Investigating Coordinate Rotation Methods in Non-Ideal Agricultural Fields for reliable Carbon Dioxide Flux Estimates

Abstract:

The Eddy Covariance (EC) remains a popular method to measure fluxes of water, trace gases, and energy fluxes. Determining a suitable reference frame is necessary to ensure the vertical velocity is not affected by large mean flows. There are two popular methods to achieve this, either double rotation or planar fit. When measuring over a dynamic canopy, with a sloping terrain, the difficulties in determining the reference frame is compounded. Models for sector-wise planar fit, the single plan planar fit as well as a continuous planar fit were fitted over the entire data set to determine the influence of surface cover and stability regimes. This method was carried out using model-based recursive partitioning. Recursive partitioning allows for modeling nonlinear relationships, automated detection of interactions among the explanatory variables, as well as providing an easily interpretable segmented model. Momentum, heat, and flux of carbon dioxide and water vapor fluxes were then calculated to compare the viability of the differing methods. The study found multiple methods both in the processing of the EC data and post-processing procedures to identify sources of vertical velocity bias and potential workflows to overcome it. The enumerated results and workflows will have a direct impact to provide farmers and scientists reliable estimates on agricultural water use and carbon dioxide flux in a changing climate.