时滞和平衡含水率直接估计法的有效性分析

通过对不同直径落叶松枯枝含水率和环境条件的室内连续观测,分别使用Nelson模型和Simard模型作为平衡含水率响应模型,估计可燃物时滞和平衡含水率响应函数,然后以参数估计值预测可燃物含水率,分析比较建模样本数和平衡含水率模型不同对参数估计和含水率预测误差的影响。结果表明:1)基于Nelson模型的直接估计法在建模样本数较大时(84个),结果稳健,预测误差小,方法是有效的。2)采用Simard模型直接估计可燃物时滞和平衡含水率时,在建模样本数较少时,其预测效果不如Nelson模型,但当建模样本数较多(超过84)时,2个模型预测效果没有显著差别。     

樟子松针叶床层结构对失水过程中含水率参数的影响

以樟子松针叶床层为例,初步研究可燃物结构床层结构特征对其失水过程中时滞和平衡含水率参数(基于Nelson模型)的影响。通过对15个不同厚度、载量、密度的樟子松可燃物床层的失水过程的分析,估计这些床层的时滞和平衡含水率参数,利用这些参数进行可燃物含水率模拟的平均绝对误差和均方根误差均不超过0.01,所估计的可燃物含水率参数有效。对可燃物床层结构特征对这些参数的影响的研究表明,可燃物床层厚度和载量对时滞和平衡含水率的2个参数a、b具有有显著影响,与时滞和平衡含水率参数b正相关,与平衡含水率参数a负相关,据此建立可燃物含水率参数的预测模型。该模型高估了时滞和平衡含水率参数b,低估了平衡含水率参数a。所得结果还有一些不确定性,特别是可燃物结构特征对平衡含水率参数的影响,需要在更宽的温湿度范围内、针对其他平衡含水率模型和在失水吸水2个过程上进一步研究。     

气象法预测盘古林场可燃物含水率的外推精度

在春防期和秋防期对东北地区大兴安岭盘古林场的3种典型林分:樟子松林、兴安落叶松林、白桦林下的地表细小可燃物含水率进行长时间测定。得到其预测模型,使用模型对大兴安岭现有的4种可燃物含水率预测模型进行外推分析。结果表明:这些模型外推时得到的平均绝对误差(MAE)是自建模型的1.6~7.2倍。以1.5倍左右作为可替代的标准,则这4个模型与自建模型应具有一定的可替代性,即可用外来模型预测本地可燃物的含水率。春防期的模型预测精度高于秋防期。从外推精度的可用性来看,4个模型外推误差在绝对值上最小为3.6%,最大为17.5%。因此,现在的这4种可燃物含水率预测模型的外推能力都不理想。模型外推误差与建模地区和外推地区的微环境差异有关,与距离不是完全成正比。     

森林可燃物含水率测试仪研制

森林可燃物(主要指枯枝落叶层、易燃杂草、针叶、直径小于一厘米的枯枝等)含水率是计算林火潜在能量、火强度、林火蔓延、火险等级的划分、林火预报及计划烧除等的重要指标。多少年来,森林可燃物含水率的确定,一直沿用称重法和简单的手感法。这些方法在测定过程中,不仅测试麻烦,而且耗费时间和人力;特别是手感法误差极大。为了解决这个矛盾,我们研制成功电子式森林可燃物含水率测试仪。在此项研制中,对微机小系统的应用做了开发性的尝试。实践     

季节和降雨对细小可燃物含水率预测模型精度的影响

观察大兴安岭盘古地区的典型林分樟子松、白桦、兴安落叶松林的地表细小死可燃物含水率随不同季节变化的动态变化,应用气象要素回归法,分别使用无降雨数据、无降雨和有降雨混合数据及降雨数据,分春季、秋季和混合季节,建立了该地区森林地表可燃物含水率的统计预测模型,并研究了季节以及降雨对该类模型精度的影响。结果表明:季节和降雨对模型精度具有显著的影响,对于3种林型整体而言,混合模型的误差最大,可高达30%以上;秋季误差小于混合模型,大于春季预测模型;春季含水率预测模型精度最高,误差小于10%。无降雨模型预测效果最好,模型误差控制在3%以内,有降雨时段误差也可超过30%。如果采用区分季节和降雨时段建立可燃物含水率预测模型,据此做出的森林火险等级预报不会产生实质的影响,有助于提高火险等级预报的准确性。     

