High-Resolution Vertical Profile Measurements for Carbon Dioxide and Water Vapour Concentrations Within and Above Crop Canopies

Abstract

We present a portable elevator-based facility for measuring $\hbox {CO}_{2}$, water vapour, temperature and wind-speed profiles between the soil surface and the atmospheric surface layer above crop canopies. The end of a tube connected to a closed-path gas analyzer is continuously moved up and down over the profile range (in our case, approximately 2 m) while concentrations are logged at a frequency of $20 \hbox { s}^{-1}$. Using campaign measurements in winter wheat, winter barley and a catch crop mixture (spring 2015 to autumn 2016) during different stages of crop development and different times of the day, we demonstrate a simple approach to correct for time lags, and the resulting profiles of 30-min mean mole fractions of $\hbox {CO}_{2}$ and $\hbox {H}_{2}\hbox {O}$ over height increments of 0.025 m. The profiles clearly show the effects of soil respiration and photosynthetic carbon assimilation, varying both during the diurnal cycle and during the growing season. Profiles of temperature and wind speed are based on a ventilated finewire thermocouple and a hot-wire anemometer, respectively. Measurements over bare soil and a short plant canopy were analyzed in the framework of Monin–Obukhov similarity theory to check the validity of the measurements and raw-data-processing approach. Derived fluxes of $\hbox {CO}_{2}$, latent and sensible heat and momentum show good agreement with eddy-covariance measurements.

Appendices

Appendix 1

In 2015, the soil heat flux, temperature and moisture measurements (Sect. 2.2) were performed in a single location near the eddy-covariance station in representatively managed soil. In July 2016, this array was uninstalled due to cultivation, but installation of an ICOS-compliant (Op de Beeck et al. 2015) distributed array of five such locations was started. Until after repeated cultivation and seeding (September 2016), however, only the emergency plot of this array, directly next to the eddy-covariance station on unmanaged soil, was available. Heat-flux plates were installed at a depth of 0.08 m in the pre-ICOS set-up and of 0.05 m in the ICOS-compliant set-up.

The two sampling tubes for the moving and the fixed $\chi _{\hbox {CO}_{2}}$ and $\chi _{\hbox {H}_{2}\hbox {O}}$ measurements (Sect. 2.3) are of the same length to assure identical time lags. However, this length was changed from 3 to 4.5 m before 31 May 2016 to allow for a longer conduit for large canopy heights and a larger tolerance radius for setting up the analyzer. The extension ensured that the tube end dipped into the plant canopy at a horizontal distance of 0.3 m from all other installations and 1.1 m below the carriage, preventing the carriage itself from dipping into the canopy, and thus minimizing mechanical stress.

In 2016, the tubes were equipped with an optional heating system to prevent condensation of moist air on the inner surface of the tubing, particularly during nighttime conditions, with a 2.5-m long heating wire (bed heater for aquarium, Eco-Line ThermoTronic 5 Watt, Dennerle GmbH, Germany) wrapped helically around the first 1.2 m of both inlets tubes, insulated over the entire length of the tube by insulating hoses with an insulation thickness of 0.013 m, and covered by self-adhesive aluminum tape. The signals of the wind sensors were logged on the same file as the gas concentrations via the auxiliary ports of the LI-7000. The thermocouple temperatures were logged at intervals of 0.05 s to a logger (CR1000, Campbell Scientific, Inc., Logan, Utah, USA).

The final wind and temperature set-up, used in 2016, was the result of stepwise improvements to a preliminary set-up during 2015. Initially no fixed-height anemometer existed, and the moving thermocouple was operated in the open and without ventilation, shielded only by a wire mesh, which however failed to secure the delicate thermocouple junction for more than a few hours. As a result, temperature measurements during the first year are partly missing and were partly performed with an improvised repair, where the more rugged compensation lines of the thermocouple were directly connected to each other. These temperature data, however, were not used. Profiles of $\chi _{\hbox {CO}_{2}}$ and $\chi _{\hbox {H}_{2}\hbox {O}}$ are used during this time because they are not affected by missing temperature measurements. The most important changes to the set-up are indicated in Table 1.

Appendix 2

Flux derivation from surface-layer profiles is based on the integrated flux-profile relations for momentum, heat and mass,

$$\begin{aligned} \frac{u}{u_{*}}= & {} \frac{1}{\kappa }\left[ \ln \frac{z-d}{z_{0}}-\psi _{m}\left( \frac{z-d}{L}\right) \right] , \end{aligned}$$

(9)

$$\begin{aligned} \frac{\theta -\theta _{0}}{\theta _{*}}= & {} \frac{1}{\kappa }\left[ \ln \frac{z}{z_{0\theta }}-\psi _{h}\left( \frac{z-d}{L}\right) \right] , \end{aligned}$$

(10)

and

$$\begin{aligned} \frac{X-X_{0}}{X*}=\frac{1}{\kappa }\left[ \ln \frac{z}{z_{0\theta }}-\psi _{h}\left( \frac{z-d}{L}\right) \right] , \end{aligned}$$

(11)

where $u_{*}$ is the friction velocity, $\kappa =0.4$ is the von Karman constant, $z_{0}$ and d are the aerodynamic roughness length and displacement height, $z_{0\theta }$ is the scalar roughness length, and L is the Obukhov length,

$$\begin{aligned} L=-\frac{u_{*}^{3}}{\kappa \,\frac{g}{\theta }\,\frac{H}{\rho _{air}\,c_{p}}}, \end{aligned}$$

(12)

with the acceleration due to gravity g, $\theta$ is potential temperature, H is the sensible heat flux, $\rho _{air}$ is the density of air and $c_{p}$ is the specific heat at constant pressure. Potential temperature was computed by applying an adiabatic lapse rate, based on the 30-min mean temperature and pressure, such that the 2-m-a.s.l. level served as a reference. The largest deviations from air temperature, occurring thus at the surface, were $0.02\,^{\circ }\hbox {C}$, and the effects on computed fluxes were $\le \, 0.2 \hbox { W m}^{-2}$. $\theta _{0}$ and $X_{0}$ are the potential temperature or the fractional concentration by mass of the scalar X at $z-d=z_{0\theta }$; and $\theta _{*}$ and $X_{*}$ are the scaling parameters for the temperature and a concentration X, expressed by

$$\begin{aligned} \theta _{*}= & {} \frac{-H}{c_{p}\,\rho \,u_{*}}, \end{aligned}$$

(13)

$$\begin{aligned} X_{*}= & {} \frac{-F_{X}}{\rho \,u_{*}}. \end{aligned}$$

(14)

The stability corrections required in Eqs. (9) and (10) (integrated form universal functions) for momentum exchange $\psi _{m}$ and the exchange of sensible heat $\psi _{h}$ after Businger et al. (1971) are used in the modified version after Högström (1988). The universal function for the exchange of sensible heat $\psi _{h}$ is also used in the profile (Eq. 11) for the calculation of moisture exchange and for the exchange of trace gases like $\hbox {CO}_{2}$ (Panofsky and Dutton 1984).