Determining Evapotranspiration with the Eddy Covariance Method: Fast-Response Dry- and Wet-Bulb Thermocouples for Humidity Measurements Can Provide a Cheap Alternative to Infrared Hygrometers.

Tuesday, 16 December 2014
Friso Holwerda and Maria Susana Alvarado-Barrientos, Universidad Nacional Autonoma de Mexico, Centro de Ciencias de la Atmósfera, Mexico City, Mexico
Field data on evapotranspiration are of crucial importance for ecohydrological and hydrometeorological studies in the tropics. Probably the most direct way to measure evapotranspiration is with the eddy covariance method, in which the latent heat flux (λE) is calculated from turbulent fluctuations of vertical wind velocity and humidity. The humidity fluctuations are typically measured with some type of fast-response infrared hygrometer. However, these sensors are expensive, which can be problematic if research budgets are limited. Turbulent fluctuations of humidity can also be measured with fast-response dry- and wet-bulb thermocouples, which can be constructed easily and at a fraction of the price of infrared sensors. The idea of using dry- and wet-bulb thermocouples for measuring λE with the eddy covariance method is not new, but hasn't been tested recently, possibly because experiments in the late seventies showed that this approach is not without problems due to the slow response of the wet-bulb thermocouple. In the present study, values of λE derived from dry- and wet-bulb thermocouple measurements were compared with those obtained using a fast-response KH20 hygrometer. Measurements were made above a shaded coffee plantation and a sugarcane crop in central Veracruz, Mexico. The agreement between λE obtained with the thermocouples (y) and the hygrometer (x) was very good for both vegetation covers: y = 0.98x + 5.0 (W m-2), r2 = 0.93 (coffee plantation); y = 0.99x – 13.3 (W m-2), r2 = 0.88 (sugarcane). However, the correction factor (CF) for high frequency loss in the wet-bulb temperature signal was considerably higher for the low-statured sugarcane crop (CF = 1.33) as compared to the taller shaded coffee plantation (CF = 1.09). Nevertheless, as long as care is taken in the derivation of this correction factor, reliable λE data can be obtained using the dry- and wet-bulb thermocouples, offering a cheap alternative to infrared hygrometers.