Uncertainty analysis of estimating aboveground biomass of Pinus densata by remote sensing
-
摘要:
目的 采用遥感数据估算森林地上生物量仍存在一些不确定性问题，研究估算过程中的误差来源及其占比，对提高森林地上生物量的估测精度具有重要意义。 方法 从遥感影像提取因子，结合高山松Pinus densata外业调查数据，建立多元线性回归、梯度提升回归树、随机森林等3种地上生物量估测模型，对样地尺度与3种模型的不确定性进行分析和度量。 结果 ①高山松单株生物量模型不确定性为16.43%，样地尺度的不确定性为7.07%；②多元线性回归模型残差不确定性为34.86%，参数不确定性为21.30%，与样地不确定性合成后总不确定性为41.45%；③非参数模型中，梯度提升回归树估测高山松地上生物量的总不确定性为23.12%，随机森林为19.42%。 结论 3种遥感估算模型中，非参数模型的不确定性明显低于参数模型。相较于样地尺寸，遥感估算模型的不确定性对地上生物量估算精度的影响较大。图3表3参26 Abstract:Objective To improve the estimation accuracy of forest aboveground biomass, this study is aimed to conduct an uncertainty analysis, trying to figure out the percentage error of estimating forest aboveground biomass by remote sensing and the causes behind. Method With factors extracted from remote sensing images and combined with the data of Pinus densata from field surveys, three types of aboveground biomass estimation model were established, namely Multiple Linear Regression (MLR), Gradient Boost Regression Tree (GBRT), and Random Forest (RF), before the uncertainty of sample plot scale and three models was measured and analyzed. Result (1) The uncertainty of tree biomass model for P. densata is 16.43%, and the uncertainty of the scale up to the sample plot is 7.07%; (2) The residual uncertainty of the MLR model is 34.86%, the parameter uncertainty is 21.30% whereas the total uncertainty combined with the sample plot uncertainty is 41.45%. (3) In the non-parametric model of the GBRT modeling estimates, the total uncertainty of the aboveground biomass is 23.12%, and the RF is 19.42%. Conclusion Among the three remote sensing models, the uncertainty of the non-parametric model is obviously lower than that of the parametric model. Compared with the uncertainty of the sample plot scale, the remote sensing estimation model has a great effect on the accuracy of the aboveground biomass estimation. [Ch, 3 fig. 3 tab. 26 ref.] -
Key words:
- Pinus densata /
- aboveground biomass /
- remote sensing model /
- uncertainty
-
图 1 基于多元线性回归的地上生物量预测效果和残差不确定性
Figure 1 Prediction effect and the residual uncertainty of aboveground biomass based on multiple linear regression
图 2 基于梯度提升回归树的地上生物量预测效果和残差不确定性
Figure 2 Prediction effect and the residual uncertainty of aboveground biomass based on gradient boost regression tree
图 3 基于随机森林的地上生物量预测效果和残差不确定性
Figure 3 Prediction effect and the residual uncertainty of aboveground biomass based on random forest
表 1 外业调查样地统计
Table 1. Basic statistics of the sample plots of field surveys
指标 平均胸径/cm 平均树高/m 林分密度/(株·hm−2) 最大值 22.76 14.18 3 100.00 最小值 6.85 4.48 222.22 平均值 14.80 8.70 1 181.30 标准差 3.68 2.11 714.14 
下载: 导出CSV
表 2 建模和检验样本的地上生物量实测值
Table 2. Measured values of aboveground biomass of modeling and testing samples
样本集 样地数/个 地上生物量/(t·hm−2) 最大值 最小值 平均值 标准差 建模样本 48 169.75 12.95 56.50 33.85 检验样本 12 94.97 15.44 62.56 26.48
下载: 导出CSV
表 3 3种地上生物量估测模型的不确定性
Table 3. Uncertainty results of the three aboveground biomass estimation models
模型 遥感估算地上生物量的不确定性 样地尺度地上生物量的不确定性 总不确定性/% 模型残差
变异/%模型参数
变异/%遥感估测模型
不确定性/%模型残差
