不同矫正位置对落叶松分位数削度方程预测精度的影响

doi:10.12302/j.issn.1000-2006.201908036

国家林草科技领军期刊
中国精品科技期刊
中国高校百佳科技期刊
江苏省新闻出版政府奖期刊奖
RCCSE林学权威期刊（A+）
CSCD核心期刊
Scopus数据库收录期刊
中文核心期刊
SCD核心期刊

作者加群：102861116

微信公众号：南京林业大学学报

高级检索

南京林业大学学报（自然科学版） ›› 2021, Vol. 45 ›› Issue (1): 182-188.doi: 10.12302/j.issn.1000-2006.201908036

不同矫正位置对落叶松分位数削度方程预测精度的影响

辛士冬(), 何培, 姜立春^*()

东北林业大学林学院,森林生态系统可持续经营教育部重点实验室,黑龙江哈尔滨 150040

收稿日期:2019-08-28 接受日期:2019-12-05 出版日期:2021-01-30 发布日期:2021-02-01
通讯作者: 姜立春
基金资助:
国家自然科学基金项目(31570624);黑龙江省应用技术研究与开发计划项目(GA19C006);中央高校基本科研业务费专项资金项目(2572019CP15)

Effects of different calibration positions on prediction precision of quantile taper function for Larix gmelinii

XIN Shidong(), HEI Pei, JIANG Lichun^*()

Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, School of Forestry, Northeast Forestry University, Harbin 150040, China

Received:2019-08-28 Accepted:2019-12-05 Online:2021-01-30 Published:2021-02-01
Contact: JIANG Lichun

摘要/Abstract

摘要：

【目的】基于非线性分位数回归方法构建大兴安岭落叶松(Larix gmelinii)树干削度方程,并分析比较基本模型与不同分位数(τ=0.1、0.2、0.3、0.4、0.5、0.6、0.7、0.8、0.9)模型,利用树干不同高度的上部直径进行矫正分位数组合模型预测精度,为落叶松天然林干形的精准预测提供理论依据。【方法】以大兴安岭壮志林场212株落叶松树干干形数据为研究对象,基于非线性分位数回归方法和Max and Burkhart分段削度方程,利用SAS软件中NLP过程拟合各分位数分段削度方程,把树干相对高20%、30%、40%、50%、60%、70%处的直径以及胸径到树尖的中间位置(50%^*)的树干上部直径引入到分段削度方程中进行矫正,并以平均误差(MAB)和相对误差(MPB)为评价指标对削度方程进行对比分析。【结果】Max-Burkhart分段削度方程在9个不同的分位点都可以得到参数估计值,因此分位数回归削度模型可以评价在不同分位数的预测能力。未矫正的分位数(τ=0.5、0.6)模型的预测精度略优于基本模型。准确地选择矫正位置至关重要,与未矫正的基本模型相比,利用树干相对高20%和70%处的直径进行矫正不能提高各分位数组合模型的预测精度,利用树干相对高30%、40%、50%、60%处的直径以及胸径到树尖中间位置的树干上部直径进行矫正的大多数分位数组合(3、5、7、9个分位数组合)模型的预测精度都能得到提高,总体使用矫正位置分位数组合模型的预测精度顺序为40%>50%^*>50%> 60%>30%>20%>70%。最佳的矫正位置为树干相对高40%处,并以3个分位数的组合(τ=0.3、0.5、0.7)模型预测精度最高,与未矫正的基本模型相比,MAB和MPB均下降13.5%。【结论】在削度方程中引入一个合理的矫正位置可以提高模型的预测精度,其中,最佳矫正位置为树干相对高40%处,最优模型为3个分位数组合(τ=0.3、0.5、0.7)模型。在实际应用中,如果不考虑矫正时,建议采用分位数τ=0.5削度方程的参数估计值。

关键词: 落叶松, 组合分位数, 矫正位置, 削度方程, 预测精度

Abstract:

