• 中国中文核心期刊
  • 中国科学引文数据库(CSCD)核心库来源期刊
  • 中国科技论文统计源期刊(CJCR)
  • 第二届国家期刊奖提名奖
Volume 36 Issue 2
Apr.  2023
Article Contents
Turn off MathJax

Citation:

Variation of Negative Air Ions and Its Influencing Factors in Typical Plantations in Rocky Mountain Area of North China

  • Corresponding author: SANG Yu -qiang, syuqiang@163.com
  • Received Date: 2022-11-24
    Accepted Date: 2023-01-11
  • Objective To explore the variation characteristics of Negative air ion (NAI) of typical plantations and the relationship between NAI and environmental factors in growing period and non-growing period in rocky mountain area of north China, and reveal environmental factors affecting NAI in different seasons in this area. Method The variation characteristics of NAI of typical plantations in Henan Xiaolangdi Earth Critical Zone National Research Station were measured from May to December in 2021 using the air negative ions, PM 2.5, PM 10 and meteorological data. Besides, the main environmental factors and variable importance measures affecting NAI of typical plantations in the area were analyzed by random forest algorithm. Results The diurnal variation of NAI in Quercus variabilis showed single peak curve in growing period but not obvious in non-growing period. The NAI of Platycladus orientalis showed single peak cure during the experiment period. The NAI concentration of Quercus variabilis plantations (740.32 ion·cm−3) was higher than that of Platycladus orientalis (703.74 ion·cm−3) during the observation period. The daily NAI of Quercus variabilis (858.94 ion·cm−3) was higher than that of Platycladus orientalis (724.33 ion·cm−3) during the growing period. The daily NAI of Quercus variabilis (621.70 ion·cm−3) was lower than that of Platycladus orientalis (683.16 ion·cm−3) during the non-growing period. The meteorological factors such as Air temperature (Ta), Relative humidity (RH), Vapor pressure deficit (VPD) and photosynthetically active radiation (PAR) in the growing period were higher than those in the non-growing period, while the particulate matter (PM2.5 and PM10) in the non-growing period was higher than that in the growing period. The concentration of PM10 was higher than that of PM2.5 during the experiment period. There was no significant difference between Wind speed (WS) between the growing period and non-growing period. The random forest method revealed that the main environmental factors affecting the NAI concentrations of Quercus variabilis and Platycladus orientalis in the growing period were VPD, PAR and WS, and their variable importance measures were 20.22, 15.08 ,14.71, respectively, and 25.08, 16.76, 16.49, respectively. The main environmental factors affecting the NAI concentration of Quercus variabilis and Platycladus orientalis during the non-growth period were PM 2.5, WS and PM 10, and their variable importance measures were 33.36, 17.58, 14.28, respectively, and 15.89, 17.51, 14.62, respectively. Conclusion The diurnal variation of NAI concentration of Quercus variabilis and Platycladus orientalis in growing period both showed a single peak curve; the diurnal variation of NAI concentration of Quercus variabilis plantations was not obvious, while the diurnal variation NAI concentration of Platycladus orientalis plantations showed a single peak curve in non-growing period. There were significant differences in NAI concentration between Quercus variabilis and Platycladus orientalis plantations, NAI concentration of Quercus variabilis was higher than that of Platycladus orientalis during the growing period, while NAI concentration of Quercus variabilis was lower than that of Platycladus orientalis during the non-growing period. The NAI concentration of Quercus variabilis was higher than that of Platycladus orientalis during the observation period. Differences of environmental factors affecting NAI of typical plantations in the area were obvious. VPD and PAR were the key factors during the growing period, while PM2.5, PM 10 and WS were the key factors during the non-growing period.
  • 加载中
  • [1]

    SHI G Y, ZHOU Y, SANG Y Q, et al. Modeling the response of negative air ions to environmental factors using multiple linear regression and random forest[J]. Ecological Informatics, 2021, 66: 101464. doi: 10.1016/j.ecoinf.2021.101464
    [2]

    WANG R, CHEN Q, WANG D X. Effects of altitude, plant communities, and canopies on the thermal comfort, negative air ions, and airborne particles of mountain forests in summer[J]. Sustainability, 2022, 14(7): 3882. doi: 10.3390/su14073882
    [3] 李爱博, 赵雄伟, 李春友, 等. 基于控制试验的植株数量及空气温湿度与负离子的关系[J]. 应用生态学报, 2019, 30(7):2211-2217.

    [4] 施光耀, 周 宇, 桑玉强, 等. 基于随机森林方法分析环境因子对空气负离子的影响[J]. 中国农业气象, 2021, 42(5):390-401. doi: 10.3969/j.issn.1000-6362.2021.05.004

    [5]

    GOLDSTEIN N. Reactive oxygen species as essential components of ambient air[J]. Biochemistry (Moscow), 2002, 67(2): 161-170. doi: 10.1023/A:1014405828966
    [6]

    LIN W, ZENG C C, NIE W B, et al. Study of the vertical structures, thermal comfort, negative air ions, and human physiological stress of forest walking spaces in summer[J]. Forests, 2022, 13(2): 335. doi: 10.3390/f13020335
    [7]

    LIU S, LI C, CHU M T, et al. Associations of forest negative air ions exposure with cardiac autonomic nervous function and the related metabolic linkages: A repeated-measure panel study[J]. Science of the Total Environment, 2022, 850: 158019. doi: 10.1016/j.scitotenv.2022.158019
    [8]

    HAN H S, JEONG G J, LEE H W, et al. Innovative use of negative air ions as an alternative therapy for acne vulgaris: a report of three cases[J]. Annals of Dermatology, 2022, 34(3): 216-220. doi: 10.5021/ad.2022.34.3.216
    [9]

    LI A B, LI Q L, ZHOU B Z, et al. Temporal dynamics of negative air ion concentration and its relationship with environmental factors: results from long-term on-site monitoring[J]. The Science of the Total Environment, 2022, 832: 155057. doi: 10.1016/j.scitotenv.2022.155057
    [10] 施光耀, 桑玉强, 张劲松, 等. 不同光照强度下植物电信号变化特征及其与空气负离子的关系[J]. 应用生态学报, 2022, 33(2):439-447.

    [11]

    WANG H, WANG B, NIU X, et al. Study on the change of negative air ion concentration and its influencing factors at different spatio-temporal scales[J]. Global Ecology and Conservation, 2020, 23: e01008. doi: 10.1016/j.gecco.2020.e01008
    [12]

    ZHANG J, YU Z L. Experimental and simulative analysis of relationship between ultraviolet irradiations and concentration of negative air ions in small chambers[J]. Journal of Aerosol Science, 2006, 37(10): 1347-1355. doi: 10.1016/j.jaerosci.2006.03.003
    [13] 余 娟, 高占冬, 王德远, 等. 天缘洞空气负离子时空分布特征及影响因素分析[J]. 环境化学, 2021, 40(4):1078-1087.

    [14] 冯燕珠. 公园不同植物配置群落空气负离子变化特征研究[D]. 福州: 福建农林大学, 2018.

    [15]

    DENG L. Review on research of the negative air ion concentration distribution and its correlation with meteorological elements in mountain tourist area[J]. Earth Sciences, 2019, 8(1): 60-68. doi: 10.11648/j.earth.20190801.15
    [16] 包红光, 闫晓云, 侯秀娟, 等. 半干旱城市公园绿地PM2.5与空气负离子浓度动态特征[J]. 生态学杂志2023,42(1) :170-179.

    [17] 侯秀娟, 闫晓云, 王 波, 等. 夏季干旱半干旱城市公园绿地空气负离子与空气颗粒物变化特征[J]. 南京林业大学学报(自然科学版), 2022, 46(4):212-220.

    [18] 赵 号, 郭继峰, 张龙镇, 等. 嵌入式自然空气负离子与PM2.5作用特性实验装置[J]. 洛阳理工学院学报(自然科学版), 2021, 31(2):35-41,48.

    [19] 韦赛君, 张 静, 王 翔, 等. 金钱松林挥发物季节性变化对空气负离子及微生物的影响[J]. 江西农业大学学报, 2021, 43(6):1316-1326.

    [20] 黄向华, 曾宏达, 陈 惠, 等. 基于静态箱法研究城市草坪空气负离子来源及其影响因素[J]. 应用与环境生物学报, 2021, 27(4):949-955.

    [21] 王梦楠. 上杭城区空气负离子浓度时空变化及模拟[D]. 长沙: 中南林业科技大学, 2021.

    [22] 蓝素素. 森林康养视角下的城市型森林公园规划设计[D]. 北京: 北京林业大学, 2021.

    [23] 王一荃, 周 璋, 李意德, 等. 不同热带森林空气负离子浓度评价研究[J]. 生态环境学报, 2021, 30(5):898-906.

    [24] 廖荣俊, 颜晓捷, 江 波, 等. 灵鹫山国家森林康养基地空气负氧离子浓度变化特征及其影响因素研究[J]. 浙江林业科技, 2021, 41(5):36-41.

    [25] 杨 畅, 王月容, 汤志颖, 等. 不同群落结构风景游憩林生态保健效应研究——以北京西山国家森林公园为例[J]. 生态学报, 2022, 42(16):1-15.

    [26] 施光耀, 桑玉强, 张劲松, 等. 自然状态下栓皮栎人工林空气负离子浓度与相对湿度的关系[J]. 中国农业气象, 2021, 42(1):24-33.

