Turbulent fluxes from weight-shift microlight aircraft over transitional grassland of Inner Mongolia

In environmental science spatial representativeness of measurements is a general problem. Often ground-based measurements can be commenced only at the expense of either scale or resolution. Airborne platforms have been shown to overcome this problem. However typical research aircraft are expensive to operate or not even applicable e.g. in remote areas. Unmanned aerial vehicles do not yet allow a comprehensive sensor package due to payload restrictions. Here the Weight-Shift Microlight Aircraft (WSMA) can provide an alternative. Such aircraft was successfully applied to comprehensive Atmospheric Boundary Layer (ABL) soundings over the grassland of the Xilin River Catchment (XRC), Inner Mongolia. The XRC is subject to vast human activity, which influence to capture was objective of a total of 3 month of soundings in the summers of 2006 and 2009. This paper concentrates on the recent development and first application in the XRC of Eddy Covariance (EC) flux measurement from WSMA. After an outline of the WSMA, method development and validation, the research framework in the XRC is introduced. Results for fluxes of sensible heat (Q_H), water vapor (Q_E) and CO₂ (Q_C) are shown and related to simultaneous ground measurements of control sites.

After successfully applying the WSMA to aerosol, radiation transfer and surface mapping studies the feasibility of EC flux measurements in the ABL was explored. The WSMA sensor package comprises the measurement of airframe motion relative to the earth (U_b, from inertial navigation system), air motion relative to the airframe (U_a, from five-hole pressure probe), atmospheric state (pressure, temperature, dew-point temperature), trace gas concentrations (CO₂, H₂O), radiation (actinic, global, surface temperature), infrared surface imaging and aerosol size distribution. Variables entering the EC flux calculation are stored at 10 Hz or 2.6 m horizontal resolution. Due to its flexible wing- and aircraft architecture, WSMA dynamics can be significantly enhanced as compared to fixed wing aircraft (FWA). Most of all the wind vector (U) has to be evaluated carefully, since it is calculated as vector difference U = U_b - U_a from quantities larger by one order of magnitude. Well established algorithms for FWA were adapted and corrected for the WSMAs varying motions. The system uncertainty in wind components (±0.3 ms^-1) and flux measurements (±5 Wm^-2 for Q_H and Q_E) was estimated by propagating sensor and calibration uncertainties. The final algorithm was applied to 50 soundings in the ABL and validated against ground truth from sonic detection and ranging, tall tower EC-, cup- and vane as well as large aperture scintillometer. Here wind components compared to 0.2 ms^-1, fluxes from WSMA differed on average by -2±5 % for Q_H (assuming a linear Vertical Flux Gradient, VFG) and +15±17 % for Q_E (no VFG considered).

In succession 36 WSMA EC flux measurements have been carried out in the XRC to directly capture the regions present energy, water and CO₂ exchange. Simultaneously to the WSMA soundings, half-hourly tower EC fluxes were measured at 3 control sites, two of them non-grazed (C4 dominated Leymus, C3 dominated Stipa) and one heavily grazed (3 sheep unit ha^-1y^-1, HG). The idea of this setup was to a) validate the spatial representativity of process-based regional flux simulation and b) allow for scenario study of advancing grassland degradation. Flight tracks over the heterogeneous terrain were planned with the aid of a geographic information system as a compromise to a) avoid topography associated local circulation, b) minimize flux statistical error and influence of VFG, and c) represent the XRCs land cover proportion. As result WSMA soundings took place around sun apex and followed multiple repetitions of 2 to 80 km long line transects, organized in vertical stacks. Operational results at lowest flight level (50 m, AC50) averaged to Q_H = -177±78 Wm^-2, Q_E = -78±40 Wm^-2 and Q_C = 0.02±0.2 mg CO₂ m^-2s^-1 throughout the campaign. At a radiation budget Q_S* = -412±86 Wm^-2, which is 16 % lower than the control sites, the energy balance from AC50 measurements was closed to 72±18 % on average. AC50 measured Q_H was found comparable in magnitude to the HG control site, whereas Leymus and Stipa sites were 30 % higher on average. The magnitude of Q_E was comparable to the Leymus site, whereas Q_E from Stipa and HG were ≥35 % higher on average. This relation is remarkable, since the soil moisture on HG was actually lower (9 %vol) compared to the non-grazed sites (14 %vol). AC50, Stipa and HG measurements indicate marginal CO₂ release / uptake, only Leymus displayed notable CO₂ assimilation (-0.07±0.3 mg CO₂ m^-2s^-1).

This study proves that EC flux measurement is possible from WSMA with an accuracy comparable to conventional FWA. It shows that the WSMA, which can be easily shipped to and operated from remote areas, provides a suitable tool to close the gap from local to spatial measurements.