欢迎访问林业科学,今天是

林业科学 ›› 2021, Vol. 57 ›› Issue (3): 67-78.doi: 10.11707/j.1001-7488.20210307

• 论文与研究报告 • 上一篇    下一篇

基于3-PG模型的长白落叶松生物量生长预测

夏晓运1,2,庞勇2,3,*,黄庆丰1,吴荣4,陈东升5,白羽2,3   

  1. 1. 安徽农业大学林学与园林学院 合肥 230036
    2. 中国林业科学研究院资源信息研究所 北京 100091
    3. 国家林业和草原局林业遥感与信息技术重点实验室 北京 100091
    4. 西南林业大学林学院 昆明 650224
    5. 中国林业科学研究院林业研究所 北京 100091
  • 收稿日期:2019-03-01 出版日期:2021-03-25 发布日期:2021-04-07
  • 通讯作者: 庞勇
  • 基金资助:
    十三五国家重点研发项目(2017YFD0600404)

Prediction of Biomass Growth of Larix olgensis Based on 3-PG Model

Xiaoyun Xia1,2,Yong Pang2,3,*,Qingfeng Huang1,Rong Wu4,Dongsheng Chen5,Yu Bai2,3   

  1. 1. School of Forestry and Landscape, Anhui Agricultural University Hefei 230036
    2. Research Institute of Forest Resource and Information Techniques, CAF Beijing 100091
    3. Key Laboratory of Forestry Remote Sensing and Information System, National Forestry and Grassland Administration Beijing 100091
    4. Faculty of Forestry, Southwest Forestry University Kunming 650224
    5. Research Institute of Forestry, CAF Beijing 100091
  • Received:2019-03-01 Online:2021-03-25 Published:2021-04-07
  • Contact: Yong Pang

摘要:

目的: 基于3-PG模型预测长白落叶松生物量生长变化,为长白落叶松林分生长规律研究提供依据。方法: 以5块长白落叶松密度试验林连续28年监测数据和24块长白落叶松固定样地3期调查数据为基础,结合各组分(叶、干和根)生物量计算公式,获得每块样地不同调查时间的密度、胸径、蓄积和各组分生物量。根据密度试验林数据校正模型生理参数,结合立地参数和气象参数,通过参数率定、迭代拟合与敏感性分析方法确定长白落叶松3-PG模型的生理参数。采用决定系数(R2)、平均误差(ME)、平均绝对误差(MAE)、平均相对误差(MRE)和均方根误差(RMSE)评价模型预测能力。选取冠层量子效率(alpha)和初级生物量分配到根的最小值(pRn)进行敏感性分析,并预测肥力等级(FR)为0.2、0.4和0.6时长白落叶松生物量生长变化趋势。结果: 1)3-PG模型预测值与实测值之间R2在0.77以上;除叶干生物量比为25.6%外,其他各指标的MRE绝对值均在10.97%以内,预测结果较可靠;2)alpha和pRn具有较高敏感性,是模型的关键参数;3)模型预测不同FR下的长白落叶松生物量变化符合树木生长机理过程,且各组分生物量随FR增加而增加。结论: 基于地面数据的参数率定后,3-PG模型能够很好模拟长白落叶松生物量生长变化,可作为一种有效的森林经营预测工具。对于长白落叶松3-PG模型,冠层量子效率(alpha)和初级生物量分配到根的最小值(pRn)是影响预测结果的关键参数。

关键词: 长白落叶松, 3-PG模型, 敏感性分析, 生物量

Abstract:

Objective: This study was executed to predict the biomass growth of Larix olgensis based on 3-PG model in order to provide the basis for studying its growth rules. Method: The data used in this article was obtained from 5 plots of experimental forests continuously observed for 28 years and 24 plots with 3 re-measurements. The stem density, DBH(diameter at breast height), volume and biomass of each plot in different times were calculated using the biomass calculation formula for each component (leaf, stem and root). The physiological parameters of the 3-PG model of L. olgensis were calibrated by density data. And the parameter values were determined based on soil data and meteorological data through parameter calibration, iterative fitting and sensitivity analysis. The result accuracy was examined by calculating coefficient of determination (R2), mean error(ME), mean absolute error(MAE), mean relative error(MRE), and root mean square error(RMSE). Two factors, the canopy quantum efficiency(alpha) and minimum biomass fraction of NPP to roots (pRn), were chosen to procced sensitivity analysis. Then, the growth biomass of L. olgensis was predicted under the conditions of the fertility ratings(FR) being 0.2, 0.4 and 0.6. Result: 1) The prediction results were reliable and the coefficient of determination (R2) was above 0.77. The absolute values of MRE of all the other indicators were within 10.97%, except for the foliage to stem biomass ratio with a value of 25.6%. 2) The sensitivity analysis showed that alpha and pRn were the key parameters of the model with high sensitivities. 3) The predicted biomass growth in this study was consistent with the growth mechanisms of plants under different FRs. At the same time, the biomass of larch was increased with the growth of FR. Conclusion: After the parameter calibration based on field data, the 3-PG model could simulate the biomass growth of L. olgensis well and be used as an effective forest management tool. For the 3-PG model of L. olgensis, alpha and pRn might be the key parameters affecting the prediction results.

Key words: Larix olgensis, 3-PG model, sensitivity analysis, biomass

中图分类号: