should be enhanced in consideration of the cold-air flow from the north. Zone C is the zone where the effect of WGFs is weak, but WSFs and WCFs play an important role. Moreover, the districts in zone C have a relatively low speed of cold air flow, especially in Hansol-dong, where the value is considerably low compared to other administrative districts. The wind corridor forests of zone C require planning of WSFs in as many regions as possible as well as WCFs using rivers and major roads connected to zone A. Lastly, zone D has WGFs located in the south of the residential area and a large area of WSFs such as Mt. Goihwasan. In addition, the Geumgang River, which serves as a wind corridor, is adjacent, so the highest speed of cold air flow is considered to have been high in Daepyeong-dong. The average value of zone D is placed in the mid-range. The wind corridor forests of zone D should strengthen the cold-air flow by securing the connection with WGFs and implement WCFs to allow cold air that flows along Geumgang River into the city (Figure 14).

Land 2021, 10, 607 14 of 17 Figure 13. Division of zones for wind corridor forest implantation in HBC region based on administrative districts.

Land 2021, 10, x FOR PEER REVIEW 15 of 18 (a) (b)

FigureFigure 14. 14. StrategiesStrategies for for wind wind corridor corridor forest forest planning planning by byzone: zone: (a) (zonea) zone A; ( A;b) zone (b) zone B; (c B;) zone (c) zone C; C; ((dd)) zone zone D. D.

Finally,Finally, the the strategies strategies for for planning planning by by three three types types of of wind wind corridor corridor forests forests are are sug- sug- gested.gested. WGSs WGSs should should strengthen strengthen the thefunction function of wind of wind production production through through changes changes in for- in estforest structure structure and andvegetation vegetation in existing in existing moun mountainoustainous neighborhood neighborhood parks. parks. The vegetation The vegeta- indextion index may be may increased be increased through through dense dense planting, planting, while whiledecreasing decreasing the density the density of the ofpe- the ripheralperipheral areas areas of WGF of WGF so that so thatthe generated the generated wind wind can circulate can circulate well. well.To create To create forests, forests, the speciesthe species suitable suitable for climate for climate and andsoil soilproperties properties of the of thetarget target areas areas should should be beidentified. identified. DamagedDamaged forests forests should should be be managed managed and and restored restored to to increase increase their their vegetation vegetation index. index. Fur- Fur- thermore,thermore, WSFs WSFs should should be be planned planned in in vulnerable vulnerable districts districts of of cold cold air air flow flow by by creating creating new new urbanurban parks parks as as well as improving existing ones. By complementing planting inin thethe existingexist- ingurban urban parks, parks, the the production production functions functions of of cold cold airair shouldshould be enhanced. enhanced. In In addition, addition, the the improvementimprovement and and expansion expansion of of green green spaces spaces in inthe the existing existing public public institutes institutes and and schools schools shouldshould be be implemented. implemented. Through Through various various green green projects, projects, such such as as green green parking parking lots lots and and demolitiondemolition of of walls, walls, the the small small area area with with the the function function of of WSFs WSFs should should be be secured secured in ineach each administrativeadministrative district. district. To To allow allow WSFs WSFs to to expand expand the the flow flow of of cold cold air air to to the the surrounding surrounding areas,areas, low-rise low-rise plants plants should should be be planted planted along along the the boundaries boundaries of of WSFs. WSFs. The The land land cover cover of of naturalnatural pastures, pastures, such such as as grass, grass, should should be be maintained maintained to to secure secure cold cold air air production production func- func- tions,tions, and and impermeable impermeable paving should be removed.removed. WCFs shouldshould enhanceenhance thethe functionfunction of of wind flow connection using roadside trees and riverside green spaces. In the case of roads, the forest planning may be implemented according to guidelines of street tree planting in consideration of the surrounding land use and building structures as well as the conditions of roads such as widths of driveways and sidewalks. Old street trees with a low vegetation index should be replaced, and tree shading should be strengthened by adding a row of new street trees to prevent a rapid increase of surface temperature on the road during the day. In the case of rivers, the ground surface may be covered with grasslands to allow maintaining the function of generating cold air better.

