What is the effect of connecting the tip of an oscilloscope probe to a ground lead? Here is one of the few links that I've found that shows the set-up: The Shorted Scope Probe. What are the possible uses of such set-up and what is the cause for the behaviour one might observe when shorting the probe (reduced susceptibility to environment noise, etc.)?
I always do this when measuring any fast signals or when the circuit is noisy. Putting the tip to the ground clip should give you zero volts to the scope, but often it doesn't. That is because the potential at the ground point of the scope can be moving relative to earth (or the local reference point of the supply feeding the scope). Even if your circuit is "floating" there will be capacitance to ground. As rapid transient occur at the ground clip point, currents will flow in the probe ground wire. This will in turn generate voltage transients between the ground clip and the actual scope ground. These can be measured by placing the tip onto the ground clip. Any waveform that you observe when doing this will be superimposed on the real signal that you are trying to measure when you connect the tip to some point varying with respect to the ground clip.
If you see high speed transients when using a scope probe, try this test to check if they are due to this affect. This also applies to differential probes, connect the +ve probe to the same point as the -ve one to check that the differential probe is rejecting all the common mode signal.
By forming a loop with the ground lead, you form an antenna. This can be useful for observing EMI issues with a circuit, or for triggering on an EMI spike. I have used a scope probe connected as shown to trigger on an ESD test pulse (connecting directly to the DUT is risky, as the ESD could damage the scope). Additionally, you can use the probe to find an area of a circuit that that is radiating excessively (set scope to show FFT, wave the probe around over the circuit).