The estimation of aerodynamic drag in cycling through road tests has the advantage of considering actual cycling conditions. The main drawback is that its accuracy depends on factors of the testing scenario as the wind speed and the road grade . For that reason, the effect of and on the estimation of drag area () was studied. Numerical and experimental approaches were combined. The numerical approach investigated the sensitivity of to and perturbations. The experimental approach analysed the effect of including or not and on the identification of by comparing the changes in the prediction of power delivered. It was found that disregarding small values of (e.g. 0.5 m/s) and (e.g. gradient of 0.05%) leads to errors in the estimation of of around 10%, referred to the actual value. It was also obtained that the average error of the power prediction when considering and for the identification of the parameters is about 4.4% and about 25.5% when and are neglected. It is concluded that including and data reduces the error on the identification of through outdoor road experiments.
Keywords: Aerodynamics; anemometry; bicycle; resistive forces.