    [27]

    ZHANG C Y, WU Z N, LI Z H, et al. Inhibition effect of negative air ions on adsorption between volatile organic compounds and environmental particulate matter[J]. Langmuir, 2020, 36(18): 5078-5083. doi: 10.1021/acs.langmuir.0c00109
    [28] 余 娟. 龙打岩洞空气负离子时空分布特征及影响因素研究[D]. 贵阳: 贵州师范大学, 2021.

    [29] 李少宁, 李 嫒, 鲁绍伟, 等. 北京西山国家森林公园中空气负离子浓度与气象因子的相关性研究[J]. 生态环境学报, 2021, 30(3):541-547.

    [30] 齐 冰, 杜荣光, 邵碧嘉. 杭州市空气负离子变化特征分析[J]. 气象与减灾研究, 2011, 34(4):68-71.

    [31] 陈 欢, 章家恩. 空气负离子浓度分布的影响因素研究综述[J]. 生态科学, 2010, 29(2):181-185. doi: 10.3969/j.issn.1008-8873.2010.02.016

    [32] 赵怡宁, 史常青, 许荡飞, 等. 崇礼区典型林分空气负离子浓度及影响因素[J]. 林业科学研究, 2018, 31(3):127-135.

    [33] 李萌萌. 河北太行山低山区水土保持林空气离子特征研究[D]. 保定: 河北农业大学, 2014.

    [34] 李少宁, 李 嫒, 赵 旭, 等. 北京西山国家森林公园空气负离子与大气污染物关系研究[J]. 西南农业学报, 2021, 34(10):2269-2273.

    [35] 杜万光, 王 成, 王 茜, 等. 北京香山公园主要植被类型的夏季环境效应评价[J]. 林业科学, 2018, 54(4):155-164.

    [36] 潘剑彬, 李佳妮, 李树华, 等. 城市绿地植物群落与空气负离子空间分异特征相关关系研究——以北京奥林匹克森林公园为例[J]. 中国园林, 2022, 38(6):57-62.

    [37] 童 睿, 阚丽虹, 朱中生. 基于Logistic回归和随机森林的心力衰竭预后预测建模[J]. 复旦学报(医学版), 2022, 49(5):656-664.

    [38] 冷寒松, 张丰铎, 任立杰, 等. 基于随机森林算法和摩擦起电传感器的人体运动识别研究[J]. 河北水利电力学院学报, 2022, 32(2):29-34.

    [39] 李爱民, 王海隆, 许有成. 优化随机森林算法的城市湖泊DOC质量浓度遥感反演[J]. 郑州大学学报(工学版), 2022, 43(6):90-96.

    [40] 孙雪莲, 舒清态, 欧光龙, 等. 基于随机森林回归模型的思茅松人工林生物量遥感估测[J]. 林业资源管理, 2015(1):71-76.

    [41] 余 海, 郭 嘉, 李恩杰, 等. 北京九龙山侧柏林空气负离子时空分布特征[J]. 林业科学研究, 2021, 34(3):174-179.

    [42] 周 斌, 余树全, 张 超, 等. 不同树种林分对空气负离子浓度的影响[J]. 浙江农林大学学报, 2011, 28(2):200-206.

    [43] 祁舒展. 滨水植物群落对气象因子与空气质量的影响[D]. 南京: 南京农业大学, 2017.

    [44] 金竹秀. 临安市城区绿地植物群落结构与生态效益研究[D]. 杭州: 浙江农林大学, 2011.

    [45] 陈亚静. 游憩小径康养功能提升设计研究[D]. 泰安: 山东农业大学, 2020.

    [46] 袁瑞瑞, 黄萧霖, 郝 璐. 近40年中国饱和水汽压差时空变化及影响因素分析[J]. 气候与环境研究, 2021, 26(4):413-424.

    [47] 王 珣, 杨小龙, 叶子飘, 等. 不同温度下甜高粱叶片光合作用的气孔限制和非气孔限制特征分析[J]. 植物生理学报, 2022, 58(7):1245-1253.

    [48] 叶彩华, 王晓云, 郭文利. 空气中负离子浓度与气象条件关系初探[J]. 气象科技, 2000(4):51-52.

    [49] 刘 洋, 段文标. 莲花湖库区水源涵养林负离子含量研究[J]. 森林工程, 2009, 25(1):8-12,20.

    [50] 裘彦挺, 吴志军, 尚冬杰, 等. 我国城市大气PM2.5与O3浓度相关性的时空特征分析[J]. 科学通报, 2022, 67(18):2008-2017.