直接估计法在帽儿山林场白桦林可燃物含水率的适用性分析

在春季防火期内按林型内凋落物的不同空间位置的分类标准采集白桦林的地表死可燃物,并依此放入由小到大的铁丝网中,在室内进行试验,记录数据,对两种林型分别采用Nelson模型和Simard模型作为平衡含水率对温湿度响应模型,使用直接估计法对地表死可燃物的含水率结果进行了预测。结果表明:(1)基于Nelson模型的可燃物含水率预测误差往往小于基于Simard模型,一般来说基于Nelson模型的含水率预测效果要好于使用Simard模型预测含水率效果。但是,对于白桦林的腐殖质和混合可燃物则两者相差不大,预测效果都比较好,误差要求在3%以内都可以使用;(2)对于白桦林内凋落物及半腐殖质,误差要求在3%以内时,使用Nelson模型预测可燃物含水率的预测精度要更好;(3)白桦林内不同层可燃物含水率的预测精度由高到低依次为:混合可燃物、腐殖质、半腐殖质和凋落物。由于试验在室内进行,尽最大可能减少可外界误差影响,因此可以作为使用直接估计法预测白桦林含水率模型误差的最低值。     

室内模拟降雨对蒙古栎和红松凋落物床层饱和含水率的影响

【目的】地表凋落物作为森林火灾的引火物,其含水率大小决定凋落物被引燃的难易程度和发生火灾后一系列火行为指标等。降雨作为林火预测预报中必不可少的气象因子,直接影响地表凋落物的含水率。但由于降雨的不确定性,其对凋落物含水率影响的研究较少。为了搞清降雨对凋落物含水率的影响,分析降雨条件下凋落物含水率动态变化和凋落物床层饱和含水率情况。【方法】以蒙古栎和红松地表凋落物为研究对象,设置不同床层密实度和初始含水率,利用降雨模拟器在室内模拟不同降雨量,每隔10 min称量一次凋落物床层至饱和,得到凋落物含水率动态变化情况,分析床层密实度、初始含水率和降雨量对2种凋落物床层饱和含水率的影响,并建立响应预测模型。【结果】降雨条件下2种凋落物床层含水率呈对数增加;不论蒙古栎还是红松,床层初始含水率对床层饱和含水率没有显著影响,饱和含水率受床层密实度和降雨量的影响显著。降雨量和床层密实度对凋落物床层饱和含水率的影响相互制约,随着床层密实度的增加,降雨量对饱和含水率的作用下降;不同凋落物床层密实度时,建立了Ms=a×exp(b×R)的床层饱和含水率预测模型,模型预测误差均在可接受范围内。【结论】本研究揭示了降雨对不同结构的凋落物床层含水率的影响,对于含水率预测模型研究和火险预报研究具有重要意义。     

平衡含水率法预测死可燃物含水率的研究进展

死可燃物含水率预报是森林火险天气预报的重要内容,准确预测死可燃物含水率是做好森林火险天气预报和火行为预报的关键.平衡含水率法预测死可燃物含水率在物理上十分可靠,若研究对象可精确描述,理论上其含水率的预测是准确的.因此,该方法是重要的可燃物含水率预测方法.本文对该方法的理论基础和应用情况进行综述.结果表明:1)平衡含水率的预测模型主要有4种,其中Simard模型、Van Wagner模型和Anderson模型都是统计模型,其应用具有一定的局限性;而Nelson模型为半物理模型,在预测可燃物含水率上的效果好,应用广.2)可燃物类型影响平衡含水率,但具体机理还没有系统研究.3)对时滞的影响因子研究相对较少.可燃物的种类、物理性质对时滞都有影响.4)现有平衡含水率法中,Catchpole等的方法因采用Nelson的半物理模型而具普适性,有良好的应用前途.5)平衡含水率法在实际预测中得到广泛应用,是美国火险等级系统和加拿大森林火险等级系统及其他类似系统中可燃物含水率预测的主要方法.     

区域尺度森林地上生物量估计与不确定性度量

多源误差导致的不确定性普遍存在于大尺度生物量估算的整个过程中,并影响最终的估计精度。本文在传统的模型分析法中引入Monte Carlo模拟方法,对区域尺度地上生物量进行估算,分别分析抽样误差和模型误差导致的不确定性,并研究建模样本量对生物量和不确定性估计的影响。结果表明:该方法不仅可针对生物量进行稳健估计,且可分别度量由不同误差源导致的不确定性。此外,本文方法可显著降低模型误差对不确定性的影响,从而降低总不确定性。建模样本量对于大尺度生物量估计值没有显著影响,但对模型误差的影响较大,该影响不能通过增加循环次数加以改善,且总不确定性随建模样本量的减小而增加。因此,较大的模型样本量可有效地减少不确定性,从而提高生物量估计精度和效率。     

