2009-05-042024-05-17https://scholars.lib.ntu.edu.tw/handle/123456789/681689摘要：本計畫將建立克利金法和Gandin法程式，將方法技術引進氣象局。透過數值天氣預報模式建立高解析度的OSSE溫度數據，評估此二方法和Barnes法、Cressman法、距離倒數權重法、周思運博士溫度內插方法的誤差行為比較，將數據、方法論和思維過程，呈現為報告，提供氣象局參考。 利用克利金法和Gandin法，以氣象局自動化氣象站的數據，內插估計台灣本島高解析度網格地面溫度，資料時間至少一年。評估比較此二方法和Barnes法、Cressman法、距離倒數權重法、周思運博士溫度內插方法計算結果的可能優劣。最後，利用克利金法和Gandin法，建立Gridded MOS的方法論，聘評估其可應用性。<br> Abstract: This study will result programs of Universal Kriging and Universal Gandin techniques. They together with their theory will be delivered to Central Weather Bureau. OSSE surface temperature data are to be produced through careful designed WRF simulations. These data will be used to evaluate the differences in estimation error behavior of the two methods as well as Barnes, Cressman, Inverse Distance Weighting schemes, and that of Dr. Jou’s. The data, methodology, thinking and procedures are all to be written in the report to CWB. Utilize the verified Kriging and Gandin model to estimate temperatures at high resolution grid points. The processed data duration is at least 1 year. Infer the pros and cons of the estimates using these two method with the other methods mentioned above. Lastly, derive the methodology or applicability of using Universal Kriging or Gandin techniques for Gridded MOS.