查詢結果分析
來源資料
相關文獻
- 協合遙測與數值地形模型於臺灣冷杉生育地之模擬
- 埔里地區Landsat TM影像地形效應糾正之探討
- 衛星影像地形效應糾正方法適用性之探討
- 航照遙測及數值地形模型技術在農林土地使用規劃上之應用
- 三維遙測影像處理技術於活動斷層調查之應用
- 小型SCADA系統之可靠度與架構探討
- 應用DTM資料探討合歡山地區臺灣冷杉空間分佈之特性
- Mapping Indicators of Land Degradation Induced by Bush Invasion in Jornada Southern New Mexico Using Landsat TM Imagery
- 從地理期刊之發表探討臺灣近卅年來地圖學之發展趨勢
- 遙測技術在農業上之應用
頁籤選單縮合
題 名 | 協合遙測與數值地形模型於臺灣冷杉生育地之模擬=Synergism of Remote Sensing and Digital Terrain Model for Abies Kawakamii Habitat Simulation |
---|---|
作 者 | 黃凱易; 李旻旻; | 書刊名 | 中華林學季刊 |
卷 期 | 32:1=124 1999.03[民88.03] |
頁 次 | 頁91-107 |
分類號 | 436.718 |
關鍵詞 | 臺灣冷杉; 遙測; 數值地形模型; 屏障區地形假說; 協合增效作用; Abies kawakamii; Taiwan fir; Remote sensing; Digital terrain model; Shelter district terrain hypothesis; Synergism; |
語 文 | 中文(Chinese) |
中文摘要 | 本研究之目的係結合衛星影像與數值地形模型(Digital Terrain Model, DTM)探討合 歡山地區台灣冷杉(Abies Kawakamii ; Taiwan fir)之空間分佈特性,並且建立地理資訊系統 (Geographic Information System, GIS)模式推測冷杉之適生育地。本研究將土地利用數值圖檔 與DTM圖檔套疊,產生並分析全區與冷杉在海拔、坡度、坡向及水分梯度之統計值,以了 解冷杉在各因子分佈特性及其對冷杉生育地之重要性;同時將其與Landsat TM影像套疊並 分析累計頻率圖及直方圖,以產生最佳冷杉辨識波段。卡方及K-S(Kolmogorov-Smirnovr)檢 定用以驗證分析結果。本研究建立兩類GIS套疊推測模式:一為等權模式;另一為非等權 之卡方模式,其中非等權模式中各因子之權重係由卡方檢定計算而得。模式準確度評估,係 以模式建立用樣本及全部預測冷杉像元完成。卡方檢定及K-S檢定結果顯示海拔為生育地 之極重要因子,而坡度、坡向及水分梯度則為次要因子,模式準確度評估結果亦是如此。本 研究提出之「屏障區地形假說」可用以解釋為何坡向及水分梯度在冷杉生育地模式中退居次 要因子。更重要的是模式加入TM影像導出波段SBI,其推測準確度稍高於單因子海拔模式 之推測結果,表示二者之結合於冷杉生育地之模擬,可以產生協合增效作用,但不顯著,乃 係因為模式建立用樣本取樣差異、TM衛星影像空間解析力不適用於台灣崎嶇地形與雜異且 細碎之地覆及缺乏有效之衛像地形效應糾正。後續研究將針對這些問題加以深入探討。 |
英文摘要 | The objectives of this study were to investigate the characteristics of spatial distribution of Abies kawakamii (Taiwan fir) in Hohuan mountain areas by combining satellite image and digital terrain model (DTM) data and to predict the potential habitat for Taiwan fir. The digitized land use map was overlaid with DTM maps generating and then analyzing the statistics of elevation, slope, aspect, and moisture gradient for the entire area and fir areas to understand the distribution characteristics of each factor and its importance for Taiwan fir habitat. This map was also overlaid with Landsat TM image to identify the best discriminating band between fir and non-fir pixels via spectral pattern analysis and cumulative frequency graphs. Two predictive GIS overlay models were developed: an equal-weight model and a chi-square model. For an unequal-weight model, the importance of each habitat parameter was evaluated with a chi-square test. Two types of accuracy assessment were performed using the fir samples in model development and all of the fir pixels in a predicted habitat map. Chi-square test and K-S test results indicated that elevation was an extremely important habitat parameter, whereas slope, aspect, and moisture gradient were minor ones. Model accuracy assessments provided the same results as those acquired from the chi-square tests and K-S tests. The "shelter district terrain hypothesis" proposed in the study might explain why both aspect and moisture gradient factors became the minor parameters for Taiwan fir habitat modeling. Of more significance was that two-factor models of elevation and TM derivative band (SBI) had slightly higher accuracy than one-factor model of elevation did. The results indicated that synergism of remote sensing and DTM could be achieved to improve the accuracy of predicting Taiwan fir potential habitat, but not significant. The reasons included sampling bias associated with fir samples for model development, inappropriate spatial resolution of TM image for Taiwan's rugged terrain and heterogeneous land cover, and scarcity of effective topographic normalization of TM image. Attention will be paid to these problems in the follow-up studies. |
本系統中英文摘要資訊取自各篇刊載內容。