The AirDat TAMDAR in-situ multi-function aircraft sensor includes in its measurements an estimation of atmospheric turbulence. Turbulence is reported as an eddy dissipation rate (EDR) and is calculated based on true airspeed (TAS) samples which are calculated from the TAMDAR pitot and static port pressures, and the TAMDAR temperature. The MacCready method is used to estimate the EDR based on the power spectrum of the TAS signal. An EDR is estimated every 3 seconds using a 6 second block of 10.67 Hz TAS data. EDR estimates can be averaged if desired for a smoother result—normally a 9 second average is used. In order to be compatible with accepted standards, the standard TAMDAR EDR reporting is modeled after the NCAR method used on several Delta and Southwest Airlines planes. The report includes the mean and peak EDR, and the time of peak for each one minute period. The TAMDAR data are available to participating airlines for operational use.

A previous NASA sponsored study done by the University of North Dakota (UND) in 2005 used in-situ data, including accelerometer data, gathered from the University of Wyoming's King Air 200 research aircraft. There was good correlation between EDR computed from the TAMDAR data and the King Air reference data; however, there were a relatively large percentage of high EDR outliers with the TAMDAR data. Subsequent testing revealed that noise, probably due to locally generated turbulence, on the TAMDAR static port was a significant contributor. A TAMDAR sensor that addresses the noise issue with hardware and software changes has since been installed on the UND Citation II for evaluation.

The TAMDAR sensor is currently installed on about 170 commercial aircraft. Included in this group are 36 AeroMexico and 56 Chautauqua Embraer ERJ-145s. TAMDAR has an ARINC-429 data bus interface to the avionics system on these planes. This allows the aircraft TAS bus data to be used by the TAMDAR algorithm to calculate an EDR. Since there are no EDR reference data available on these planes, evaluation is based on the expected -5/3 power spectra slope of the Kolmogorov model and histograms of the EDR peak and mean levels.