[1]黄金君,舒清态,席磊,等.基于层次贝叶斯法的高山松生物量估测模型[J].江苏农业学报,2022,38(05):1265-1271.[doi:doi:10.3969/j.issn.1000-4440.2022.05.013]
　HUANG Jin-jun,SHU Qing-tai,XI Lei,et al.Research on biomass estimation model for Pinus densata based on hierarchical Bayesian method[J].,2022,38(05):1265-1271.[doi:doi:10.3969/j.issn.1000-4440.2022.05.013]
点击复制

基于层次贝叶斯法的高山松生物量估测模型()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
38
期数:
2022年05期
页码:
1265-1271
栏目:
农业信息工程
出版日期:
2022-10-31

文章信息/Info

Title:
Research on biomass estimation model for Pinus densata based on hierarchical Bayesian method
作者:
黄金君1; 2; 舒清态2; 席磊2; 孙杨2; 刘玥伶2
( 1.广西壮族自治区中国科学院广西植物研究所，广西桂林541006；2.西南林业大学林学院，云南昆明650224)
Author(s):
HUANG Jin-jun1; 2; SHU Qing-tai2; XI Lei2; SUN Yang2; LIU Yue-ling2
(1.Guangxi Institute of Botany, Chinese Academy of Sciences, Guilin 541006, China；2.College of Forestry, Southwest Forestry University, Kunming 650224, China)
关键词:
层次贝叶斯法; 非层次贝叶斯法; 异速生物量模型; 高山松
Keywords:
hierarchical Bayesian method; non-hierarchical Bayesian method; allometric biomass model; Pinus densata
分类号:
S711
DOI:
doi:10.3969/j.issn.1000-4440.2022.05.013
文献标志码:
A
摘要:
为研究层次贝叶斯法在高山松单木及不同组分生物量模型中的运用，基于香格里拉市Ⅰ区和Ⅱ区共115株高山松天然林数据，分别利用层次贝叶斯法与非层次贝叶斯法拟合高山松单木及各组分异速生物量模型，最后使用十折交叉方法进行模型精度验证。结果表明：(1)层次贝叶斯法拟合模型的效果优于非层次贝叶斯法，利用层次贝叶斯法拟合单木及不同组分生物量模型的决定系数（R2）精度提高区间为[0.000 1,0.012 0]，均方根误差（RMSE）降低区间为[0.03 kg,8.94 kg]，平均绝对误差（MAD）降低区间为[0.03 kg,3.31 kg]。(2)对比层次与非层次贝叶斯法拟合单木及不同组分生物量模型的结果发现，树干生物量、木材生物量和单木生物量模型效果最优，树皮生物量、树冠生物量和树枝生物量模型效果较优，树叶生物量模型效果较差。层次贝叶斯法拟合的R2区间为[0.365 0,0.965 0]，非层次贝叶斯法拟合的R2区间为[0.437 0,0.964 7]。与非层次贝叶斯法相比，层次贝叶斯法可以有效提高生物量模型的估测精度(树枝与树叶除外)，且这2种方法均可使用传统方法的估测结果作为先验信息，更新模型的参数值，提高建模灵活性。
Abstract:
To study the application of hierarchical Bayesian method in biomass models for single tree and different components of Pinus densata, hierarchical Bayesian method and non-hierarchical Bayesian method were used respectively to fit the allometric biomass models of single P. densata tree and differents tree components based on data of 115 P. densata trees from natural forests in Districts Ⅰ and Ⅱ of Shangri-La City. Finally, ten-fold crossover method was used to verify the accuracy of the models. The results showed that, firstly, effect of fitting model based on hierarchical Bayesian method was better than that based on non-hierarchical Bayesian method. The improvement interval of determination coefficient (R2) accuracy in biomass fitting models for single tree and different tree components by using hierarchical Bayesian method was [0.000 1, 0.012 0]. The reduction ranges of root-mean-square error (RMSE) and mean absolute deviation (MAD) were [0.03 kg, 8.94 kg] and [0.03 kg, 3.31 kg], respectively. Secondly, by comparing the results of hierarchical Bayesian and non-hierarchical Bayesian methods in fitting biomass models of single tree and different tree components, it was found that models for trunk biomass, wood biomass and single tree biomass showed the best effects, while models for bark biomass, crown biomass and branch biomass showed good results, but models for leaf biomass showed poor results. The R2 intervals fitted by hierarchical Bayesian method and non-hierarchical Bayesian method were [0.365 0, 0.965 0] and [0.437 0, 0.964 7], respectively. Compared with non-hierarchical Bayesian method, hierarchical Bayesian method can effectively improve the estimation accuracy of biomass models (except for branches and leaves), and both methods can use the estimated results of traditional methods as prior information to update the parameter values of the models and improve flexibility of modeling.