哑变量在森林蓄积量模型估测中的应用

【目的】建立含哑变量的林分蓄积量估测模型，分析哑变量在香格里拉高山松林分蓄积量模型中的意义与作用。【方法】以香格里拉为研究区，基于2008—2009年3幅TM遥感影像与2008年抽样控制样地数据，对香格里拉高山松林分神经网络模型与考虑龄组构造的哑变量神经网络模型两种类型建立蓄积量遥感估测模型，并进行精度评价。对比模型的估测值与实测值，计算模型残差，检验各龄组残差均值与0之间的差异性；同时对模型的预测值结果进行组间均值的差异性检验，以此作为确定龄组分类形式构建哑变量的标准与依据。【结果】2个模型的独立样本检验结果表明，引入哑变量的神经网络估测模型比神经网络模型拟合效果要好，其决定系数要高于神经网络模型，决定系数从0.516提高到0.783。模型预估精度从神经网络模型的66.3%提高至哑变量模型的74.8%，估算误差优于神经网络模型。【结论】根据模型的残差差异性结果得出，哑变量模型可以在一定程度上解决在估测幼龄林、中龄林蓄积量低值高估的问题；可见引入哑变量估测森林蓄积量的方法是相对有效的。     

Verification for generalizability and accuracy of a thinning-trees selection model with the ensemble learning algorithm and the cross-validation method

For the purpose of making a highly effective model in relation to the selection of trees for thinning for various forestry goals, the author examined the generalizability and accuracy of models using various ensemble learning algorithms and the m-fold cross-validation method. These techniques make it possible to improve discrimination accuracy by combining or integrating multiple learning results whose accuracies are not very high. WEKA, which is a machine learning tool for data mining programmed in Java machine language, was used to verify the results of the simulation models. The number of samples was 503. Pattern-recognition algorithms in this study used five classification-type models and one function-type model. It was found that: (1) without cross validation, two pattern-recognition algorithms can be classified as having comparatively high discrimination accuracy; (2) with cross validation, discrimination accuracy decreased as a whole, but was not very different from that without cross validation, and (3) from the viewpoint of generalizability, we constructed a model at around 70% discrimination accuracy. In order to construct more effective models, we need to design the model to utilize certain algorithms or to build in re-sampling methods such as ensemble learning and cross validation. Additionally, in the case of small sample datasets, ensemble learning is an effective method for constructing efficient models. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

DYNAMIC MODELS OF MOISTURE CONTENT OF FOREST FUELS

INTRODUCTIoNMoisturecontentoffOrestfuelsaffects.ignitionprobability,rateofspread,radia-tionefficiencyandenergyrelease.Itisoneofthemajorvariablesfordeterminingfireoccurrence,evaluatingfiredangerandpre-dictingfirebehavior.Obviouslyfuelmois-turecontentisevidenceonwhichisbasedtodetermineprescribedburning.Thechangeoffuelmoisturecontentisaffectedbymanyfactors.Thedynamicmodelsoffuelmoisturecontentcanbeestablishedinaccordancewiththosefactors.lnthispaper,thetheoryofrelativemoisturecontent,develope…     

Modeling stem profile of Lebanon cedar,Brutian pine,and Cilicica fir in Southern Turkey using nonlinear mixed-effects models

The Max and Burkhart segmented taper model was fitted using nonlinear mixed-effects modeling techniques to account for within- and between-individual stem profile variation for Lebanon cedar (Cedrus libani A. Rich.), brutian pine (Pinus brutia Ten.), and cilicica fir (Abies cilicica Carr.) in Turkey. About 75% of the trees were randomly selected for model development, with the remainder used for model validation. Diameter measurements from various heights were evaluated for tree-specific calibrations by predicting random-effects parameters using an approximate Bayesian estimator. The procedure was tested with a validation dataset. Predictive accuracy of the model was improved by including random-effects parameters for a new tree based on upper stem diameter measurements. Prediction in stem diameter was less biased and more precise across the all sections of bole when compared to predictions based only on fixed-effects parameters. In the future, the proposed mixed models can be applied to region wide three species stands by fitting the model to a larger data set that more closely represents regional variation.     

Application of VIS/NIR spectroscopy for estimating chemical,physical and mechanical properties of cork stoppers

