The GFZ CHAllenging Minisatellite Payload (CHAMP) measured significantly higher cross-track wind speeds in the thermosphere, almost twice the magnitude, than zonal winds observed by the UCL Fabry-Pérot interferometers (FPIs) situated in Kiruna and Svalbard, and simulated in the UCL Coupled Middle Atmosphere and Thermosphere general circulation model (CMAT2) in the period 2000-2010. Equipotential zonal wind contours are produced in CMAT2 from 240km, the approximate height of the 630.0nm atomic oxygen red emission line observed by the FPIs, upwards; thus it follows their extrapolation to the orbiting altitude of CHAMP (~325km in December 2008) dictate similar zonal wind magnitudes to both the FPI measurements and CMAT2 generated values. Here we investigate the viscosity within CMAT2 as a possible determinant of the equipotential contours and thus contributor to the underestimation of zonal winds, namely by varying the coefficients of viscosity: molecular and turbulent. The latter is investigated by altering the turbulent Prandtl number (where Pr = 0.7, 2, 100), the former its dependence on temperature. We confirm the values of the coefficients necessary to create these high winds alone as physically unlikely, although acknowledge their potential in having some small contribution. Other possible sources of error are also briefly discussed including the chosen CHAMP drag parameter and the reliability of the FPI data. This study focuses on a 10-day period from the 9th-19th December 2008, at solar minimum.