Effect of open-path gas analyzer wetness on eddy covariance flux measurements: A proposed solution

Abstract

Open-path gas analyzers are popular in eddy covariance flux measurements of trace gasses (i.e. CO2). The quality of the data, however, may be influenced by several factors. Exposure in an outdoor environment invariably causes the instrument to become colder or warmer than the air temperature. Instruments with internal temperature regulation and/or from heat generated by active electrical components can also influence the sensor temperature. In addition, sensors can have condensation problems on their optical windows thus affecting the quality of the measurement. Unreasonable measurements have been widely discussed, especially in moist, high-latitude regions. As this is a very important research problem facing flux studies, we examined how wetness (dew and raindrops) on the surface of the focus lens of the popular LI-COR LI-7500 infrared gas analyzer may affect flux measurements from the open-path eddy-covariance system. Field experiments showed that additional sensor heating may inhibit dew formation yet greatly improve the quality of flux measurements. A detailed energy balance approach was used to model the gas analyzer window temperature under environmental conditions and dew effect through a pair of LI-COR LI-7500, with and without heat treatment, in a grassland ecosystem in the Netherlands. With the proposed model, existing datasets can be filtered for dew events. Data from three different flux measurement sites were then used to assess the magnitude of dew effects on longer time-scales; 2 years from the Netherlands and 3 weeks of data from an arid coastal desert. About 30% of the measurements were affected by dew in the grassland area versus 4% in the arid region during the dry season. Sensor heating suppresses dew formation but might lead to errors in trace gas fluxes evaluated over long periods, thus we analyzed how sensor heating or cooling affects trace gas flux measurements. Additions to a recent (2006) correction and application to a horizontally and vertically oriented LI-COR LI-7500 are presented as they deal with sensor heating problems in eddy-covariance systems. The sensor energy balance model, together with the proposed modified sensor heating corrections, were used to estimate sensor temperature effects on long-term scale CO2 flux measurements and showed that additional heating does affect the turbulent trace gas CO2 fluxes but is very minor, especially for a horizontally mounted LI-COR LI-7500 gas analyzer. Further efforts are urgently needed to improve the data quality and quality of flux measurements. © 2008 Elsevier B.V. All rights reserved.