【Objective】 The aim of this study was to develop a stem taper equation based on the quantile regression method for Larix gmelinii in Greater Khingan Mountains. We compared and analyzed the prediction precision of basic models, different quantile (τ=0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9) models, and quantile group models using different upper-stem diameters. These studies provide a theoretical basis for the precise prediction of the natural forest taper of L. gmelinii. 【Method】 The stem taper data of 212 L. gmelinii in Greater Khingan Mountains, Songling Forest Bureau, were the research object. Based on the nonlinear quantile regression method and Max and Burkhart segmented taper equation, the nonlinear programming (NLP) method in SAS software was used to fit the stem taper equation of different quantiles. Calibration of the taper equations was carried out using the upper-stem diameter measured at the relative heights of 20%, 30%, 40%, 50%, 60%, 70%, and the midpoint between the breast height and the tip of the tree (50%^*). The mean absolute bias (MAB) and mean percentage of bias (MPB) were used as statistical criteria to compare the calibrated stem taper equations. 【Result】 The parameters of the Max and Burkhart segmented taper equations based on nine different quantiles were obtained, allowing the prediction ability of the quantile regression taper model in different quantiles to be evaluated. The prediction accuracy of the uncalibrated quantile models at the quantile (τ=0.5, 0.6) was slightly better than the uncalibrated basic model. Accurate choices of the localized position were essential. Compared with the uncalibrated basic model, the prediction accuracy of each quantile group model for the calibrated stem taper using the upper-stem diameter measured at relative heights of 20% and 70% was not improved. The prediction accuracy of the most quantile group (3, 5, 7, 9) models calibrated using the upper stem measurements at relative heights of 30%, 40%, 50%, 60% and the midpoint between the breast height and the tip of the tree was improved. The overall prediction accuracy using a quantile group model for the calibrated positions was 40%>50%^*>50%>60%>30%>20%>70%. The best calibration position was the relative height of 40% of the tree, and the three quantile (τ=0.3, 0.5, 0.7) group model had the highest prediction accuracy. Both MAB and MPB decreased by 13.5% compared to the uncalibrated basic model. 【Conclusion】 Introducing a reasonable calibration position in the taper equation improved the prediction accuracy of the models. The best calibration position was the relative height of 40% of the tree with the group of three quantiles (τ =0.3, 0.5, 0.7). In practical applications, if the calibration is not considered, it is recommended that the quantile (τ =0.5) should be used.

Key words: Larix gmelinii, combined quantile, calibration position, taper equation, prediction precision

中图分类号:

S758.2

辛士冬,何培,姜立春. 不同矫正位置对落叶松分位数削度方程预测精度的影响[J]. 南京林业大学学报（自然科学版）, 2021, 45(1): 182-188.

XIN Shidong, HEI Pei, JIANG Lichun. Effects of different calibration positions on prediction precision of quantile taper function for Larix gmelinii[J].Journal of Nanjing Forestry University (Natural Science Edition), 2021, 45(1): 182-188.DOI: 10.12302/j.issn.1000-2006.201908036.

图/表 5

表1

表2

图1

表3

表4

基于不同矫正位置落叶松削度方程的统计量"

