Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/65720
標題: 森林資源取樣調查規劃與資料分析─以新化實驗林場為例─
Forest Inventory Sampling Plan and Analysis- An Illustration of Hsin-Hua Experiment Forest Station -
作者: 陳志賢
Chen, Chih-shian
關鍵字: Inventory;資源調查;Sampling Plan;Sampling desings;Plot sizes;Sampling;取樣規劃;取樣設計;樣本大小;取樣形狀;樣本面積;永久樣區
出版社: 森林學系
摘要: 
本文研究結果摘述如下:(1)研擬建立本省資源調查體系,以提供長期、
連續及整合性的資源調查在蒐集資料時之參考,藉以改善現有及過去的組
織架構與野外調查中獲得資訊的過程。(2)比較四種取樣調查方法(長方
形樣區、圓形樣區、樣線法、樣點法),結果指出,在同時考慮精確度及
精密度的原則下,選擇長方形樣區做為永久樣區的設立為最佳的取樣方法
。(3)樣區面積利用不同的取樣面積所造成之材積變異曲線,求得372.58
平方公尺為一最適面積。(4)研擬永久樣區的有效取樣規劃理論藉以整體
性的考量整個取樣過程,進而有效的提高精確度及精密度。(5)研擬一完
整的調查制度,以達成標準化及一致性調查的流程步驟與方法,做為設立
永久樣區之參考。(6)利用地理資訊系統疊圖、擷取及邏輯判斷分析的功
能,解決私有林地籍分散分佈的情形,以提供林地管理之參考。(7)新化
林場調查結果指出:林場內之各主要林型分佈面積為262.75公頃,總蓄積
量為26162.7605立方公尺。並得知三種不同容許誤差範圍下信賴區間之統
計值取樣個數,分別為誤差率30% 精密度70% 時應設30個樣區,誤差
率20 %精密度80% 時應設82個樣區,誤差率10%精密度90% 時應設 373個
樣區。並建立了新化林場內各樹種之樹高曲線式與材積式資訊庫。目前設
有62處永久樣區提供連續調查使用。(8)本次新化林場樣區材積在95%信
賴區間時,為4.9784±0.6862(m/0.05ha),並且求得機差百分率為13.78%
(顯著水準為5%),稍微超過10% ,可知調查樣區62區,在樣區材積推估
時的代表性仍嫌不夠,所以未來應再增設永久樣區。

(1)Some consideration in establishing forest inventory system
in Taiwan were studied in the paper. It offers a reference to
support a long-term、successive and intergration forest
inventory in data collection and improves the organization and
field survey to got the information process. (2)Comparing the
four sampling methods (rectangle sampling ,circle sampling,
line sampling, point samply), we got the conclusion that
rectangle sampling is the best permanet plot if we considered
both precision and accuracy. (3)From differen plot areas we got
the volume variation curve and got a calculation that the
smallest area is 372.58m.(4)In the paper, we established an
effective theory of sampling plan to heighten the accuracy and
precision. (5)In the paper, a complete inventory system with
standard and unifying process could be established and be a
reference for the permanent plot. (6)The function of overlay,
selection and logical analysis of GIS could solve the problems
of the spread of private forest lands. (7) According to the
inventory conclusion of Hsin-hua Experimental Forest Station,
the area of main spread land is 262.75 ha, and the total volume
is 26162.7605m. Three different confidence interval to sample
size are as following:when the error is 30% and the precision
is 70%,30 plots should be established;when the error is 20%
and precision is 80%, 82 plots should be established; when
the error is 10% and precision is 90%, 373 plots should be
established. There are also tree height and volume equations
database about species of trees in Hsin-hua Experimental Forest
Station. 64 permanent plots are established for successive
inventories. (8)When the confidence interval is 95%, the
volume of a plot in Hsin-hua Experimental Forest Station is
4.9784±0.6862 (m/0.05ha),and the percent of standard error is
13.78%,(Significance level is 5%) more than 10%. We got the
conclusion that 62 plots is not enough to estimate he volume,
so it is necessary to increase
URI: http://hdl.handle.net/11455/65720
Appears in Collections:森林學系

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.