变异/%模型参数
变异/%单株生物量模型
不确定性/%样地尺度不
确定性/%多元线性回归 34.86 21.30 40.84 10.76 12.42 16.43 7.07 41.45 梯度提升回归树 22.01 22.01 23.12 随机森林 18.09 18.09 19.42 说明：梯度提升回归树与随机森林为非参数模型，不考虑模型参数变异的不确定性
下载: 导出CSV
-
[1] 唐守正, 张会儒, 胥辉. 相容性生物量模型的建立及其估计方法研究[J]. 林业科学, 2000, 36(专刊1): 19 − 27. TANG Shouzheng, ZHANG Huiru, XU Hui. Study on establish and estimate method of compatible biomass model[J]. Sci Silv Sin, 2000, 36(spec 1): 19 − 27. [2] 王维枫, 雷渊才, 王雪峰, 等. 森林生物量模型综述[J]. 西北林学院学报, 2008, 23(2): 58 − 63. WANG Weifeng, LEI Yuancai, WANG Xuefeng, et al. A review of forest biomass models [J]. J Northwest For Univ, 2008, 23(2): 58 − 63. [3] 徐婷, 曹林, 佘光辉. 基于Landsat 8 OLI的特征变量优化提取及森林生物量反演[J]. 遥感技术与应用, 2015, 30(2): 226 − 234. XU Ting, CAO Lin, SHE Guanghui. Feature extraction and forest biomass estimation based on Landsat 8 OLI [J]. Remote Sensing Technol Appl, 2015, 30(2): 226 − 234. [4] MCROBERTS R E, WESTFALL J A. Effects of uncertainty in model predictions of individual tree volume on large area volume estimates [J]. For Sci, 2014, 60(1): 34 − 42. doi: 10.5849/forsci.12-141 [5] LU Dengsheng, CHEN Qi, WANG Guangxing, et al. A survey of remote sensing-based aboveground biomass estimation methods in forest ecosystems [J]. Int J Digital Earth, 2016, 9(1): 63 − 105. doi: 10.1080/17538947.2014.990526 [6] BAO Rui, ZHANG Jialong, LU Chi, et al. Estimating above-ground biomass of Pinus densata Mast. using best slope temporal segmentation and Landsat time series[J/OL]. J Appl Remote Sensing, 2021, 15(2)[2021-03-25]. doi: 10.1117/1.JRS.15.024507. [7] ZHAO Panpan, LU Dengsheng, WANG Guangxing, et al. Examining spectral reflectance saturation in landsat imagery and corresponding solutions to improve forest aboveground biomass estimation [J]. Remote Sensing, 2016, 8(6): 469 − 495. doi: 10.3390/rs8060469 [8] 张加龙, 胥辉, 陆驰. 应用Landsat 8 OLI和GBRT对高山松地上生物量的估测[J]. 东北林业大学学报, 2018, 46(8): 25 − 30. doi: 10.3969/j.issn.1000-5382.2018.08.005 ZHANG Jialong, XU Hui, LU Chi. Estimating above ground biomass of Pinus densata based on Landsat 8 OLI and gradient boost regression [J]. J Northeast For Univ, 2018, 46(8): 25 − 30. doi: 10.3969/j.issn.1000-5382.2018.08.005 [9] COHEN R, KAINO J, OKELLO J A, et al. Propagating uncertainty to estimates of above-ground biomass for Kenyan mangroves: a scaling procedure from tree to landscape level [J]. For Ecol Manage, 2013, 310: 968 − 982. doi: 10.1016/j.foreco.2013.09.047 [10] 秦立厚, 张茂震, 钟世红, 等. 森林生物量估算中模型不确定性分析[J]. 生态学报, 2017, 37(23): 7912 − 7919. QIN Lihou, ZHANG Maozhen, ZHONG Shihong, et al. Model uncertainty in forest biomass estimation [J]. Acta Ecol Sin, 2017, 37(23): 7912 − 7919. [11] SHETTLES M, TEMESGEN H, GRAY A N, et al. Comparison of uncertainty in per unit area estimates of aboveground biomass for two selected model sets [J]. For Ecol Manage, 2015, 354: 18 − 25. doi: 10.1016/j.foreco.2015.07.002 [12] CLARK D B, KELLNER J R. Tropical forest biomass estimation and the fallacy of misplaced concreteness [J]. J Veg Sci, 2012, 23(6): 1191 − 1196. doi: 10.1111/j.1654-1103.2012.01471.x [13] CHEN Qi, LAURIN G V, VALENTINI R. Uncertainty of remotely sensed aboveground biomass over an African tropical forest: propagating errors from trees to plots to pixels [J]. Remote Sensing Environ, 2015, 160: 134 − 143. doi: 10.1016/j.rse.2015.01.009 [14] 秦立厚. 