基于随机森林模型的毛竹林CO2通量模拟及其影响因子

doi:10.11707/j.1001-7488.20180801

摘要/Abstract

摘要： [目的]探究环境因子对毛竹林CO₂通量的影响，为亚热带典型森林碳循环模拟提供技术和理论支撑。[方法]基于浙江省安吉县山川乡2011—2014年毛竹林通量塔观测数据，采用随机森林模型对毛竹林CO₂通量进行模拟，以均方根误差（RMSE）、决定系数（R²）和林氏调和系数（LCCC）3个指标评价模型精度，并通过计算环境因子重要性得分来分析环境因子对毛竹林CO₂通量的影响。[结果]随机森林模型能以较高精度模拟毛竹林CO₂通量，但由于2013年7和8月发生极端高温干旱，模型难以刻画这种短时、剧烈的干扰，最终导致测试阶段模型精度（R²=0.845 5，RMSE=0.437 7 mg·m^-2s^-1，LCCC=0.914 1）低于训练阶段模型精度（R²=0.961 5，RMSE=0.005 4 mg·m^-2s^-1，LCCC=0.980 1）；十折交叉验证表明，随机森林模型拟合效果稳定，且模型内部参数设置合理，模型误差主要来自于输入数据；在月尺度上，环境因子对毛竹林CO₂通量影响的重要性得分表现为光合有效辐射PAR （63.332）> 土壤5 cm深处温度T_S（29.932）> 空气相对湿度R_H（25.839）> 大气温度T_A（25.581）> 空气CO₂浓度C_CO₂（25.095）> 饱和水汽压差VPD （24.123）> 风速W_S（23.504）> 生态系统有效能量A_E（19.323）> 土壤热通量Q_S（18.502），PAR对毛竹林CO₂通量变化影响最大，PAR、T_S和VPD对毛竹林CO₂通量的影响较显著（P<0.05），这3个因子是影响月尺度上毛竹林CO₂通量变化的主导因子。[结论]随机森林模型能以较高精度拟合毛竹林CO₂通量；在相关环境因子中，光合有效辐射、土壤5 cm深处温度和饱合水汽压差对毛竹林CO₂通量影响的贡献最显著，这3个因子对毛竹林CO₂通量月尺度上的变化具有控制作用。

Abstract: [Objective]This paper aims to investigate the influence of environmental factors on the CO₂ flux of moso bamboo forest and to provide technical and theoretical support for carbon cycle simulation of typical subtropical forests.[Method]The CO₂ flux of moso bamboo forest was simulated using random forest model based on eddy covariance flux data collected from 2011 to 2014, and the accuracy of model was evaluated using the mean squared root error (RMSE), the coefficient of determination (R²) and the lin's concordance correlation coefficient (LCCC). The contribution of each environmental factor to CO₂ flux was analyzed based on importance score calculated using the random forest algorithm.[Result]The random forest model accuracy of testing data (R²=0.845 5, RMSE=0.437 7 mg·m^-2s^-1; LCCC=0.914 1) was lower than that of training data (R²=0.961 5; RMSE=0.005 4 mg·m^-2s^-1; LCCC=0.980 1), because the model was hard to accurately depict this kind of short, intense interference to CO₂ flux during the extreme drought occurred in July and August in 2013. The accuracy was stable for different training data based on 10-fold cross-validation method and the parameters of the model were appropriately set. The error in the model was mainly caused by the input data. The importance score of each environmental factor was decreased in the following order:PAR (63.332) > T_S (29.932) > R_H (25.839) > T_A (25.581) > C_CO₂ (25.095) > VPD (24.123) > W_S(23.504) > A_E (19.323) > Q_S(18.502). PAR was the dominant factor for explaining the change of CO₂ flux in moso bamboo forest. Based on the significance test, monthly CO₂ flux was significantly influenced by PAR, T_S, and VPD (P<0.05).[Conclusion]The random forest model can simulate the CO₂ flux of moso bamboo with a high accuracy; PAR, T_S, and VPD remarkably affect the CO₂ flux at the 0.05 significance level, according to their importance score, indicating that PAR, T_S, and VPD play an important role in controlling the CO₂ flux of moso bamboo forest at a monthly time scale.

