驗證GPS衛星訊號隱含之水氣含量

摘要水氣對於地球上來說是原本存在的物理現象，對於氣候上的變遷，水氣扮演一種能量傳遞方式，相較於整個大氣層中的其他氣體時，在呈現上屬於比較不穩定的狀態，也對人類的生活帶來許多的優缺點。因此，若能有效的獲取大氣中水氣含量（PWV）資訊，對於天氣狀態的研究與分析以及氣象的預報會有相當之協助，本研究以地面GPS接收訊號來求解對流層天頂向的水氣含量，並藉由地面的雨量資料來釐清與PWV之間的關係。GPS衛星資料是以Bernese 5.0求解水氣含量，並利用水氣微波輻射儀（WVR）所觀測到的濕延遲量來驗證本研究所計算之成果。成果顯示，WVR觀測值及GPS計算值呈現一致的趨勢，相關係數都在0.9以上，平均誤差則介於-1至-2 mm之間。若將計算成果與美國大氣研究中心（NCAR）相對應，由數據顯示NCAR的估計值有5~6 mm的低估現象。換言之若以WVR的觀測值為標準，我們計算的成果低估PWV約1~2 mm，NCAR的計算成果則低估PWV約4~6 mm。PWV較高的情況下通常即有降雨的跡象，平地的PWV在60 mm、山區的PWV在40 mm就容易有降雨的現象，本研究藉由數據上的統計分析了解天氣的狀態，希望對氣象的預報上提供參考的數據。

關鍵字

全球定位系統；水氣微波輻射儀；對流層；濕延遲；水氣含量；降雨量

並列摘要

Abstract Water vapor is part of the evaporation process—a long existing physical phenomenon on earth. It transfers energy in the Nature as the weather changes, and is therefore less stable than other types of gasses in the atmosphere. Because of its instability, it affects people’s lives in both good and bad ways. If we can more effectively obtain information on Precipitable Water Vapor (PWV) in the atmosphere, we will be able to conduct more detailed researches and analyses about the weather and give more accurate forecasts. This research aims to measure water vapor in tropospheric zenith delay through signals received from the global positioning system (GPS). It also seeks to understand the relations between GPS and PWV through precipitation data. Bernese 5.0 is used to measure water vapor, and the wet delay data obtained through Water Vapor Radiometer (WVR) is used to verify such calculation. Results show that, in terms of measurement accuracy, WVR and GPS works just as well as one another, with their correlation coefficients standing over 0.9 and average errors between -1 and -2 mm. If comparing GPS measurement with that of the National Center for Atmospheric Research (NCAR), however, the NCAR statistics appear to have an underestimation of 5 to 6 mm. In other words, based on the WVR measurement, GPS appears to underestimate PWV by only 1 to 2 mm, while NCAR appears to underestimate PWV by 4 to 6 mm. In addition, there is often rain when PWV is high, such as 60 mm in plain areas and 40 mm in the mountains. We hope the statistical results of this research can not only help researchers understand more about the weather, but help improve the quality of weather forecasting in the future.