AmeriFlux Meeting

Day 2

Gaby Katul presented preliminary results from the Duke study on spatial variability of turbulent fluxes. Overall, there appeared to be planar homogeneity among velocity statistics (u^, w^, and u^*), momentum and scalar fluxes (H>LE>CO2), particularly during daylight hours -- suggesting reasonably good representation by a single point turbulence measurement in the uniform pine stand.

In response to earlier concerns about representativeness of single point measurements, Hans Peter Schmid proposed a quantitative approach to assessing spatial heterogeneity. Representativeness of a variable (e.g. turbulent, scalar, ecological) could be determined by comparison with a surrogate variable for which the spatial variability is known - for example a variable determined with remote sensing data.

Dennis Baldocchi discussed the measurement of subcanopy fluxes. The viability of subcanopy fluxes can be assessed, as with above canopy fluxes, though closure of the energy balance. However, because of the high spatial variability below the canopy, this approach is particularly sensitive to the spatial coverage of the net radiation measurements. Employing a moving radiometer (e.g. on a tram), reasonable energy balance closure can be observed. The appropriate length of a tram will depend upon the spatial variability of the subcanopy environment at each site (i.e. longer for open canopy, shorter for closed).

Michael Unsworth addressed the issue of nocturnal flux measurements. Under stable conditions (e.g. u^* <0.15 m/s) the storage CO₂ flux may provide an estimate of the nightime respiration rate. Under windy conditions, above and below canopy eddy covariance should provide reasonable flux densities

Ken Davis discussed the interaction between the surface and planetary boundary layer and related research at the WLEF tower in Wisconsin.

Dave Bowling reviewed the essentials of flux measurement using relaxed eddy accumulation (REA). While this technique appears quite promising for some compounds such as isoprene and non-methane hydrocarbons as well as CO₂ and H₂O, sampling requirements for many others quantities (e.g. isotopes) remain difficult.

Joe Berry discussed data requirements for ecosystem modeling using examples from SiB2 application at FIFE. It was strongly emphasized that continuous meteorological data is crucial to timely and successful model development and application. Individual sites can help by providing continuous records or contacts and references of other "backup" data sources that would facilitate filling the gaps.

Bill Munger and Kevin Tu discussed software intercomparison using the EUROFLUX "Gold" file. Preliminary results from the Harvard and Howland groups were presented. Good agreement was found at Level 1 (covariances) and Level 2 (flux densities) despite the use of linear (Harvard) and a recursive filter (Howland) detrending methods. Nearly all of the differences were attributed to Level 1 processing. It was suggested (Peter Bakwin) and subsequently agreed that all groups in the future should use a common lag (2.5s for CO₂, 3.0s for H₂O) for intercomparison of the EUROFLUX "Gold" file.

Representatives from LiCor, Campbell and ATI led question/answer sessions on relevant products, instruments and sensors. According to Dayle McDermitt, the difference in lags between CO₂ and H₂O measurements (CO₂ lag < H₂O lag) is due to adsorption/desorption of water on tube walls - this difference is apparently not due to electronics. A simple check of this would be to see if lag times change with temperature (higher temp -> lower retention time).

Ken Clark and Dave Hollinger discussed methods of lightning protection. A wealth of information on this topic is available on the web. The Florida group isolates instruments and cable from the tower every several meters using wood or plastic mounts. The importance of ensuring that the grounding rods are truly grounded (i.e. deep enough) was stressed. The suggestion was made to test this by measuring the resistance across two rods. Also, the use of RS-232 data stream "in line" surge suppressers can minimize any electrical damage.

The meeting was adjourned at 4:00 October 29, 1997.