The CSAT3 should not be used to measure absolute sonic temperature. Aside from the effects of water vapor in the sonic path, the speed of sound measurement is highly dependent on the ability of the CSAT3 to maintain its geometry. A 0.2 mm change in the sonic path is equal to a 1°C change in the measured absolute sonic temperature at 25°C. This is equal to a 0.33% error in the absolute sonic temperature. The wind speed measurements are also dependent on the sonic path distance, but to a lesser degree.