透過分散式光纖溫度感測器探索複雜地形中微氣象過程：揭示溫度動態和氣流模式

Abstract

大氣的觀測能提供目前與過去大氣變化的重要訊息，這些資料有助於我們了解大氣中能量的傳遞與平衡以及短期或長期的天氣變化，而目前地面的大氣觀測方式主要是單點的觀測，不過在城市或複雜地形中其空間變異性很大，如果我們使用單點的觀測資料代表這個區域的數值可能造成資料的失真，因此傳統上我們需要增加測站的密度才能解決這一問題，不過測站的增加會使得人力與維護成本大幅提高，因此我們時常需要在成本與觀測密度間取得平衡點。 分散式光纖溫度感測器(DTS)是一種新型態的溫度觀測方式，它通過量測雷射光在光纖纖維不同位置上產生的反向拉曼散射(Raman scattering)強度得到整段光纖不同位置的溫度，能提供我們長距離(>2000公尺)與高空間解析度(<0.5公尺)的溫度資料，目前此儀器已成熟的應用在不同領域，而近年來國際上開始有許多研究通過此方式觀測大氣的變化，不過目前國內對此方式的觀測還相對陌生。 為了瞭解DTS在大氣觀測的特性與應用於森林中的可行性，本研究首先於花蓮七星潭測試光纖在不同包覆方式下(PVC與不鏽鋼)對於氣溫觀測的特性，並且測試主動加熱光纖法推估二維風速的可能性。對於氣溫的觀測，最主要的干擾來自太陽輻射，不鏽鋼包覆的光纖對於短波輻射的變化相當敏感而PVC包覆的光纖則幾乎不受影響，應用此特性能很好的使用兩種不同包覆方式光纖間的溫差推估短波輻射強度，結果顯示光纖間溫差與現場短波輻射的觀測值的Pearson相關係數可達0.98。另一方面是通過加熱與未加熱的不鏽鋼包覆光纖間的溫差推估風速，其估計出的一分鐘平均風速與WXT觀測值的平均絕對誤差為0.06 m/s。 實際應於森林觀測，DTS有效的觀測到過往傳統儀器無法記錄到的微氣象過程，其中發現光斑分布與冠層開闊度的關係並不明顯，而在冬季晴朗天氣下陽光能直接照射森林地表約只有五小時。另一方面在水平溫度剖面的觀測中風對於氣溫的影響最大，在日間平均風速較高的區域呈現較低的氣溫，夜間則是相反。使用垂直風速與溫度剖面能很好呈現冠層內外氣流的互動過程，計算夜間平均紊流強度剖面得到0.06的常數值，使用此數值檢查2/23-24發現約42%的時間冠層內外的氣體可能缺乏垂直混合，顯示處於解耦狀態。 結果顯示利用DTS可以記錄森林冠層內部氣溫、日照、風速等細微的變化，以利於我們對於複雜環境中的微氣象過程有不同的見解。

Atmospheric observations provide crucial data about current and historical atmospheric changes, aiding our understanding of energy transmission and balance as well as short-term and long-term weather variations. Currently, ground-based atmospheric measurements are primarily point-based. However, in urban areas or complex terrains, these measurements can vary significantly in space, and using point-based data to represent regional values may lead to data distortion. Traditionally, increasing the density of monitoring stations has been the solution to this issue, but this leads to significantly higher labor and maintenance costs. Thus, a balance often needs to be struck between cost and observation density. Distributed temperature sensing (DTS) represents a new form of temperature observation, utilizing measurements of backscattered Raman light at various points along a fiber to determine temperature profiles over long distances (>2000 m) with high spatial resolution (<0.5 m). This technology is well-established across various fields and has begun to be used internationally for atmospheric studies, although it is still relatively unfamiliar domestically. In order to understand the characteristics of DTS in atmospheric observation and its feasibility in forests, this study first tested the characteristics of temperature observation using DTS under PVC coated fiber (FO-PVC) and stainless steel jacketed fiber (FO- STL) in Qixingtan, Hualien, and tested the possibility of estimating two-dimensional wind speed using the active heating fiber method. The main interference in air temperature observations comes from solar radiation, with FO-STL cables being highly sensitive to changes in shortwave radiation while FO-PVC cables are largely unaffected. Using this characteristic, the temperature difference between the two types of coated fibers can be used to estimate shortwave radiation intensity effectively, yielding a correlation r = 0.98 with actual shortwave radiation measurements on-site. On the other hand, wind speed is estimated by the temperature difference between the heated and unheated FO-STL cables. The mean absolute error (MAE) of the one-minute mean wind speed estimated by this method compared to the WXT observed values is 0.06 m/s. In practical forest observations, DTS can effectively monitor various micrometeorological processes. It was found that the relationship between the distribution of light spots and the openness of the canopy is not obvious. During clear winter weather, sunlight can directly reach the forest floor for about five hours. On the other hand, in the observation of horizontal temperature profiles, wind has the greatest impact on air temperature. Areas with higher average wind speeds during the day exhibit lower temperatures, while the opposite is true at night. Using vertical wind speed and temperature profiles can effectively show the interaction of airflows inside and outside the canopy. Calculating the average nighttime turbulence intensity profile yields a constant value of 0.06. Using this value to check data from February 23-24, it was found that about 42% of the time, the air above and under the canopy lacks vertical mixing, indicating a decoupled state. The results show that using DTS can study processes such as temperature, solar radiation, and wind speed within the forest canopy in greater detail, providing us with different insights into micrometeorological processes in complex environments.