Maize canopies under two soil water regimes: IV. Validity of Bowen ratio–energy balance technique for measuring water vapor and carbon dioxide fluxes at 5-min intervals

Abstract

The feasibility of determining at 5-min intervals the fluxes of water vapor and CO 2 using a Bowen ratio/energy balance/CO 2 gradient (BREB+) apparatus was studied. Details of the BREB+ apparatus are given. The results obtained over maize canopies at the short 5-min signal averaging time are examined in several aspects. One aspect is the fact that on days when net radiation ( R n) changed rapidly because of cloud movement, latent heat flux ( λE) fluctuated nearly in unison at 5-min intervals with the fluctuations in independently measured R n, Bowen ratio, and canopy temperature. The direction of the fluctuations is consistent with the principles of energy balance and gradient transport. Hence, the results support the use of the short signal averaging time. Another aspect is the coincidental fluctuations in the measured differential in water vapor (Δ q) and in carbon dioxide (ΔCO 2) at two heights above the canopy on clear days when the trend in radiation was steady, which appeared to be the result of measurement noise at first glance. It turned out that Δ q and ΔCO 2 fluctuated largely in unison, indicating the likelihood of real accumulation of water vapor and depletion of CO 2 near the top of the canopy at times of low turbulence, and the opposite effects of reduced accumulation and depletion at times of high turbulence. The calculated values of eddy transfer coefficient for water vapor, K w, is consistent with this interpretation. In spite of the fluctuations in Δ q, the time trend in λ E was steady on clear days. In contrast, the time trend in the 5-min canopy CO 2 flux ( A cf) often exhibited substantial short term peaks and dips. Parts of these fluctuations in A cf may be attributed to the measurement noise in ΔCO 2 and the inherent larger errors in calculating A cf. The error contributed by the changes in CO 2 and water vapor stored within the canopy air volume from one measurement interval to the next were evaluated. The limited results show that the errors due to storage were largely minor and were similar in relative size for λ E and for A cf. The 5-min mean data were compared with the 30-min mean data to evaluate whether the shorter averaging time is sufficient to make horizontal divergence of energy components insignificant. The comparison indicates negligible mean horizontal movements in λE and sensible heat H and demonstrates the fulfillment of the assumption of one-dimensional transport underlying the BREB+ technique. On days of variable clouds, it is shown that the 30-min mean values distorted or obliterated the expected responses of the canopy to sudden changes in radiation. Some limitations and other inaccuracies of the BREB+ technique are discussed. Overall, the results are strongly supportive of the validity of the BREB+ technique in general, and the use of the novel averaging time of 5 min for λ E and H over field crops in particular. The necessary averaging time for A cf is less certain and further study is desirable.