Key words: moso bamboo forest, eddy covariance, flux tower, CO₂ flux, random forest model, environmental factors, Anji County

中图分类号:

S718.55⁺6

陈亮, 周国模, 杜华强, 刘玉莉, 毛方杰, 徐小军, 李雪建, 崔璐, 李阳光, 朱迪恩. 基于随机森林模型的毛竹林CO₂通量模拟及其影响因子[J]. 林业科学, 2018, 54(8): 1-12.

Chen Liang, Zhou Guomo, Du Huaqiang, Liu Yuli, Mao Fangjie, Xu Xiaojun, Li Xuejian, Cui Lu, Li Yangguang, Zhu Di. Simulation of CO₂ Flux and Controlling Factors in Moso Bamboo Forest Using Random Forest Algorithm[J]. Scientia Silvae Sinicae, 2018, 54(8): 1-12.

参考文献

方精云,于贵瑞,任小波,等. 2015. 中国陆地生态系统固碳效应——中国科学院战略性先导科技专项"应对气候变化的碳收支认证及相关问题"之生态系统固碳任务群研究进展. 中国科学院院刊,30(6):848-857.
(Fang J Y, Yu G R, Ren X B, et al. 2015. Carbon sequestration in China's terrestrial ecosystems under climate change—progress on ecosystem carbon sequestration from the CAS strategic priority research program. Bulletin of Chinese Academy of Sciences, 30(6):848-857. [in Chinese])
李旭青,刘湘南,刘美玲,等. 2014. 水稻冠层氮素含量光谱反演的随机森林算法及区域应用. 遥感学报,18(4):923-945.
(Li X Q, Liu X N, Liu M L, et al. 2014. Random forest algorithm and regional applications of spectral inversion model for estimating canopy nitrogen concentration in rice. Journal of Remote Sensing, 18(4):923-945. [in Chinese])
刘敏,郎荣玲,曹永斌. 2015. 随机森林中树数量. 计算机工程与应用,51(5):129-131.
(Liu M, Lang R L, Cao Y B. 2015. Number of trees in random forest. Computer Engineering and Applications, 51(5):126-131. [in Chinese])
彭镇华,王妍,任海青,等. 2009. 安庆杨树林生态系统碳通量及其影响因子研究. 林业科学研究,22(2):237-242.
(Peng Z H, Wang Y, Ren H Q, et al. 2009. Research on the variation of carbon flux and the relationship of environmental factors and carbon flux of populous forest. Forest Research, 22(2):237-242. [in Chinese])
孙成,江洪,周国模,等. 2013. 我国亚热带毛竹林CO₂变异特征. 应用生态学报,24(10):2717-2724.
(Sun C, Jiang H, Zhou G M, et al. 2013. Variation characteristics of CO2 flux in Phyllostachys edulis forest ecosystem in subtropical region of China. Journal of Applied Ecology, 24(10):2717-2724. [in Chinese])
王春林,于贵瑞,周国逸,等. 2006. 鼎湖山常绿针阔叶混交林CO₂通量估算. 中国科学D辑:地球科学, 36(suppl.1):119－129.
(Wang C L, Yu G R, Zhou G Y, et al. 2006. CO2 flux evaluation over the evergreen coniferous and broad-leaved mixed forest in Dinghushan, China. Science in China Series D:Earth Sciences, 36(suppl.1):119－129. [in Chinese])
王兴昌, 王传宽. 2015. 森林生态系统碳循环的基本概念和野外测定方法评述. 生态学报, 35(13):4241-4256.
(Wang X C, Wang C K. 2015. Fundamental concepts and field measurement methods of carbon cycling in forest ecosystems:a review. Acta Ecologica Sinica, 35(13):4241-4256. [in Chinese])
于贵瑞,孙晓敏. 2006. 陆地生态系统通量观测的原理与方法. 北京:高等教育出版社,195, 222, 218.
(Yu G R, Sun X M.2006. Principles of flux measuremnet in terrestrial ecosystems. Beijing:Higher Education Press,195, 222, 218. [in Chinese])
张军辉,于贵瑞,韩士杰,等. 2006. 长白山阔叶红松林CO₂通量季节和年际变化特征及控制机制,中国科学.D辑:地球科学, 36(S1):60-69.
(Zhang J H, Yu G R, Han S J,et al. 2006. Seasonal and annual variation of CO₂ flux above a broad-leaved korean pine mixed forest.Science in China Series D:Earth Sciences, 36 (S1):60 -69. [in Chinese])
张雷,王琳琳,张旭东,等. 2014. 随机森林算法基本思想及在生态学中的应用——以云南松分布模拟为例. 生态学报, 34(3):650-659.
(Zhang L, Wang L L, Zhang X D, et al.2014. The basic principle of random forest and its applications in ecology:a case study of Pinus yunnanensis. Acta Ecologica Sinica, 34(3):650-659. [in Chinese])