    [51]

    MIAO S, ZHANG XY, HAN YJ, et al. Random forest algorithm for the relationship between negative air ions and environmental factors in an urban park[J]. Atmosphere, 2018, 9(12), 463.
    [52] 赵春彦, 司建华, 冯 起, 等. 风对极端干旱区胡杨蒸腾速率的影响[J]. 冰川冻土, 2015, 37(4):1104-1111.

    [53] 张 鹏. 不同饱和水汽压亏缺和土壤干旱环境对作物水分利用效率的影响机理及模拟研究[D]. 杨凌: 西北农林科技大学, 2021.

  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(3) / Tables(1)

Article views(2238) PDF downloads(66) Cited by()

Proportional views

Variation of Negative Air Ions and Its Influencing Factors in Typical Plantations in Rocky Mountain Area of North China

    Corresponding author: SANG Yu -qiang, syuqiang@163.com
  • 1. College of Forestry, Henan Agricultural University, Zhengzhou 450002, Henan, China
  • 2. Research Institute of Forestry, Chinese Academy of Forestry,Key Laboratory of Tree Breeding and Cultivation of National Forestry and Grasland Administration, Beijing 100091, China
  • 3. Henan Xiaolangdi Earth Critical Zone National Research Station on the Middle Yellow River, Jiyuan 459000, Henan, China
  • 4. Jiyuan Forestry Workstation, Jiyuan 459000, Henan, China
  • 5. Nanshan National Forest Farm, Jiyuan 459000, Henan, China

Abstract:  Objective To explore the variation characteristics of Negative air ion (NAI) of typical plantations and the relationship between NAI and environmental factors in growing period and non-growing period in rocky mountain area of north China, and reveal environmental factors affecting NAI in different seasons in this area. Method The variation characteristics of NAI of typical plantations in Henan Xiaolangdi Earth Critical Zone National Research Station were measured from May to December in 2021 using the air negative ions, PM 2.5, PM 10 and meteorological data. Besides, the main environmental factors and variable importance measures affecting NAI of typical plantations in the area were analyzed by random forest algorithm. Results The diurnal variation of NAI in Quercus variabilis showed single peak curve in growing period but not obvious in non-growing period. The NAI of Platycladus orientalis showed single peak cure during the experiment period. The NAI concentration of Quercus variabilis plantations (740.32 ion·cm−3) was higher than that of Platycladus orientalis (703.74 ion·cm−3) during the observation period. The daily NAI of Quercus variabilis (858.94 ion·cm−3) was higher than that of Platycladus orientalis (724.33 ion·cm−3) during the growing period. The daily NAI of Quercus variabilis (621.70 ion·cm−3) was lower than that of Platycladus orientalis (683.16 ion·cm−3) during the non-growing period. The meteorological factors such as Air temperature (Ta), Relative humidity (RH), Vapor pressure deficit (VPD) and photosynthetically active radiation (PAR) in the growing period were higher than those in the non-growing period, while the particulate matter (PM2.5 and PM10) in the non-growing period was higher than that in the growing period. The concentration of PM10 was higher than that of PM2.5 during the experiment period. There was no significant difference between Wind speed (WS) between the growing period and non-growing period. The random forest method revealed that the main environmental factors affecting the NAI concentrations of Quercus variabilis and Platycladus orientalis in the growing period were VPD, PAR and WS, and their variable importance measures were 20.22, 15.08 ,14.71, respectively, and 25.08, 16.76, 16.49, respectively. The main environmental factors affecting the NAI concentration of Quercus variabilis and Platycladus orientalis during the non-growth period were PM 2.5, WS and PM 10, and their variable importance measures were 33.36, 17.58, 14.28, respectively, and 15.89, 17.51, 14.62, respectively. Conclusion The diurnal variation of NAI concentration of Quercus variabilis and Platycladus orientalis in growing period both showed a single peak curve; the diurnal variation of NAI concentration of Quercus variabilis plantations was not obvious, while the diurnal variation NAI concentration of Platycladus orientalis plantations showed a single peak curve in non-growing period. There were significant differences in NAI concentration between Quercus variabilis and Platycladus orientalis plantations, NAI concentration of Quercus variabilis was higher than that of Platycladus orientalis during the growing period, while NAI concentration of Quercus variabilis was lower than that of Platycladus orientalis during the non-growing period. The NAI concentration of Quercus variabilis was higher than that of Platycladus orientalis during the observation period. Differences of environmental factors affecting NAI of typical plantations in the area were obvious. VPD and PAR were the key factors during the growing period, while PM2.5, PM 10 and WS were the key factors during the non-growing period.

  • 空气负离子(Negative air ion,NAI)是指由于空气中氧分子因其化学性质优先获得自由电子而带负电荷的离子或离子团[1-2]。NAI具有降尘杀菌[3]、有效清除空气中有机污染物[4]、增强人体免疫系统和协助治疗各种疾病的功能[1,5-8],因而,被称为“空气中的维生素”[3,9],已经成为衡量空气质量优劣的重要指标之一[10-11]。根据来源不同,NAI可以分为物理和生物来源2大类,物理来源以宇宙辐射、雷电活动、降雨、土壤放射性成分、水分子分解等为主[4,12];生物来源以植物光合作用、光电效应、针叶植物尖端放电等为主[13-14]。NAI大小因受多种环境因素如气候条件[4,15]、颗粒物[16-18]等限制导致其影响因素多变复杂[19-20]

    目前,对NAI的研究集中于不同林分配比[21-23]下NAI浓度的时空变化[2,11,13,24-25]及其影响因素[1,4,26]、开发利用[7,27]等。森林作为陆地生态系统中产生NAI的重要场所之一[4,19,23],通过植物的光合作用和尖端放电产生大量的空气负离子,并且在森林生态系统复杂的结构和环境因子的影响下延长了NAI的存留时间[11,23]。现有研究表明,不同时期环境因子对植物的NAI浓度影响不同,如Wang等[11]发现,空气温度、相对湿度、风速、颗粒物与NAI的相关性四季均有差异;余娟[28]则发现,夏季空气温度、相对湿度与NAI相关性与春秋冬季节相反;李少宁等[29]则认为,夏季空气温度、相对湿度、太阳辐射与NAI的相关性与冬季时期相反。因研究对象和研究区域的不同,导致当前NAI与环境因子的关系尚未形成统一的结论[4,30-31]。当前,关于华北山区的NAI虽然已有不少研究[29,32-36],但研究区域多位于华北土石山区北端,且研究对象以混交林为主,研究时段多仅限于林分生长季NAI浓度变化及影响因素进行比较,鲜有将不同生长季针叶、阔叶林的NAI浓度变化及影响因素[29,33,34]进行比较。因此,本研究以河南黄河小浪底地球关键带国家野外科学观测研究站的典型人工林侧柏(Platycladus orientalis L. Franco)和栓皮栎(Quercus variabilis Blume)为试验对象,利用随机森林模型(Random Forest, RF)探究该地区生长季与非生长季栓皮栎和侧柏人工林NAI的变化特征以及其影响因子,旨在揭示栓皮栎和侧柏人工林NAI的差异性及不同季节影响NAI的主导因子。

    • 研究区位于河南黄河小浪底地球关键带国家野外科学观测研究站(35°01′45″ N,112°28′08″ E),地处黄土丘陵-南太行交错带,平均海拔410 m,该地区属暖温带大陆季风气候,年平均气温13.4 ℃,全年日照时数2 367.7 h,年平均降水量641.7 mm,年蒸发量1 400 mm,无霜期220~230 d,0 ℃以上年平均有效积温 5 282 ℃,10 ℃以上年均积温达 4 847 ℃,植物生长期为210~220 d。受季风气候的影响,降水季节性分配不均匀,6—9月平均降水量438.0 mm,占全年的68.3%。栓皮栎人工林林龄46 a,平均树高10 m,林分密度998株·hm−2,郁闭度0.75;侧柏人工林林龄30 a,平均树高8 m,林分密度625株·hm−2,郁闭度0.80。

    2.   研究方法
    • 采用RR-9411A型空气负离子自动监测仪(北京雨根公司,中国)同步观测NAI、PM2.5和PM10。该仪器测量部分对进气口吸入的空气进行测量,测量范围0~1.2 × 106 ion·cm−3,测量气流750 cm3·s−1,分辨率10 ions·cm−3,测量精度≤ ± 10%,采集频率1 s,存储周期10 min。在侧柏与栓皮栎人工林的综合观测塔距离地面垂直高度5 m处均安装一台仪器进行实时观测,观测时间为2021年5—12月。利用自动气象站系统(北京雨根公司,中国)同步监测记录空气温度(TA)、光合有效辐射(PAR)、相对湿度(RH)、风速(WS)等气象因子。

      饱和水汽压亏缺(VPD)是表示空气温度和相对湿度的一个综合指标,相比于空气温度和相对湿度更具有表性,利用经验公式[4]计算VPD,其计算公式为:

        式中:Ta为空气温度,RH为相对湿度,取值(0,1)。

    • 随机森林模型是在变量和数据的使用上进行随机抽样,生成一定数量的决策树,再将决策树的结果进行汇总得出最终结果,能很好地解决单一决策树过拟合的问题[37-39]。利用Bootsrap重抽样方法从原始样本中抽取多个样本,对每个Bootsrap样本进行决策树建模,然后组合多棵决策树的预测,通过投票得出最终预测结果[4,40]。模型拟合效果采用决定系数($ \mathit{R}) $对模拟结果进行精度检验,采用重要性得分$ {\mathit{V}\mathit{I}}_{\mathit{n}}\left({\mathit{X}}_{\mathit{j}}\right) $对所选变量进行排序,计算公式如下:

