New methodology to subtract noise effects from turbulence parameters computed from ADV velocity signals

The Acoustic Doppler Velocimeter (ADV) signals are affected by Doppler noise which is intrinsic to the Doppler measurement technique. It has the characteristics of white noise and its integral effects must be subtracted from some turbulence parameters. New tools are introduced in this paper to evaluate the relative importance of the noise energy on the total energy as well as to define the characteristic frequency in the measured energy spectrum where noise energy is more important than turbulent energy (flat plateau). To develop these tools a model for the power spectrum is adopted which allows the analysis of a range of flow conditions. Results show that noise energy can have an important contribution on the total measured energy, mainly in low energy flows. However in those cases, the noise energy level can be defined from the spectrum because the characteristic frequency is smaller than the Nyquist frequency. After the noise energy level is defined, corrections to the rest of the turbulent parameters (length and time scale, convective velocity, etc) must be performed.