查詢結果分析
相關文獻
- Geostatistical Cross-Validation for the Design of Additional Sampling Regimes in Heavy-Metal Contaminated Soils
- 以地理統計結合X射線螢光光譜分析推估土壤銅濃度的空間分佈
- 逐步高斯模擬法與克利金法於流通係數空間變異研究
- 條件模擬法模擬小樣區土壤鋅重金屬污染
- 克利金法於環境規劃之應用領域探討
- 雨量站網設計與評估--區域化變數理論之應用
- 地理統計預測污染土壤中重金屬的空間分佈(1)--極端值與半變異圖模式的影響
- Adaptive Sampling Used for Kriging Maps of Heavy Metals in Contaminated Soils
- 標準貫入試驗值之水平空間分布模擬
- 利用地理資訊系統評估臺南安順污染廠址周圍之土壤污染
頁籤選單縮合
題 名 | Geostatistical Cross-Validation for the Design of Additional Sampling Regimes in Heavy-Metal Contaminated Soils=以地理統計的交叉驗證法決定重金屬污染土壤的再次採樣策略 |
---|---|
作 者 | 莊愷瑋; 李達源; 陳尊賢; | 書刊名 | 中國環境工程學刊 |
卷 期 | 9:2 1999.06[民88.06] |
頁 次 | 頁89-96 |
分類號 | 445.6 |
關鍵詞 | 克利金法; 空間變異; 土壤污染; Kriging; Spatial variability; Soil contamination; |
語 文 | 英文(English) |
中文摘要 | 地理統計的克利金法目前已常被使用在土壤中污染物含量的空間內插估計 ,可有助於界定土壤污染的範圍並做為污染風險評估與復育的參考。然而,污染土壤中污染 物的資料分佈常表現極大的變異與高度的偏歪,這種現象反映了污染物在土壤中分佈的不均 質性。以重金屬為例,不均質性主要發生在相鄰採樣位置的觀測值間有很大的差異時 (當高 的觀測值週圍有較低的觀測值或低的觀測值週圍有較高的觀測值時),這樣的不均質性常會 使得以克利金法所估計之空間分佈圖來界定污染的範圍變為困難,一旦界定污染範圍錯誤之 可能性大於可容忍的範圍時,則必須進行再次採樣。在本研究中,我們提出以使用地理統計 的交叉驗驗法決定應優先再次採樣的位置與樣本個數,並以臺灣桃園縣境內污染土壤中鎘濃 度資料為例做說明。結果顯示,由交叉驗證法所得之絕對估計誤差 (克利金推估值與觀測值 之差的絕對值 ) 可用來決定優先再次採樣的區域,並配合採樣預算決定採樣個數。 |
英文摘要 | To delineate areas of contaminated soils for risk assessment and remediation, the kriging technique is often used for the spatial interpolation of a pollutant. However, the data distributions of heavy metal concentrations in contaminated soils usually exhibit great variation and high skewness. This indicates the heterogeneity of spatial distributions of pollutants in contaminated soils. The heterogeneity of heavy metals in contaminated soils is usually observed from the great deviations among some neighboring observations (i.e. a high-valued observation surrounded by low-valued observations or a low-valued observation surrouhnded by high-valued observations). It is difficult to correctly delineate hazardous areas of heavy-metal contaminated soils based on the spatial distribution obtained by kriging. If the possiblity of incorrect delineation is no longer tolerated, them further sampling should be adopted to identify the hazardous areas of contaminated soils. In this study, we used geostatistical cross-validation to determine the locations of prime candidates for additional sampling in contaminated soils. A data set of the soil Cd concentrations in a contaminated site in Taoyuan County, Taiwan, is used for illustration. The results show that he magnitudes of the absolute estimation errors (AEEs), the absolute values of kriging estimated-minus-observed values, generated by the cross-validation procedure and based on the budget available for sampling, can be used to determine where the sampling density should be increased for additional sampling. |
本系統中英文摘要資訊取自各篇刊載內容。