Allard V, Ourcival J M, Rambal S, et al. 2008. Seasonal and annual variation of carbon exchange in an evergreen mediterranean forest in southern France. Global Change Biology, 14 (4):714-725.
Baldocchi D, Falge E, Gu L H, et al. 2001. Fluxnet:a new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities. Bulletin of the American Meteorological Society, 82(11):2415-2434.
Barr A G, Black T A, Hogg E H, et al. 2007. Climatic controls on the carbon and water balances of a boreal aspen forest, 1994—2003. Global Change Biology, 13 (3):561-576.
Beer C, Reichstein M, Tomelleri E, et al. 2010. Terrestrial gross carbon dioxide uptake:global distribution and covariation with climate. Science, 329(5993):834-838.
Breiman L. 2001. Random forests. Machine Learning, 45 (1):5-32.
Campbell J L, Rustad L E, Boyer E W, et al. 2009. Consequences of climate change for biogeochemical cycling in forests of northeastern North America. Canadian Journal of Forest Research, 39(2):264-284.
Delgado-Baquerizo M, Maestre F T, Reich P B, et al. 2016. Microbial diversity drives multifunctionality in terrestrial ecosystems. Nature Communications, 7(10541):10541.
Desai A R, Bolstad P V, Cook B D, et al. 2005. Comparing net ecosystem exchange of carbon dioxide between an old-growth and mature forest in the upper midwest, USA. Agricultural and Forest Meteorology, 128(1/2):33-55.
Falge E, Baldocchi D, Olson R, et al. 2001. Gap filling strategies for defensible annual sums of net ecosystem exchange. Agricultural and Forest Meteorology, 107(1):43-69.
Fernández-Martínez M, Vicca S, Janssens I A, et al. 2014. Spatial variability and controls over biomass stocks, carbon fluxes, and resource-use efficiencies across forest ecosystems. Trees, 28(2):597-611.
Froelich N, Croft H, Chen J M, et al. 2015. Trends of carbon fluxes and climate over a mixed temperate-boreal transition forest in southern ontario, canada. Agricultural and Forest Meteorology, (211/212):72-84.
Gonsamo A, Croft H, Chen J M, et al. 2015. Radiation contributed more than temperature to increased decadal autumn and annual carbon uptake of two eastern north America mature forests. Agricultural and Forest Meteorology, 201(4):8-16.
Grossiord C, Granier A, Gessler A, et al. 2013. The influence of tree species mixture on ecosystem-level carbon accumulation and water use in a mixed boreal plantation. Forest Ecology and Management, 298(3):82-92.
He H L, Yu G R, Zhang L M, et al. 2006. Simulating CO₂ flux of three different ecosystems in China FLUX based on artificial neural networks. Science in China, 49(S2):252-261.
Jha C S, Thumaty K C, Rodda S R, et al. 2013. Analysis of carbon dioxide, water vapour and energy fluxes over an Indian teak mixed deciduous forest for winter and summer months using eddy covariance technique. Journal of Earth System Science, 122(5):1259-1268.
Katul G, Lai C T, Schäfer K, et al. 2001. Multiscale analysis of vegetation surface fluxes:from seconds to years. Advances in Water Resources, 24(9/10):1119-1132.
Keenan T F, Hollinger D Y, Bohrer G, et al. 2013. Increase in forest water-use efficiency as atmospheric carbon dioxide concentrations rise. Nature, 499(7458):324-327.
Law B E, Falge E, Gu L, et al. 2002. Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation. Agricultural and Forest Meteorology, 113(1/4):97-120.
Li X L, Jia Q Y, Liu J M. 2015. Seasonal variations in heat and carbon dioxide fluxes observed over a reed wetland in northeast China. Atmospheric Environment, 127:6-13.
Lin L I. 1989.A concordance correlation coefficient to evaluate reproducibility. Biometrics, 45(1):255-268.
Lloyd J, Taylor J A. 1994.On the temperature dependence of soil respiration. Functional Ecology, 8(3):315-323.
Lopatin J, Dolos K, Hernández H J, et al. 2016. Comparing generalized linear models and random forest to model vascular plant species richness using LiDAR data in a natural forest in central Chile. Remote Sensing of Environment, 173(315):200-210.
Rahmati O, Pourghasemi H R, Melesse A M. 2015. Application of GIS-based data driven random forest and maximum entropy models for groundwater potential mapping:a case study at Mehran Region, Iran. Catena, 137:360-372.
Richardson A D, Anderson R S, Arain M A, et al. 2012. Terrestrial biosphere models need better representation of vegetation phenology:results from the north American carbon program site synthesis. Global Change Biology, 18(2):566-584.
Saigusa N, Yamamoto S, Murayama S, et al. 2002. Gross primary production and net ecosystem exchange of a cool-temperate deciduous forest estimated by the eddy covariance method. Agricultural and Forest Meteorology,112(2):203-215.
Tang X L, Fang S H, Qi L H, et al. 2015. Soil respiration and net ecosystem production in relation to intensive management in moso bamboo forests. Catena, 137:219-228.
Wen X D, Zhao Z H, Deng X W, et al. 2014. Applying an artificial neural network to simulate and predict Chinese fir (Cunninghamia lanceolata) plantation carbon flux in subtropical China. Ecological Modelling, 294:19-26.
Wilson K B, Baldocchi D D, Hanson P J. 2001. Leaf age affects the seasonal pattern of photosynthetic capacityand net ecosystem exchange of carbon in a deciduous forest. Plant Cell and Environment, 24(6):571-583.
Xie J, Jia X, He G M, et al. 2015. Environmental control over seasonal variation in carbon fluxes of an urban temperate forest ecosystem. Landscape and Urban Planning, 142:63-70.
Yang R M, Zhang G L, Liu F, et al. 2016. Comparison of boosted regression tree and random forest models for mapping topsoil organic carbon concentration in an alpine ecosystem. Ecological Indicators, 60:870-878.
Yang Z, Chen J, Becker R, et al. 2014. Disentangling the confounding effects of PAR and air temperature on net ecosystem exchange in time and scale. Ecological Complexity, 19(10):74-75.
Yu G R, Chen Z, Piao S L, et al. 2014. High carbon dioxide uptake by subtropical forest ecosystems in the east Asian monsoon region. Proceedings of the National Academy of Sciences of the United States of America, 111(13):4910-4915.
Zhou G M, Xu X J, Du H Q, et al. 2011. Estimating aboveground carbon of moso bamboo forests using the k-nearest neighbors technique and satellite imagery. Photogrammetric Engineering and Remote Sensing, 77(11):1123-1131.