5. Conclusions This study aims to provide practical planning strategies for implementing wind corridor forests utilizing the properties of cold air with a study area of Sejong recently founded in South Korea. To achieve this goal, the study investigated the current conditions of wind corridor forests and cold air flow across regions in Sejong and carried out a detailed analysis in the HBC region. After dividing the HBC region into four zones based on the analysis results, the study provided strategies for different plans in each zone and set out planning strategies by three types of wind corridor forests. The HBC region was surrounded by the wind-generating forests (WGF) on the outskirts of the city, and there were major WGFs, such as Mt. Guksabong, Mt. Ohsan, and Land 2021, 10, 607 15 of 17

wind ﬂow connection using roadside trees and riverside green spaces. In the case of roads, the forest planning may be implemented according to guidelines of street tree planting in consideration of the surrounding land use and building structures as well as the conditions of roads such as widths of driveways and sidewalks. Old street trees with a low vegetation index should be replaced, and tree shading should be strengthened by adding a row of new street trees to prevent a rapid increase of surface temperature on the road during the day. In the case of rivers, the ground surface may be covered with grasslands to allow maintaining the function of generating cold air better.

5. Conclusions This study aims to provide practical planning strategies for implementing wind corridor forests utilizing the properties of cold air with a study area of Sejong recently founded in South Korea. To achieve this goal, the study investigated the current conditions of wind corridor forests and cold air flow across regions in Sejong and carried out a detailed analysis in the HBC region. After dividing the HBC region into four zones based on the analysis results, the study provided strategies for different plans in each zone and set out planning strategies by three types of wind corridor forests. The HBC region was surrounded by the wind-generating forests (WGF) on the outskirts of the city, and there were major WGFs, such as Mt. Guksabong, Mt. Ohsan, and Mt. Wonsusan, in the north. The WGFs were graded according to the forest density; grade 1, which is classified as forests advantageous for generating cold air, occupies the largest proportion at 68.82%, followed by grade 2 (30.41%) and grade 3 (0.77%). Wind-spreading forests (WSF) were evenly distributed in the form of neighborhood parks. WSFs were graded according to the vegetation index; the mountainous parks with large area were classified as grade 1, and small parks were generally classified as grade 2. The wind-connecting forests (WCF) were well-established in the HBC region, but some areas were disconnected. The ventilation network analysis was carried out to complement the disconnected WCFs. As a result, the study identified available areas where WCFs can be utilized around river areas, such as Geumgang River, Jaechun Stream, and Bangchukchun Stream. The study also identified unfavorable types of wind corridor forests that should be implemented for each administrative district. The analysis of cold air flow using the cold air analysis model, KALM, was carried out by focusing on the speed of cold air flow and height of cold air layer. Each analysis confirmed the time-series change of value. The height of the cold air layer gradually increased over time, but the speed of the cold air flow increased and decreased repeatedly. In addition, the highest and average speed of cold air flow were compared by administrative district. Eojin-dong and Daepyeong-dong showed high values, while Hansol-dong was the administrative district with low values. According to the functional characteristics of the HBC region, four zones were distinguished, and the planning strategies suitable for wind corridor forests in each zone were provided. Zone A should strengthen the connection between WGFs surrounding the outskirts and WSFs in the city center. Zone B should complement large WGFs and enhance WCFs of major roads and Bangchukchun Stream. Zone C has disadvantageous wind, and hence, it should secure WSFs and implement WGFs connected to zone A. Lastly, zone D should secure a connection between WGFs located in the south and the Geumgang River located in the north and thus enhance the cold-air flow. Moreover, the planning strategies for three types of wind corridor forest were also presented. The strategies strengthen each function of wind corridor forests—the wind production of WGF, wind diffusion and generation of WSF, and wind flow connection of WCF. WGFs should be changed in forests’ structures and vegetation. WSFs should be planned by creating new urban parks and complement existing urban parks. WCFs should be implemented by using roadside trees and riverside green spaces. The results of this study can be used as the fundamental data for implementing wind corridor forest projects and utilized as resources for selecting regions to establish wind Land 2021, 10, 607 16 of 17

