本文選題：森林資源調查 + 誤差分析�。� 參考：《北京林業(yè)大學》2014年博士論文

【摘要】：森林是地球生態(tài)系統(tǒng)的重要組成部分,是人類社會得以持續(xù)健康發(fā)展的基礎,我國為了掌握宏觀森林資源現(xiàn)狀與動態(tài),經過五十余年的努力,構建了一整套較為完善的國家森林資源連續(xù)清查技術體系,并以5年為一個調查周期最終匯總形成面向全國的森林資源清查結果報告。其關鍵技術是精確測定同一固定樣地的初查和復查數(shù)據,并利用所取得的成對值進行直接分析比較,以得到森林資源現(xiàn)狀及其消長變化。 本文圍繞“森林資源調查技術與方法”這一重要的林業(yè)科學技術問題展開了較為創(chuàng)新和系統(tǒng)的研究,主要研究內容包括利用誤差傳播定律和數(shù)理統(tǒng)計方法分析森林資源調查中包含的各種誤差及其相互關系,為了盡可能消除這些誤差,研發(fā)高精度的森林調查儀器并驗證其在固定樣地調查中的使用效果和精度,研究利用多邊形法進行森林蓄積量估測。 首先,將森林資源調查中產生的誤差分為抽樣誤差和觀測誤差。我國森林資源調查精度是以抽樣誤差的大小來表示,不包括觀測誤差,因此可能會出現(xiàn)抽樣精度很高,但實際精度較低的情況。抽樣誤差是指由于樣地抽樣設計不當,導致樣本平均數(shù)與總體平均數(shù)產生誤差,而根據誤差產生的不同環(huán)節(jié),我們將觀測誤差分為樣地面積誤差、樣地量測誤差和樣地模型誤差,利用誤差傳播定律和數(shù)理統(tǒng)計方法,結合若干實際算例介紹了每種誤差和總誤差的計算方法及其相互關系,并提出了相應的系統(tǒng)改進設想。 其次,為了盡可能的消除誤差提高調查精度,對森林資源調查的關鍵技術與方法進行了研究,以自主研發(fā)的FTS-2型測樹全站儀為森林資源調查裝備,以30塊大型喬木固定樣地,約24500棵樣木為研究對象進行調查試驗,結果表明,利用高精度FTS-2型測樹全站儀進行森林固定樣地調查,可以得到固定樣地中每棵樹木的三維坐標,其中坐標測量中誤差±4mmm,高程中誤差±7mm；每木樹高值測量精度約為97%,平均用時5秒；每木胸徑值測量相對誤差僅為2%,平均用時約為3秒；每木冠幅測量值相對誤差10%,并實現(xiàn)樹冠體積量測和三維表達；立木材積測量值精度可達95%。同時實現(xiàn)樣木分布成圖、樣地樣木精準復位與樣地數(shù)字化三維建模；此外,以全站儀每木三維坐標為離散點信息,利用Voronoi多邊形算法實現(xiàn)林分空間結構參數(shù)信息化提取和分析,最后介紹了其他先進設備--三維激光掃描儀和電子測樹槍在樣地調查中的應用,旨在為我國森林資源調查關鍵技術與方法提供先進精準的儀器裝備和調查手段。 最后,以FTS-2型測樹全站儀為調查儀器,以多邊形法為森林蓄積量調查方法,在北京市區(qū)及周邊區(qū)縣實驗林場中隨機抽取共計70個角規(guī)抽樣點進行調查。通過與傳統(tǒng)角規(guī)蓄積量調查方法比較得出多邊形樣地形狀為六邊形即樣地邊界木選擇6棵時,林分蓄積量估測精度較高。同傳統(tǒng)角規(guī)樣地調查相比,利用多邊形樣地進森林蓄積量、林分密度、每公頃斷面積、林分平均高等林分因子的調查具有更好的穩(wěn)定性和準確性,精度均可達90%左右,該方法為我國森林蓄積量估測提供了新的方法和技術支持。
[Abstract]:The forest is an important part of the earth's ecosystem, is the basis for sustained and healthy development of the human society to our country, in order to grasp the current situation and dynamic macro forest resources, after fifty years of effort, establishes a perfect continuous assessment system of national forest resources, and in 5 years as a final summary form survey cycle the results of the national forest resources inventory report. The key technology is a fixed sample of the preliminary investigation and review the data accurately measured, and the use of paired values obtained by direct analysis and comparison, in order to get the change status and dynamics of forest resources.
This paper studied more innovation and system around the "technology and method of forest resources investigation of this important forestry science and technology, the main research contents include all kinds of error analysis in forest resource survey included the use of error propagation law and the method of mathematical statistics and their relations, as far as possible in order to eliminate the errors, the forest survey instrument with high precision the development and validation of the plots in the use of performance and accuracy, using the polygon method to estimate forest volume.
First of all, the errors in forest resource survey is divided into sampling error and observation error. The accuracy of the survey of forest resources in China is in terms of the size of sampling error does not include the observation error, so that there might be sampling accuracy is very high, but the actual accuracy is too low. The sampling error is due to sampling design improper, cause sample average and ensemble average error, and according to the different aspects of the error, we will be divided into the observation error area error plots, error and sample error model of sample volume, the error propagation law and the method of mathematical statistics, combined with some practical calculation methods for each error and the total error and their relations, and put forward the corresponding improvement suggestion system.
Secondly, as far as possible in order to eliminate the error and improve the accuracy of survey, key technologies and methods of investigation of forest resources were studied by FTS-2 independent research and development of measuring total station for forest resources survey equipment, with 30 pieces of large trees in fixed samples, about 24500 sample trees as the research object of investigation results show that test. Forest plots, tree measuring total station with high precision FTS-2, can obtain the three-dimensional coordinates of every tree plots, in which the error in the coordinate measuring elevation error + 4mmm + 7mm; each tree is about 97% high value measurement precision, the average time of 5 seconds per DBH; the value of the relative measurement error is only 2%, the average time of about 3 seconds; each tree crown measured relative error of 10%, and the canopy volume measurement and 3D expression; standing volume measurement precision can reach 95%. and wood distribution map, kind of like wood Accurate reduction and sample digital 3D modeling; in addition, the total station per wood three-dimensional coordinates of discrete points of information, realize the extraction and analysis of forest spatial structure parameter information using Voronoi polygon algorithm, finally introduces the application of other advanced equipment -- 3D laser scanner and electron gun in tree measurement plots, in order to provide advanced precision the apparatus and method for the investigation of key technologies and methods of investigation of forest resources in China.
Finally, the FTS-2 tree measuring total station survey instrument, the polygon method of forest volume for investigation, investigation in Beijing and the surrounding counties were randomly selected in the experimental forest of a total of 70 points. The sampling with the traditional gauge volume survey method to obtain the polygon like shape is hexagonal that kind of comparison 6 tree boundary trees, stand volume higher accuracy. Compared with the traditional gauge field survey, using polygonal plots into forest volume, forest density, basal area per hectare, higher forest stand average factors has better stability and accuracy, precision can reach 90% or so, the method for me the forest volume provides support for new methods and techniques of estimation.