[1]	侯晓静, 明金科, 秦荣水, 朱霁平. 基于随机森林模型的交界域火灾风险分析[J]. 林业科学, 2019, 55(8): 194-200.
[2]	辛福梅, 闫小莉, 张长耀, 贾黎明. 西藏拉萨河谷区藏川杨和北京杨树干液流特征及其对环境因子的响应[J]. 林业科学, 2019, 55(2): 22-32.
[3]	任学敏, 朱雅, 陈兆进, 丁传雨, 李玉英, 杨改河. 太白山锐齿槲栎林乔木更新特征及其影响因子[J]. 林业科学, 2019, 55(1): 11-21.
[4]	杨水生, 徐秀英, 石道金. 劳动力老龄化对竹林地利用效率的影响——基于中介效应的实证研究[J]. 林业科学, 2018, 54(10): 132-142.
[5]	杜满义, 封焕英, 范少辉, 苏文会, 毛超, 唐晓鹿, 刘广路. 闽西毛竹林不同施肥处理下土壤有机碳含量垂直分布与季节动态[J]. 林业科学, 2017, 53(3): 12-20.
[6]	李翀, 周国模, 施拥军, 周宇峰, 徐林, 范叶青, 沈振明, 李少虹, 吕玉龙. 不同经营措施对毛竹林生态系统净碳汇能力的影响[J]. 林业科学, 2017, 53(2): 1-9.
[7]	高国龙, 杜华强, 韩凝, 徐小军, 孙少波, 李雪建. 基于特征优选的面向对象毛竹林分布信息提取[J]. 林业科学, 2016, 52(9): 77-85.
[8]	王海珍, 韩路, 徐雅丽, 牛建龙, 于军. 灰胡杨叶片气孔导度特征及数值模拟[J]. 林业科学, 2016, 52(1): 136-142.
[9]	徐小军, 周国模, 杜华强, 施拥军, 周宇峰. 缺失数据插补方法及其参数估计窗口大小对毛竹林CO₂通量估算的影响[J]. 林业科学, 2015, 51(9): 141-149.
[10]	王茜, 王成, 王艳英. 毛竹林森林浴对小白鼠自发行为的影响[J]. 林业科学, 2015, 51(5): 78-86.
[11]	刘会良, 张玲卫, 张宏祥, 布海丽且姆·阿卜杜热合曼, 张道远, 管开云. 基于物种分布模型的新疆野生果树物种丰富度分布格局[J]. 林业科学, 2015, 51(12): 1-8.
[12]	汤孟平, 沈利芬, 赵赛赛, 仇建习, 徐文兵, 庞春梅, 赵明水, 梅爱君. 基于谱分析的天目山毛竹林生长周期[J]. 林业科学, 2014, 50(9): 184-188.
[13]	郭宝华, 刘广路, 范少辉, 杜满义, 苏文会. 不同生产力水平毛竹林碳氮磷的分布格局和计量特征[J]. 林业科学, 2014, 50(6): 1-9.
[14]	李密, 周刚, 何振, 谭济才, 刘跃进, 伍义平. 油茶幼林茶角胸叶甲成虫空间分布型及其环境解释[J]. 林业科学, 2014, 50(10): 173-180.
[15]	徐小军, 周国模, 杜华强, 周宇峰, 胡军国, 陆国富. 样本分层对毛竹林地上部分碳储量估算精度的影响[J]. 林业科学, 2013, 49(6): 18-24.

基于随机森林模型的毛竹林CO₂通量模拟及其影响因子

Simulation of CO₂ Flux and Controlling Factors in Moso Bamboo Forest Using Random Forest Algorithm

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

作者中心

期刊获奖

引证指标

联系方式