组合模型 group models	20%		30%		40%		50%		60%		70%		50%^*
组合模型 group models	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB	平均误差 MAB	相对误差 MPB
τ₃₁	1.409 1	6.227 0	1.250 5	5.526 1	1.168 9	5.165 7	1.194 0	5.276 7	1.228 6	5.429 4	1.456 7	6.437 5	1.168 1	5.162 1
τ₃₂	1.388 5	6.136 2	1.265 2	5.591 0	1.185 1	5.237 1	1.212 4	5.357 9	1.266 1	5.595 3	1.490 9	6.588 4	1.186 2	5.242 0
τ₃₃	1.332 5	5.888 5	1.224 4	5.410 7	1.154 6	5.102 4	1.181 2	5.219 8	1.214 2	5.365 7	1.436 6	6.348 4	1.157 9	5.116 8
τ₃₄	1.503 5	6.644 3	1.382 1	6.107 6	1.305 0	5.767 2	1.302 2	5.754 8	1.358 2	6.001 9	1.561 2	6.899 2	1.292 3	5.711 0
τ₅₁	1.412 4	6.241 7	1.254 4	5.543 4	1.167 7	5.160 4	1.193 4	5.274 0	1.229 1	5.431 5	1.460 2	6.453 1	1.168 2	5.162 5
τ₅₂	1.412 4	6.241 4	1.254 1	5.542 1	1.169 6	5.168 5	1.192 9	5.271 8	1.228 6	5.429 6	1.458 1	6.443 5	1.168 5	5.163 6
τ₅₃	1.410 1	6.231 4	1.250 8	5.527 4	1.169 1	5.166 4	1.193 4	5.273 7	1.227 9	5.426 1	1.455 4	6.431 6	1.167 4	5.158 9
τ₅₄	1.333 6	5.893 3	1.225 7	5.416 8	1.155 1	5.104 6	1.181 1	5.219 4	1.214 1	5.365 5	1.433 7	6.336 0	1.158 2	5.118 4
τ₅₅	1.389 3	6.139 3	1.265 7	5.593 4	1.185 6	5.239 4	1.212 2	5.356 8	1.265 2	5.591 2	1.489 1	6.580 6	1.185 3	5.238 0
τ₅₆	1.389 4	6.139 9	1.266 0	5.594 9	1.185 2	5.237 4	1.212 5	5.358 4	1.265 2	5.591 1	1.487 9	6.575 1	1.185 3	5.238 2
τ₇₁	1.413 1	6.244 8	1.255 0	5.545 8	1.168 3	5.162 7	1.193 2	5.272 9	1.228 2	5.427 5	1.458 5	6.445 3	1.167 3	5.158 5
τ₇₂	1.413 4	6.246 2	1.255 5	5.548 2	1.170 1	5.170 6	1.192 9	5.271 5	1.228 6	5.429 3	1.455 3	6.431 0	1.168 8	5.165 1
τ₇₃	1.390 3	6.144 2	1.267 1	5.599 5	1.186 1	5.241 6	1.212 1	5.356 4	1.265 2	5.591 0	1.486 3	6.568 1	1.185 6	5.239 5
τ₇₄	1.413 2	6.245 3	1.255 3	5.547 3	1.167 8	5.160 7	1.193 6	5.274 5	1.228 1	5.427 4	1.457 2	6.439 7	1.167 4	5.158 8
τ₉₁	1.414 2	6.249 6	1.256 3	5.551 9	1.168 8	5.164 9	1.193 1	5.272 6	1.228 1	5.427 2	1.455 7	6.432 8	1.167 7	5.160 1