The aim of this study was to evaluate the potential of visible and near infrared spectroscopy (Vis/NIRS) in predicting the chemical, physical and mechanical behavior of single-piece natural corks stoppers used for sealing wine bottles. Two training sets of 90 and 150 cork stoppers were used to obtain four spectra per sample in different positions: two of the stopper bases (transversal section) and two of the stopper sides (tangential section and radial section). The samples were scanned in the range of 400–2,500 nm using a Foss-NIRSystems 6500 SY II spectrophotometer equipped with a remote reflectance fiber-optic probe. On each training set, two-thirds of the samples were used to develop modified partial least square (MPLS) calibration equations, and the remaining one-third of the sample for the external validation of these MPLS equations. The best equations were obtained for the transversal section, which is the recommended one when applying Vis/NIRS technology to cork. The best results for the chemical composition were obtained for waxes and total polyphenols, showing coefficient of determination of the cross validation (r _cv ² ) values of 0.64 and 0.56 and coefficient of determination of the external validation (r _EV ² ) values of 0.53 and 0.55, respectively. The best equation for the physical and mechanical parameters was obtained for moisture content (r _cv ²  = 0.86 and r _EV ²  = 0.85), with somewhat lower results for density, compression force and extraction force (r _cv ²  = 0.66, 0.72, 0.52 and r _EV ²  = 0.52, 0.49, 0.51, respectively). The SECV (standard error of cross validation) and SEP (standard error of external validation) were similar for all the physical and mechanical parameters, thus confirming the robustness of the equations. MPLS model for moisture content fulfills the requirements for screening (RPD >2.5), but MPLS models obtained for waxes, total polyphenols, density, compression force and extraction force are not good enough for routine analysis or quality control. The results obtained from the MPLS models based on Vis/NIRS technology would permit the continuous quality control of humidity in the production line as well as obtaining information about certain chemical components (extractives contents) and some physical and mechanical parameters (density, extraction force and compression force).     

Modelling diapause termination and phenology of the Japanese beetle,Popillia japonica

We deve?loped a mechanistic, stage-structured model simulating the phenology of Popillia japonica. The model simulates the influence of soil temperature on the larval diapause termination and on the development rate function of post-overwintering larvae and pupae. Model parameters are estimated based on literature evidence for pupae development and on a parameterisation process that allows estimating parameters for larval diapause termination and for the development rate function (and the related uncertainty) of post-overwintering larvae. Data used for model parameterisation and validation refer to time-series adult trap catches collected during the P. japonica monitoring programme performed by the Phytosanitary Service of Lombardy Region within the infested area in Lombardy (Italy) from 2015 to 2019. A total of 12 randomly selected locations are used to estimate biologically realistic model parameters (parameterisation dataset). We applied a Jackknife nonparametric resampling procedure on the parameterisation dataset to quantify uncertainty associated with parameters’ estimates. Parameterised model is then validated on time-series adult trap catches data referring to a different set of 12 randomly selected locations (validation dataset) surveyed in Lombardy. The model successfully predicted the beginning of adult emergence and the overall curve of adult emergence in the validation dataset. The model presented can support the definition of the best timing for the implementation of monitoring and control activities for the local and the area-wide management of P. japonica.

     

近红外光谱法测定毛竹综纤维素的含量研究

研究了用近红外光谱(NIR)结合多变量统计分析技术对毛竹综纤维素含量的快速测定。用常规实验室方法测定了54个竹材样品的综纤维素含量,用近红外光谱仪采集相应样品的光谱,对原始光谱进行二阶导数和25点平滑预处理后,从54个竹材样品中挑选41个代表性的样品建模,选择1011~1675nm和1930~2488nm波段区间,用偏最小二乘法(PLS1)和完全交互验证方式建立毛竹综纤维素含量的预测模型。结果表明,毛竹综纤维素含量和近红外光谱之间存在非常好的相关性,预测模型的相关系数(RP)为0.95,预测模型的标准偏差(SEP)为0.76%。     

On the validation of models of forest CO2 exchange using eddy covariance data: some perils and pitfalls

With the widespread application of eddy covariance technology, long-term records of hourly ecosystem mass and energy exchange are becoming available for forests around the world. These data sets hold great promise for testing and validation of models of forest function. However, model validation is not a straightforward task. The goals of this paper were to: (1) review some of the problems inherent in model validation; and (2) survey the tools available to modelers to improve validation procedures, with particular reference to eddy covariance data. A simple set of models applied to a data set of ecosystem CO2 exchange is used to illustrate our points. The major problems discussed are equifinality, insensitivity and uncertainty. Equifinality is the problem that different models, or different parameterizations of the same model, may yield similar results, making it difficult to distinguish which is correct. Insensitivity arises because the major sources of variation in eddy covariance data are the annual and diurnal cycles, which are represented by even the most basic models, and the size of the response to these cycles can mask effects of other driving variables. Uncertainty arises from three main sources: parameters, model structure and data, each of which is discussed in turn. Uncertainty is a particular issue with eddy covariance data because of the lack of replicated measurements and the potential for unquantified systematic errors such as flux loss due to advection. We surveyed several tools that improve model validation, including sensitivity analysis, uncertainty analysis, residual analysis and model comparison. Illustrative examples are used to demonstrate the use of each tool. We show that simplistic comparisons of model outputs with eddy covariance data are problematic, but use of these tools can greatly improve our confidence in model predictions.