【學位授予單位】：北京林業(yè)大學
【學位級別】：博士
【學位授予年份】：2014
【分類號】：S757.2

【參考文獻】

相關期刊論文 前10條

1 唐雪海;王田磊;袁進軍;馮仲科;劉德慶;聶敏莉;李影;;數(shù)字近景攝影測量輔助三維激光掃描用于森林固定樣地測樹原理探討[J];安徽農業(yè)科學;2010年12期

2 馮仲科,韓熙春,周科亮,南永天,付曉;全站儀固定樣地測樹原理及精度分析[J];北京測繪;2003年01期

3 馮仲科,景海濤,周科亮,吳鴻華;全站儀測算材積的原理及精度分析[J];北京林業(yè)大學學報;2003年03期

4 馮仲科;劉永霞;;森林生物量測定精度分析[J];北京林業(yè)大學學報;2005年S2期

5 梁長秀;韓光瞬;馮仲科;周心澄;;羅盤導線定位及其精度分析[J];北京林業(yè)大學學報;2005年S2期

6 馮仲科;羅旭;馬欽彥;郝星耀;陳曉雪;趙利果;;基于三維激光掃描成像系統(tǒng)的樹冠生物量研究[J];北京林業(yè)大學學報;2007年S2期

7 熊妮娜;王佳;羅旭;馮仲科;韓光瞬;;一種基于三維激光掃描系統(tǒng)測量樹冠體積方法的研究——以油松為例[J];北京林業(yè)大學學報;2007年S2期

8 施一民;電磁波測距邊歸算至投影面的公式論證及應用討論[J];測繪通報;2000年12期

9 陳霖生;;“六株木樣地法”測定落葉松人工林蓄積量的奻歐[J];東林科技;1975年02期

10 關強;尹麗麗;李志鵬;詹長書;;基于超聲測距的定高樹徑測量儀的研究[J];東北林業(yè)大學學報;2006年04期

，

本文編號：1732709