表4

参考文献 27

[1]	孟宪宇. 削度方程和出材率表的研究[J]. 南京林业大学学报(自然科学版), 1982,6(1):122-133.
	MENG X Y. Studies of taper equations and the table of merchantable volumes[J]. J Nanjing For Univ (Nat Sci Ed), 1982,6(1):122-133.
[2]	JIANG L C, BROOKS J R, WANG J X. Compatible taper and volume equations for yellow-poplar in West Virginia[J]. Forest Ecology and Management, 2005,213(1/2/3):399-409. DOI: 10.1016/J.FORECO.2005.04.006. doi: 10.1016/j.foreco.2005.04.006
[3]	JIANG L C, BROOKS J R. Taper, volume, and weight equations for red pine in West Virginia[J]. Northern Journal of Applied Forestry, 2008,25(3):151-153. DOI: 10.1093/NJAF/25.3.151. doi: 10.1093/njaf/25.3.151
[4]	曾伟生, 廖志云. 削度方程的研究[J]. 林业科学, 1997,33(2):127-132.
	ZENG W S, LIAO Z Y. Research of the taper equation[J]. Sci Silvae Sin, 1997,33(2):127-132.
[5]	CAO Q V, WANG J. Calibrating fixed-and mixed-effects taper equations[J]. Forest Ecology and Management, 2011,262(4):671-673. DOI: 10.1016/J.FORECO.2011.04.039. doi: 10.1016/j.foreco.2011.04.039
[6]	SABATIA C O. Use of upper stem diameters in a polynomial taper equation for New Zealand radiata pine: an evaluation[J]. New Zealand Journal of Forestry Science, 2016,46(1):14. DOI: 10.1186/S40490-016-0070-2. doi: 10.1186/s40490-016-0070-2
[7]	CZAPLEWSKI R L, MCCLURE J P. Conditioning a segmented stem profile model for two diameter measurements[J]. Forest Science, 1988,34(2):512-522.
[8]	CAO Q V. Calibrating a segmented taper equation with two diameter measurements[J]. Southern Journal of Applied Forestry, 2009,33(2):58-61. DOI: 10.1093/SJAF/33.2.58. doi: 10.1093/sjaf/33.2.58
[9]	KOZAK A. Effects of upper stem measurements on the predictive ability of a variable-exponent taper equation[J]. Canadian Journal of Forest Research, 1998,28(7):1078-1083. DOI: 10.1139/X98-120. doi: 10.1139/x98-120
[10]	GÓMEZ-GARCÍA E, CRECENTE-CAMPO F, DIÉGUEZ-ARANDA U. Selection of mixed-effects parameters in a variable-exponent taper equation for birch trees in northwestern Spain[J]. Annals of Forest Science, 2013,70(7):707-715. DOI: 10.1007/S13595-013-0313-9. doi: 10.1007/s13595-013-0313-9
[11]	SHARMA M, PARTON J. Modeling stand density effects on taper for jack pine and black spruce plantations using dimensional analysis[J]. Forest Science, 2009,55(3):268-282. DOI: 10.1093/FORESTSCIENCE/55.3.268.
[12]	KOENKER R, BASSETT G. Regression quantiles[J]. Econometrica, 1978,46(1):33-50. doi: 10.2307/1913643
[13]	KOENKER R, HALLOCK K F. Quantile regression[J]. Journal of Economic Perspectives, 2015,15(4):143-156. doi: 10.1257/jep.15.4.143
[14]	张期奇, 董希斌, 张甜, 等. 抚育间伐强度对兴安落叶松中龄林测树因子的影响[J]. 森林工程, 2018,34(5):1-7.
	ZHANG Q Q, DONG X B, ZHANG T, et al. The Effects of Different thinning Intensities on tree-measurement Factors of middle-aged Larch gmelinii Seedlings[J]. Forest Engineering, 2018,34(5):1-7.
[15]	SHARMA M, BURKHART H E. Selecting a level of conditioning for the segmented polynomial taper equation[J]. Forest Science, 2003,49(2):324-330. DOI: 10.1093/FORESTSCIENCE/49.2.324.
[16]	COBLE D W, HILPP K. Compatible cubic-foot stem volume and upper-stem diameter equations for semi-intensive plantation grown loblolly pipe trees in East Texas[J]. Southern Journal of Applied Forestry, 2006,30(3):132-141. DOI: 10.1093/SJAF/30.3.132. doi: 10.1093/sjaf/30.3.132
[17]	TRINCADO G, BURKHART H E. A generalized approach for modeling and localizing stem profile curves[J]. Forest Science, 2006,52(6):670-682. DOI: 10.1093/FORESTSCIENCE/52.6.670.
[18]	OZCELIK R, BROOKS J R, JIANG L C. Modeling stem profile of Lebanon cedar, Brutian pine, and Cilicica fir in Southern Turkey using nonlinear mixed-effects models[J]. European Journal of Forest Research, 2011,130(4):613-621. DOI: 10.1007/S10342-010-0453-5. doi: 10.1007/s10342-010-0453-5
[19]	KOZAK A. My last words on taper equations[J]. Forestry Chronicle, 2004,80(4):507-515. DOI: 10.5558/TFC80507-4. doi: 10.5558/tfc80507-4
[20]	DOYOG N D, LEE Y J, LEE S J, et al. Compatible taper and stem volume equations for Larix kaempferi (Japanese larch) species of South Korea[J]. Journal of Mountain Science, 2017,14(7):1341-1349. DOI: 10.1007/S11629-016-4291-X. doi: 10.1007/s11629-016-4291-x
[21]	MAX T A, BURKHART H E. Segmented polynomial regression applied to taper equations[J]. Forest Science, 1976,22(3):283-289. DOI: 10.1093/FORESTSCIENCE/22.3.283.
[22]	KOENKER R, BASSETT G. Robust tests for heteroscedasticity based on regression quantiles[J]. Econometrica, 1982,50(1):43-61. DOI: 10.2307/1912528. doi: 10.2307/1912528
[23]	CADE B S, NOON B R. A gentle introduction to quantile regression for ecologists[J]. Frontiers in Ecology & the Environment, 2003,1(8):412-420. DOI: 10.1890/1540-9295(2003)001[0412:AGITQR]2.0.CO;2. doi: 10.1890/1540-9295(2003)001[0412:AGITQR]2.0.CO;2
[24]	辛士冬, 姜立春. 利用分位数回归模拟人工樟子松树干干形[J]. 北京林业大学学报, 2020,42(2):1-8.
	XIN S D, JIANG L C. Modeling stem taper profile for Pinus sylvestris plantations using nonlinear quantile regression[J]. Journal of Beijing Forestry University, 2020,42(2):1-8. DOI: 10.12171/j.1000-1522.20190014.
[25]	SAS Institute Inc. SAS/OR 9.22 User ‘s guide: mathematical programming[M]. Cary, NC: SAS Institute Inc, 2010.
[26]	CAO Q V, WANG J. Evaluation of methods for calibrating a tree taper equation[J]. Forest Science, 2015,61(2):213-219. DOI: 10.5849/FORSCI.14-008. doi: 10.5849/forsci.14-008
[27]	RODRIGUEZ F, LIZARRALDE I, BRAVO F. Comparison of stem taper equations for eight major tree species in the Spanish Plateau[J]. Forest Systems, 2015,24(3). DOI: 10.5424/FS/2015243-06229.