corridor forests. Moreover, it is expected that the suggestions for planning strategies provided in this study contribute to providing directions for practical planning for wind corridor forests. This study offered planning strategies for wind corridor forests based on the analysis of wind corridor forests and cold air flow in a target area. However, as the application strategies may differ depending on the target city, it is suggested that the analysis of the properties of the city should be carried out when the strategies are to be made for another region. It is expected that the accumulation of various case studies will lead to creating a guideline for wind corridor forest planning that can be commonly applied for any region. In addition, since the analysis of wind corridor forest in the study is based on the results of simulation, the verification of planning effects is lacking. Thus, future studies should verify the reliability of wind corridor forest analysis using various methods, such as satellite images and field studies. In particular, since the KALM model used for this study analyzed cold air flow occurred from topographical properties and land use conditions on a calm night, it may be different from the actual wind environment of Sejong. In a trailing study, a cold air simulation model that can reflect the synoptic wind conditions will be used to analyze various wind flows for the study area.

Author Contributions: Conceptualization, U.-J.S. and J.-M.S.; methodology, J.-H.E.; investigation, J.-M.S.; resources, J.-H.O.; writing—original draft preparation, U.-J.S.; writing—review and editing, U.-J.S.; visualization, U.-J.S.; supervision, J.-H.E.; project administration, J.-H.E. All authors have read and agreed to the published version of the manuscript. Funding: This research was supported by Kyungpook National University Research Fund, 2018. Data Availability Statement: Data sharing not applicable. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Eum, J.H.; Scherer, D.; Fehrenbach, U.; Köppel, J.; Wang, J. Integrating urban climate into urban master plans using spatially distributed information—The Seoul example. Land Use Policy 2013, 34, 223–232. [CrossRef] 2. Son, J.M.; Eum, J.H.; Kim, D.P.; Kwon, J. Management strategies of thermal environment in urban area using the cooling function of the mountains: A case study of the Honam Jeongmaek areas in South Korea. Sustainability 2018, 10, 4691. [CrossRef] 3. Eum, J.H.; Kim, K.; Jung, E.H.; Rho, P. Evaluation and utilization of thermal environment associated with policy: A case study of Daegu Metropolitan City in South Korea. Sustainability 2018, 10, 1179. [CrossRef] 4. Kolokotroni, M.; Giridharan, R. Urban heat island intensity in London: An investigation of the impact of physical characteristics on changes in outdoor air temperature during summer. Sol. Energy 2008, 82, 986–998. [CrossRef] 5. Zhao, L.; Lee, X.; Smith, R.B.; Oleson, K. Strong contributions of local background climate to urban heat islands. Nature 2014, 511, 216–219. [CrossRef][PubMed] 6. Koh, B.; Kim, E.A. Comparative analysis of urban road dust compositions in relation to their potential human health impacts. Environ. Pollut. 2019, 255, 113156. [CrossRef] 7. Kwon, M.Y.; Lee, J.S.; Park, S. The effect of outdoor air pollutants and greenness on allergic rhinitis incidence rates: A cross- sectional study in Seoul, Korea. Int. J. Sustain. Dev. World Ecol. 2019, 26, 258–267. [CrossRef] 8. Kim, H.; Hong, S. Differences in the Inﬂuence of Microclimate on Pedestrian Volume According to Land-Use. Land 2021, 10, 37. [CrossRef] 9. IQAir. 2019 World Air Quality Report Region & City Pm2.5 Ranking. Available online: https://www.greenpeace.org/static/ planet4-thailand-stateless/2020/02/91ab34b8-2019-world-air-report.pdf (accessed on 10 March 2021). 10. Eum, J.H.; Scherer, D.; Fehrenbach, U.; Woo, J.H. Development of an urban landcover classiﬁcation scheme suitable for representing climatic conditions in a densely built-up Asian megacity. Landsc. Urban Plan. 2011, 103, 362–371. [CrossRef] 11. Verein Deutscher Ingenieure. Environmental meteorology-local cold air. In VDI Guideline 3787, Part 5; Beuth: Berlin, Germany, 2003; Volume 85, pp. 14–29. 12. Sievers, U.; Kossmann, M. The cold air drainage model KLAM_21-Model formulation and comparison with observations. Weather Clim. 2016, 36, 2–24. [CrossRef] 13. Ren, C.; Yang, R.; Cheng, C.; Xing, P.; Fang, X.; Zhang, S.; Wang, H.; Shi, Y.; Zhang, X.; Kwok, Y.T.; et al. Creating breathing cities by adopting urban ventilation assessment and wind corridor plan–the implementation in Chinese cities. J. Wind. Eng. Ind. Aerodyn. 2018, 182, 170–188. [CrossRef] 14. Wong, M.S.; Nichol, J.E.; To, P.H.; Wang, J. A simple method for designation of urban ventilation corridors and its application to urban heat island analysis. Build. Environ. 2010, 45, 1880–1889. [CrossRef] Land 2021, 10, 607 17 of 17