参考文献/References:

[1]钱春花,李明阳,郑超. 苏南丘陵山区森林生物量时空变化驱动因素分析[J].江苏农业学报,2021,37(2):382-388.
[2]陶惠林,冯海宽,徐良骥,等. 基于无人机高光谱遥感数据的冬小麦生物量估算[J].江苏农业学报,2020,36(5):1154-1162.
[3]张志,田昕,陈尔学,等. 森林地上生物量估测方法研究综述[J].北京林业大学学报,2011,33(5):144-150.
[4]MUUKKONEN P. Generalized allometric volume and biomass equations for some tree species in Europe[J]. European Journal of Forest Research,2007,126(2):157-166.
[5]JAVKHLAN N, MUNKH-ERDENE B, GUANGLIANG L, et al. Allometric equations for estimating above-ground biomass of Nitraria sibirica Pall. in Gobi Desert of Mongolia[J]. PLoS One,2020,15(9):1-11.
[6]张国斌,李秀芹,佘新松,等. 安徽岭南优势树种(组)生物量特征[J].林业科学,2012,48(5):136-140.
[7]薛春泉,徐期瑚,林丽平,等. 广东主要乡土阔叶树种含年龄和胸径的单木生物量模型[J].林业科学,2019,55(2):97-108.
[8]ZAPATA-CUARTAS M,SIERRA C A,ALLEMAN L. Probability distribution of allometric coefficients and Bayesian estimation of aboveground tree biomass[J]. Forest Ecology & Management,2012,277:173-179.
[9]GREEN E J, STRAWDERMAN W E. A comparison of hierarchical Bayes and empirical Bayes methods [J]. Forest Science, 1992, 38(2):350-366.
[10]GREEN E J, STRAWDERMAN W E. A Bayesian growth and yield model for slash pine plantations[J]. Journal of Applied Statistics,1996,23(2/3):285-299.
[11]CLARK J S, WOLOSIN M, DIETZE M, et al. Tree growth inference and prediction from diameter censuses and ring widths[J]. Ecological Applications,2007,17(7): 1942-1953.
[12]METCALF C J E, MCMAHON S M, CLARK J S. Overcoming data sparseness and parametric constraints in modeling of tree mortality: a new nonparametric Bayesian model[J]. Canadian Journal of Forest Research,2009,39(9): 1677-1687.
[13]张雄清,张建国,段爱国. 基于贝叶斯法估计杉木人工林树高生长模型[J].林业科学,2014,50(3):69-75.
[14]王冬至,张冬燕,李永宁,等. 基于贝叶斯法的针阔混交林树高与胸径混合效应模型[J].林业科学,2019,55(11):85-94.
[15]姚丹丹,徐奇刚,闫晓旺,等. 基于贝叶斯方法的蒙古栎林单木树高-胸径模型[J].南京林业大学学报(自然科学版),2020,44(1):131-137.
[16]FINLEY A O, BANERJEE S, CARLIN B P. spBayes: An R package for univariate and multivariate hierarchical point-referenced spatial models[J]. Journal of Statistical Software, 2007, 19(4):1-24.
[17]王柯人,舒清态,赵洪莹,等. 高山松单木地上生物量模型不确定性研究[J].西南林业大学学报(自然科学),2021,41(2):100-106.
[18]兰洁,肖中琪,李吉玫,等. 天山雪岭云杉生物量分配格局及异速生长模型[J].浙江农林大学学报,2020,37(3):416-423.
[19]王宁. 基于分层贝叶斯分析的城镇登记失业率估计方法[D].天津：天津财经大学,2012.
[20]覃朝勇. 基于层次贝叶斯方法的关键词广告转化率影响因素研究[J].统计与决策,2016(19):64-67.
[21]杨锦涛. 基于层次贝叶斯方法的滑坡易发性评价建模研究[D].成都：西南石油大学,2019.
[22]崔令军,张雄清,段爱国,等. 基于分层贝叶斯法的杉木人工林最大密度线[J].林业科学,2016,52(9):95-102.
[23]姚丹丹,雷相东,张则路. 基于贝叶斯法的长白落叶松林分优势高生长模型研究[J].北京林业大学学报,2015,37(3):94-100.
[24]张继巍,高文龙,秦天燕,等. OpenBUGS软件介绍及应用[J].中国卫生统计,2017,34(1):170-172,176.
[25]DONGSHENG C, XINGZHAO H, XIAOMEI S, et al. A comparison of hierarchical and non-hierarchical Bayesian approaches for fitting allometric larch (Larix.spp.) biomass equations[J]. Forests, 2016, 7(1):18.
[26]姚丹丹,徐奇刚,闫晓旺,等. 基于贝叶斯方法的蒙古栎林单木枯死模型[J].北京林业大学报,2019,41(9):1-8.
[27]黄兴召,王泽夫,徐小牛. 生物量转换因子连续函数的拟合方法比较[J].浙江农林大学报,2017,34(5):775-781.

备注/Memo

备注/Memo:
收稿日期：2022-02-15基金项目：国家自然科学基金项目（31860205、31460194）；国家重点研发计划项目（2018YFD0600100）；云南省教育厅科学研究基金项目（2021Y249）作者简介：黄金君(1998-)，女，广西南宁人，硕士研究生，主要从事资源与环境遥感研究。(E-mail)330061594@qq.com通讯作者：舒清态，(E-mail)shuqt@163.com
更新日期/Last Update: 2022-11-07