统计量 statistics	建模数据fitting data		检验数据validation data
统计量 statistics	直径/cm diameter	树高/m tree height	直径/cm diameter	树高/m tree height
平均值 mean	29.1	19.7	28.5	19.2
最小值 min.	5.5	7.1	6.2	7.4
最大值 max.	55.1	29.2	53.2	29.0
标准差 std.	14.1	5.4	13.5	5.3

回归方法 regression methods	分位点 quantiles	a₁	a₂	a₃	a₄	b₁	b₂
非线性回归nonlinear regression	—	-4.830 7	2.387 7	-2.344 5	267.115 0	0.789 2	0.057 6
分位数回归quantile regression	τ=0.1	-5.831 9	2.995 1	-2.670 4	174.531 9	0.862 9	0.064 8
	τ=0.2	-4.456 5	2.238 1	-2.118 5	172.336 9	0.797 2	0.068 1
	τ=0.3	-4.805 8	2.414 2	-2.391 7	174.082 7	0.791 3	0.069 1
	τ=0.4	-6.149 5	3.109 9	-3.086 8	167.964 4	0.829 5	0.070 8
	τ=0.5	-5.447 7	2.703 9	-2.749 4	174.551 1	0.812 4	0.071 0
	τ=0.6	-6.544 5	3.269 2	-3.317 9	172.806 8	0.837 0	0.072 9
	τ=0.7	-7.516 4	3.765 5	-3.839 1	172.523 2	0.850 5	0.075 4
	τ=0.8	-5.471 0	2.627 2	-2.779 5	174.657 2	0.815 0	0.078 4
	τ=0.9	-5.271 5	2.462 3	-2.725 8	174.718 1	0.808 0	0.080 5

模型 models	分位点 quantiles	平均误差 MAB	相对误差 MPB
基本模型 base model	—	1.335 1	5.900 1
各分位数模型 individual quantile models	τ=0.1	2.451 2	10.832 1
	τ=0.2	1.834 2	8.105 5
	τ=0.3	1.552 6	6.861 1
	τ=0.4	1.397 7	6.176 6
	τ=0.5	1.322 4	5.843 9
	τ=0.6	1.326 3	5.861 1
	τ=0.7	1.427 0	6.306 3
	τ=0.8	1.706 2	7.540 1
	τ=0.9	2.213 2	9.780 4

