如何正確測量或推估海－氣間的動量、可感熱和潛熱通量，是研究海－氣交互作用的一個重要課題。本文根據Lo (1993)的估計法，並且參考Liu et al. (1979)加入分子效應，估計1992年11月12日至11月26日，日本觀測船Hakuho在0°N、156°E附近的海－氣通量資料。同時，文中也採用TOGA-COARE的整體通量參數法。由這兩種方法推估的海－氣通量，也與Hakuho觀測船以渦流相關法直接量測的海－氣通量資料比較。推估結果顯示，海-氣通量具有明顯的日夜變化；白天平均海面的淨熱通量為正值，加熱海洋，夜間則為負值，冷卻海洋。海－氣通量半日變化也很明顯，可能受到海面氣象場的變化所致。兩種不同方法所估計的潛熱通量和觀測資料比較，均較觀測值為大；至於可感熱通量，本文所發展的方法，其推估結果，比TOGA COARE整體通量參數法之推估，較接近觀測值。兩種方法得到的整體傳送係數隨風速的變化都顯示，在風速約等於4 m/s時，整體傳送係數最小。風速大於4 m/s時，整體傳送係數會隨風速增加而增加；相反地，在風速小於4 m/s 時，則隨風速減小而明顯增加。此外，在一般整體通量估計時，常假設可感熱整體傳送係數和潛熱整體傳送係數相同，但觀測結果顯示兩者大小並不相同。

How to measure and estimate air-sea turbulent fluxes more accurately, including momentum, sensible and latent heat fluxes, is an important topic for air-sea interaction. According to Lo (1993) and adopted Liu et al. (1979) representation of the molecular effect, an algorithms was developed to estimate air-sea turbulent fluxes during TOGA-COARE from 11 November to 26 November 1992. The Observational turbulent fluxes at the same period, evaluated by the turbulent correlation method, of the R/V Hakuho stayed at 0 °N, 156°E were used to compare with results of this new algorithm, as well as the TOGA-COARE algorithm.Results show that the diurnal cycle of the turbulent fluxes were significant. During the daytime the net sea surface heat flux is positive, the ocean was heated, while at nighttime it is negative, the ocean was cooled. A substantial semi-diurnal cycle of turbulent fluxes is also shown. It maybe resulted from the variation of atmospheric fields near sea surface. Both of the latent heat fluxes estimated by the two algorithms are large than these of the observation. While the sensible heat flux estimated by this new algorithm is more similar with observational results than these by TOGA-COARE algorithms.When wind speed is larger than 4 m/s, the turbulent exchange coefficient for wind stress (C□) increases linearly with increasingly wind speed. But the C□ decreases rapidly with increasingly wind speed while wind speed is smaller than 4 m/s. Also, the turbulent exchange coefficients for sensible heat fluxes (C[9264]) and those for latent heat fluxes (C□) estimated from observational data are not the same.