Digital airspeed "peaks"


I’ve been bench-testing my new 3DR Pixhawk digital airspeed sensor.
When plotting the live data (Pixhawk is connected to the MP via USB), I can see a baseline that is between 1 and 3.5m/s, which seems to be normal according to the manual and wiki. However, there are random peaks which all have about the same amplitude and decay behavior. In the screenshot, you can see that there a two identical peaks.
I’ve been testing indoors with EKF enabled…

Is it a good idea to change the EKF_EAS_GATE to reject bad measurements?

Or should it work with the default settings?