不同矫正位置对落叶松分位数削度方程预测精度的影响

Effects of different calibration positions on prediction precision of quantile taper function for Larix gmelinii

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 27

相关文章 15

编辑推荐

Metrics

本文评价

作者

读者

审稿

[1]	赵金满, 韩馨悦, 程瑞明, 张志东. 塞罕坝自然保护区华北落叶松和樟子松人工林健康评价[J]. 南京林业大学学报（自然科学版）, 2024, 48(3): 199-206.
[2]	韩新宇, 高露双, 秦莉, 庞荣荣, 刘鸣谦, 朱一泓, 田益雨, 张金. 林分密度对兴安落叶松径向生长-气候关系的影响[J]. 南京林业大学学报（自然科学版）, 2024, 48(2): 182-190.
[3]	左壮, 张韫, 崔晓阳. 火烧对兴安落叶松林土壤氮形态和含量的初期影响[J]. 南京林业大学学报（自然科学版）, 2024, 48(1): 147-154.
[4]	高谢雨, 董利虎, 郝元朔. 基于TLS的抚育间伐对长白落叶松干形的影响[J]. 南京林业大学学报（自然科学版）, 2023, 47(6): 85-94.
[5]	杨乐, 黄晓君, 包玉海, 包刚, 佟斯琴, 苏都毕力格. 无人机航高对落叶松毛虫虫害遥感监测精度的影响[J]. 南京林业大学学报（自然科学版）, 2023, 47(4): 13-22.
[6]	王云霓, 曹恭祥, 徐丽宏, 陈胜楠. 内蒙古大青山华北落叶松人工林蒸散特征及其影响因子[J]. 南京林业大学学报（自然科学版）, 2023, 47(4): 148-156.
[7]	路文燕, 董灵波, 田园, 汪莎杉, 曲宣怡, 魏巍, 刘兆刚. 基于树种组成的大兴安岭天然林主要树种树高-胸径曲线研究[J]. 南京林业大学学报（自然科学版）, 2023, 47(4): 157-165.
[8]	孙宇, 李凤日, 谢龙飞, 董利虎. 基于林分及地形因子的落叶松人工林林分生物量模型构建[J]. 南京林业大学学报（自然科学版）, 2023, 47(3): 129-136.
[9]	唐依人, 贾炜玮, 王帆, 孙毓蔓, 张颖. 基于TLS辅助的长白落叶松一级枝条生物量模型构建[J]. 南京林业大学学报（自然科学版）, 2023, 47(2): 130-140.
[10]	宋磊, 金星姬, PUKKALA Timo, 李凤日. 长白落叶松人工林多目标经营模式研究[J]. 南京林业大学学报（自然科学版）, 2023, 47(2): 150-158.
[11]	高羽, 李静, 刘洋, 乌雅瀚, 巩家星, 辛启睿. 结构方程模型在兴安落叶松林生长中的应用[J]. 南京林业大学学报（自然科学版）, 2023, 47(1): 38-46.
[12]	赵凯歌, 周正虎, 金鹰, 王传宽. 长期氮添加对落叶松和水曲柳人工林土壤碳、氮、磷含量和胞外酶活性的影响[J]. 南京林业大学学报（自然科学版）, 2022, 46(5): 177-184.
[13]	聂璐毅, 董利虎, 李凤日, 苗铮, 谢龙飞. 基于两水平非线性混合效应模型的长白落叶松削度方程构建[J]. 南京林业大学学报（自然科学版）, 2022, 46(3): 194-202.
[14]	谢立红, 黄庆阳, 曹宏杰, 杨帆, 王继丰, 倪红伟. 气候变暖对黑龙江省五大连池兴安落叶松径向生长的影响[J]. 南京林业大学学报（自然科学版）, 2022, 46(2): 150-158.
[15]	辛士冬, 姜立春, 穆林. 黑龙江省红松人工林林分乔木层可加性碳储量模型[J]. 南京林业大学学报（自然科学版）, 2022, 46(1): 115-121.