15. Grunwald, L.; Kossmann, M.; Weber, S. Mapping urban cold-air paths in a Central European city using numerical modelling and geospatial analysis. Urban Clim. 2019, 29, 100503. [CrossRef] 16. Grunwald, L.; Schneider, A.K.; Schroeder, B.; Weber, S. Predicting urban cold-air paths using boosted regression trees. Landsc. Urban Plan. 2020, 201, 103843. [CrossRef] 17. Eum, J.H.; Oh, J.H.; Son, J.M.; Kim, K.; Baek, J.B.; Lee, C.Y. Analysis Schemes of Wind Ventilation Forest Types-A Case Study of Daegu Metropolitan City. J. Korean Assoc. Geogr. Inf. Stud. 2019, 22, 12–23. (In Korean) 18. Jang, G.S.; Eum, J.H.; Son, J.M.; Baek, J.B.; Sung, W.J.; Kim, J.E.; Kim, S.H. Application of Wind Ventilation Corridor Model in Urban Forest Planning and Management; National Institute of Forest Science: Daejeon, Korea, 2019; pp. 1–82. (In Korean) 19. Son, J.M.; Eum, J.H.; Sung, U.J.; Baek, J.B.; Kim, J.E.; Oh, J.H. A Proposal of Direction of Wind Ventilation Forest through Urban Condition Analysis-A Case Study of Pyeongtaek-si. J. Korean Assoc. Geogr. Inf. Stud. 2020, 23, 101–119. (In Korean) 20. Goldblatt, R.; Adda, A.; Crull, D.; Maghrabi, A.; Levin, G.G.; Rubinyi, S. Remotely sensed derived land surface temperature (LST) as a proxy for air temperature and thermal comfort at a small geographical scale. Land 2021, 10, 410. [CrossRef] 21. Sachsen, T.G.; Schneider, C. The effect of vegetation on suburban ventilation paths: Studies on cold air drainage in the city of Aachen. RWTH Aachen Univ. 2014, 229, 229472. (In German) 22. Sievers, U. Das Kaltluftabﬂussmodell KLAM_21: Theoretisch Grundlagen, Anwendung und Handhabung des PC-Modells; Deutscher Wetterdienst: Offenbach, Germany, 2005. 23. Schädler, G.; Lohmeyer, A. Simulation of nocturnal drainage ﬂows on personal computers. Meteorol. Zeitshrift 1994, 3, 167–171. [CrossRef] 24. Eum, J.H.; Son, J.M. Management of the Nakdong-Jeongmaek based on the characteristics of Cold Air–Focused on Busan, Ulsan, Pohang. J. Korean Inst. Landsc. Archit. 2016, 44, 103–115. (In Korean) [CrossRef] 25. Eum, J.H. Management Strategies of Local Cold Air in Jeongmaek for utilizing urban Ventilation Corridor—A Case Study of the Nak-nam Jeongmaek. J. Korean Assoc. Geogr. Inf. Stud. 2019, 22, 154–167. (In Korean) 26. Eum, J.H. Integration of Climate Information into Spatial Planning in Seoul, South Korea. Ph.D. Thesis, Technische Universität Berlin, Berlin, Germany, 14 April 2008. 27. Son, J.M.; Eum, J.H. Analysis and management strategies of the cold air characteristics in Hannnamgeumbuk-jeongmaek and Geumbuk-jeongmaek. J. Korean Assoc. Geogr. Inf. Stud. 2019, 22, 152–171. (In Korean)