浙江省森林碳储量估算不确定性研究[D]. 杭州: 浙江农林大学, 2017. QIN Lihou. Uncertainty of Forest Carbon Storage Estimation in Zhejiang Province[D]. Hangzhou: Zhejiang A&F University, 2017. [15] 孙雪莲. 基于Landsat 8-OLI的香格里拉高山松林生物量遥感估测模型研究[D]. 昆明: 西南林业大学, 2016. SUN Xuelian. Biomass Estimation Model of Pinus densata Forests in Shangri-La City Based on Landsat 8-OLI by Remote Sensing[D]. Kunming: Southwest Forestry University, 2016. [16] NICHOL J, HANG L K, SING W M. Empirical correction of low sun angle images in steeply sloping terrain: a slope-matching technique [J]. Int J Remote Sensing, 2006, 27(3): 629 − 635. doi: 10.1080/02781070500293414 [17] 陆驰. 基于Landsat的香格里拉市高山松地上生物量及其动态变化建模研究[D]. 昆明: 西南林业大学, 2017. LU Chi. The Estimation and Dynamic Modeling on the Aboveground Biomass of Pinus densata in Shangri-La Based on Landsat[D]. Kunming: Southwest Forestry University, 2017. [18] HARALICK R M, SHANMUGAM K, DINSTEIN I. Textural features for image classification [J]. Stud Media Commun, 1973, 3(6): 610 − 621. [19] 岳彩荣. 香格里拉县森林生物量遥感估测研究[D]. 北京: 北京林业大学, 2011. YUE Cairong. Forest Biomass Estimation in Shangri-La County Based on Remote Sensing[D]. Beijing: Beijing Forestry University, 2011. [20] 张加龙, 胥辉. 基于遥感的高山松连清固定样地地上生物量估测模型构建[J]. 北京林业大学学报, 2020, 42(7): 1 − 11. doi: 10.12171/j.1000-1522.20190394 ZHANG Jialong, XU Hui. Establishment of remote sensing based model to estimate the aboveground biomass of Pinus densata for permanent sample plots from national forestry inventory [J]. J Beijing For Univ, 2020, 42(7): 1 − 11. doi: 10.12171/j.1000-1522.20190394 [21] FRIEDMAN J H. Greedy function approximation: a gradient boosting machine [J]. Ann Stat, 2001, 29(5): 1189 − 1232. doi: 10.1214/aos/1013203450 [22] OU Guanglong, LÜ Yanyu, XU Hui, et al. Improving forest aboveground biomass estimation of Pinus densata forest in Yunnan of southwest China by spatial regression using Landsat 8 images [J]. Remote Sensing, 2019, 11(23): 738 − 762. [23] HOSMER D W, LEMESHOW S. Applied Logistic Regression [M]. New York: John Wiley & Sons, 2000. [24] 李金海. 误差理论与测量不确定度评定[M]. 北京: 中国计量出版社, 2003. LI Jinhai. Error Theory and Measurement Uncertainty Assessment[M]. Beijing: China Metrology Publishing House, 2003. [25] 陈蜀蓉, 张超, 郑超超, 等. 公益林生物量估算方法研究——以浙江省缙云县公益林为例[J]. 浙江林业科技, 2015, 35(5): 20 − 28. doi: 10.3969/j.issn.1001-3776.2015.05.005 CHEN Shurong, ZHANG Chao, ZHENG Chaochao, et al. Estimation methods for biomass of ecological forest in Jinyun [J]. J Zhejiang For Sci Technol, 2015, 35(5): 20 − 28. doi: 10.3969/j.issn.1001-3776.2015.05.005 [26] CHAVE J, CONDIT R, AGUILAR S, et al. Error propagation and scaling for tropical forest biomass estimates [J]. Philos Trans R Soc Lond Series B Biol Sci, 2004, 359(1443): 409 − 420. doi: 10.1098/rstb.2003.1425 -
-
链接本文:
https://zlxb.zafu.edu.cn/article/doi/10.11833/j.issn.2095-0756.20210473
点击查看大图
计量
- 文章访问数: 39
- 被引次数: 0