        式中:$ {o}_{i} $$ {p}_{i} $分别为NAI的观测值和模型拟合值;$ \stackrel{-}{{o}_{i}} $为观测值的均值。$ {X}_{i} $为输入变量之一, $ NOOB $为袋外样本数;$ f\left({X}_{i}\right) $为袋外数据中第$ i $个观测值;$ {f}_{n}\left({X}_{i}\right) $为在随机置换变量$ {X}_{i} $的观测值前第$ n $株树上袋外数据第$ i $个观测值所对应的预测值;${f}_{n}\left({{X{{\;{'}}}}_{i}}\right)$为在随机置换变量$ {X}_{j} $的观测值后第n株树上袋外数据第$ i $个观测值所对应的预测值;$ I[{f}_{i}\left({X}_{i}\right)={f}_{n}({X}_{i}\left)\right] $$I[{f}_{i}\left({X}_{i}\right)={f}_{n}({X{{\;{'}}}_{i}}\left)\right]$为判别函数,当$ {f}_{i}\left({X}_{i}\right)={f}_{n}\left({X}_{i}\right) $${f}_{i}\left({X}_{i}\right)={f}_{n}\left({X{{\;{'}}}_{i}}\right)$时,取值为1,否则为0[4]

    • 结合研究区人工林生长特点,本研究将观测期分为生长季(5—9月)和非生长季(10—12月),将生长季和非生长季每日对应各时刻的NAI数据和气象数据分别进行平均后,分析NAI和环境因子日变化特征。

      利用R语言对观测数据进行筛选,并剔除异常值,具体筛选标准详见文献[4],共筛选有效数据约7 400组。将生长季与非生长季栓皮栎和侧柏人工林的NAI与环境因子数据作为4个单独的模型样本,对随机森林模型进行训练。随机森林模型构建利用R语言Random Forest包来实现,采用Excel 2016、Origin2018等软件处理与分析NAI与环境因子的数据。使用SPSS 25.0软件进行单因素方差分析(One-way ANOVA),其中,p<0.05为显著,p<0.01为极显著。

    3.   结果与分析
    • 图1表明:不同生长季栓皮栎、侧柏人工林NAI浓度日内变化特征差异明显。生长季,栓皮栎、侧柏人工林的NAI浓度日内变化均呈单峰变化趋势,二者峰值均出现在9:00(分别为966.13 ± 153.82、821.03 ± 122.544 ion·cm−3),栓皮栎波谷出现在17:00(767.08 ± 156.40 ion·cm−3),侧柏波谷出现在21:00(630.55 ± 173.04 ion·cm−3)。非生长季,栓皮栎人工林因处于落叶休眠期,NAI浓度日内变化不明显,整体上较平缓且数值较低,日均NAI浓度为621.70 ± 76.90 ion·cm−3,而侧柏人工林呈单峰变化趋势,与生长季相比,峰值略有提前,出现在8:00(769.09 ± 111.21 ion·cm−3),最低值出现在18:00(636.47 ± 55.05 ion·cm−3)。生长季,人工林日均NAI浓度栓皮栎(858.94 ± 97.52 ion·cm−3)>侧柏(724.33 ± 93.78 ion·cm−3)(p<0.01);非生长季,人工林日均NAI浓度侧柏(683.16 ± 60.83 ion·cm−3)>栓皮栎(621.70 ± 76.90 ion·cm−3)(p<0.01)。整个观测期间,人工林日内平均NAI浓度栓皮栎(740.32 ± 87.21 ion·cm−3)>侧柏(703.74 ± 77.74 ion·cm−3)(p<0.01)

      Figure 1.  Diurnal variation of NAI concentration of Quercus variabilis and Platycladus orientalis plantations during the growing and non-growing season

    • 不同生长季华北土石山区主要环境因子日内变化(图2)表明:生长季和非生长季空气温度(Ta)均呈单峰变化趋势,峰值均出现在15:00左右,生长季日均Ta(23.91 ± 3.69 ℃)大于非生长季(11.48 ± 3.46 ℃)(p<0.01)。与Ta相反,相对湿度(RH)在生长季和非生长季均呈单谷变化趋势,谷值出现在13:00—14:00,生长季日均RH(62.30% ± 12.84%)大于非生长季(48.00% ± 13.52%)(p<0.01)。饱和水汽压亏缺(VPD)也呈单峰变化趋势,生长季与非生长季峰值出现在14:00—15:00,生长季日均VPD(138.63 ± 58.63 kPa)远大于非生长季(72.37 ± 27.12 kPa)(p<0.01)。生长季与非生长季风速(WS)整体均呈单峰变化趋势,生长季的峰值出现在21:00,非生长季峰值出现在14:00,生长季的日均WS(2.31 ± 0.29 m·s−1)略高于非生长季的WS(2.29 ± 0.51 m·s−1)(p<0.01),差别不大。生长季和非生长季的光和有效辐射(PAR)变化趋势一致,峰值均出现在12:00,但大小相差727.92 μmol·m−2·s−1,生长季PAR日均值(938.24 ± 68.60 μmol·m−2·s−1)高于非生长季(398.80 ± 65.22 μmol·m−2·s−1)(p<0.01)。生长季PM2.5与PM10日内变化均呈单谷变化趋势,其最低值均出现在15:00。非生长季PM2.5与PM10日内变化均呈双峰变化趋势,PM2.5与PM10首个峰值均出现在9:00,第2个峰值均出现在21:00。非生长季颗粒物浓度远大于生长季,PM2.5与PM10均值分别为103.17 ± 48.24、162.05 ± 75.32 μg·cm−3和35.92 ± 11.67、76.53 ± 31.25 μg·cm−3p<0.01)。生长季与非生长季PM10均高于同期的PM2.5。

      Figure 2.  Diurnal variation characteristics of environmental factors in growing and non-growing season

    • 根据随机森林算法对栓皮栎和侧柏人工林在生长季和非生长季的4个样本数据进行分析,模拟得到4个样本的重要性得分$ {VI}_{n}\left({X}_{j}\right) $,同时根据对不同树种不同时期的重要性得分的高低进行排序(图3)。生长季环境因子对栓皮栎人工林NAI浓度影响排序为:VPD(20.22)>PAR(15.08)>WS(14.71)>Ta(13.75)>RH(12.41)>PM2.5(12.24)>PM10(11.58);环境因子对侧柏人工林NAI浓度影响排序为:VPD(25.08)>WS(16.76)>PAR(16.49)>Ta(13.93)>PM2.5(9.69)>RH(9.35)>PM10(8.29)。生长季决定栓皮栎和侧柏NAI浓度大小的决定性因子为VPD,其次为PAR和WS,颗粒物如PM2.5和PM10影响很小。非生长季环境因子对栓皮栎人工林的NAI浓度影响排序为:PM2.5(33.36)>WS(17.58)>PM10(14.28)>Ta(12.55)>RH(8.915)>PAR(6.83)>VPD(6.49);环境因子对侧柏人工林的NAI浓度影响排序为:WS(17.51)>PM2.5(15.89)>PM10(14.62)>RH(14.26)>Ta(13.96)>VPD(11.97)>PAR(11.77)。非生长季影响栓皮栎和侧柏NAI浓度大小的关键因子为PM2.5、PM10和WS,气象因子的影响很小。利用对应4组数据集对结果进一步分析(表1)表明:使用随机森林模型得到栓皮栎和侧柏人工林在生长季与非生长季的方差解释率分别为87.6%、88.7%和87.4%、87.5%,同时NAI预测值与观测值的决定系数R值均在0.92以上,说明该模型的预测精度较好。

      Figure 3.  The random forest model outputs the ranking of the importance of environmental factors of Quercus variabilis and Platycladus orientalis plantations in the growing and non-growing season

      样本
      Sample
      方差解释率
      Variance
      explained/%
      决定
      系数
      R
      栓皮栎生长季
      Quercus variabilis growing season
      87.60.924
      侧柏生长季
      Platycladus orientalis growing season
      88.70.942
      栓皮栎非生长季
      Quercus variabilis non-growing season
      87.40.921
      侧柏非生长季
      Platycladus orientalis non-growing season
      87.50.942

      Table 1.  The coefficient of determination for the random forest model

    4.   讨论
    • 生长季,栓皮栎和侧柏NAI浓度日内变化均呈单峰曲线,与Ta、VPD、PAR变化趋势相近,与RH相反,该结果与他人的研究结果一致,但峰值出现时间存在差异,如本研究发现,栓皮栎和侧柏NAI浓度峰值均出现在9:00左右;李少宁等[29]发现,北京西山森林公园的NAI浓度峰值出现在10:00;李萌萌[33]发现,栓皮栎的NAI浓度呈双峰曲线;余海等[41]发现,不同季节北京九龙山侧柏的NAI浓度曲线和峰值出现时间不同。包红光等[16]发现,呼和浩特城市公园不同植被配置、不同季节NAI浓度趋势差别较大,说明NAI浓度因植被种类、林分配置等表现出一定的差异性,甚至相同树种因研究区域、测定时间也存在不确定性,也进一步说明了NAI浓度的多变性及其环境因子的复杂性与不确定性。

