Intercomparison of lower-cost and conventional eddy covariance systems for CO2 and H2O flux measurements above cropland monoculture and agroforestry

Agroforestry (AF) systems are recognized as a more sustainable agricultural practice compared to conventional agriculture due to its potential for increase carbon sequestration, among others. Short Rotation Alley Cropping (SRAC) is an AF practice in Central Europe in which trees and crops are cultivated in alternating rows.

The amount of carbon sequestered by a SRAC system can be estimated by the eddy covariance (EC) technique, the standard method for the continuous assessment of energy, momentum and trace gas exchanges above terrestrial ecosystems. As SRAC systems are heterogeneous, using only one EC set-up might limit the spatial representativity and, hence, the statistical power of measured fluxes. Increasing the number of EC set-ups could increase the statistical power, which is, however, cost intensive.

Therefore, the aim of this study was to test (i) the performance of a lower-cost EC (LC-EC) set-up for CO₂- and H₂O-flux measurements above SRAC and monocropping (MC) agriculture and (ii) if the sensor-to-sensor differences in fluxes are lower than differences between ecosystems (SRAC and MC).

We performed CO₂ and H₂O flux measurements above a MC system with three lower-cost and one conventional EC set-up from March to August 2022. In addition, CO₂ and H₂O fluxes were also measured with a LC-EC set-up located in a SRAC system at 520 m distance from the MC site.

The CO₂ and latent heat (LE) fluxes of the three LC-EC set-ups showed similar results compared to the conventional EC setup. The linear regression between the conventional and the LC systems showed R² coefficients in the range of 0.8-0.9 for CO₂ and 0.7-0.9 for LE, and slopes in the range of 0.9-1.0 for CO₂ and 0.8-1.0 for LE. The energy balance was consistent for all the systems, providing an average 70% closure. The total cumulative C uptake over the entire campaign was similar among the 3 LC-EC set-ups, but they underestimated the C uptake compared to the conventional EC set-up by 18% on average. The C uptake measured by the LC-EC systems at the end of the measuring campaign was 74 g C·m^-2 at the MC (mean across all the 3 LC-EC set-ups) and 111 g C·m^-2 at the SRAC. The C uptake of the conventional EC system at the MC was 90 g C·m^-2. Hence, the SRAC system had a larger C uptake than the MC system throughout the measurement campaign.

We conclude that the LC-EC provided satisfying results compared to conventional EC, with the potential to improve the spatial replication of EC measurements. Furthermore, the difference between the 3 LC-EC set-ups in the MC was much lower compared to the difference between the MC and the SRAC.