      非生长季,栓皮栎NAI浓度的日内变化规律不明显,且NAI浓度低于侧柏的,与前人研究结果一致[14,42-45]。主要原因在于该阶段栓皮栎已落叶,进入休眠期,其光合作用十分微弱,导致NAI浓度低,且无规律可循。侧柏人工林NAI浓度日内变化呈单峰曲线,而PM2.5与PM10颗粒物则呈单谷曲线。其原因在于NAI带负电,极不稳定,易与带正电性颗粒物[19]相互附着而形成大分子沉淀物[4],导致NAI浓度降低。生长季,栓皮栎人工林NAI浓度高于侧柏,主要原因在于栓皮栎为阔叶树种,叶面积大于针叶林侧柏,光合作用强,进而产生了较多的NAI;而在非生长季,栓皮栎落叶导致NAI降低,而侧柏作为常绿针叶树种,通过微弱光合作用和针叶植物尖端放电的优势产生NAI,因此,侧柏林NAI高于栓皮栎林。

    • NAI的产生除跟植物自身有关外,环境因子的影响不可忽视。本研究利用随机森林法得到生长季影响栓皮栎和侧柏人工林NAI的环境因子重要性得分从大到小分别为VPD、PAR和WS,颗粒物如PM2.5和PM10的重要性得分则很小。VPD作为空气温湿度的综合表现指标,影响叶片气孔开闭,对植物生理功能起着关键作用[46]。Ta是影响植物光合作用的重要环境因子之一[47],空气温度的升高会增加分子的运动和碰撞[4],提高了氧分子的电离,有助于NAI的形成[48]。PAR是植物光合作用的必须条件之一,PAR增强,植物能通过光合作用向空气中释放大量氧气,氧气具有较强吸附空气中自由电子的能力,同时伴随太阳辐射为空气中分子间的碰撞提供了能量[15,49]。Wang等[11]通过随机森林算法在黑龙江五大连池风景区得出,在森林中环境因子重要性得分为:O3>PM10>Ta>太阳辐射,原因在于其研究区纬度较高,易产生高浓度的O3,在地表氧化产生更多的二次微粒吸附NAI[50];Miao等[51]在城市公园所得环境因子的重要性得分为:相对湿度>辐射>空气温度>PM2.5,该研究区处于亚热带季风气候,为NAI的产生提供了有利条件;充沛的降雨,导致相对湿度保持在较高水平,相对湿度的增加会提供大量的水分子,促进NAI的形成,相对湿度会对空气中的颗粒物产生耦合效应[11],相对湿度的增加加快了颗粒物的扩散速率,颗粒物的减少堆积,进而提高空气中的NAI存活时间,维持了NAI浓度。Shi等[1]发现,影响栓皮栎林的主要环境因子重要性得分为:PM2.5>土壤湿度>空气温度>相对湿度,与本文结果差异较大。对比发现,该地区2019年生长季PM2.5浓度为2021年的1.5倍,高浓度的PM2.5抑制了气象因子对栓皮栎NAI的作用。

      非生长季影响栓皮栎和侧柏人工林NAI的环境因子重要性得分从大到小分别为PM2.5、PM10和WS,其余气象因子的影响很小。主要原因在于冬季北方城市因用煤取暖、汽车尾气排放等原因导致空气中存在了大量的颗粒物,加上逆温天气的存在使得颗粒物不容易散去。本研究发现,栓皮栎人工林PM2.5重要性得分为33.36,远高于侧柏(15.89),可能与栓皮栎林试验区所处的微地形、距离市区更近有关。整个观测期,WS对栓皮栎和侧柏NAI的重要性得分均比较大,主要原因在于WS导致叶片内外水汽压亏缺变大,叶片气孔开度变大[52-53],导致植物光合作用增强,引起NAI浓度上升。在微观角度上,风能够加快分子间的碰撞,加快分子转化成离子状态的过程[15];同时,风速的增加会导致更多气体进入观测仪器中,从而增加了仪器采集NAI量。

    5.   结论
    • (1)华北土石山区栓皮栎和侧柏人工林在生长季和非生长季NAI变化趋势及浓度差异显著:栓皮栎与侧柏生长季NAI浓度日内变化均呈单峰曲线,非生长季栓皮栎NAI浓度则不明显,而侧柏NAI浓度呈单峰曲线。生长季栓皮栎NAI浓度高于侧柏,非生长季则相反,观测期内栓皮栎平均NAI浓度高于侧柏。

      (2)影响栓皮栎和侧柏人工林NAI的环境因子差异不明显,但不同生长季差异显著,生长季主要以VPD和PAR为主,非生长季主要以PM2.5、PM10和WS为主。

Reference (53)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return