ObservingCapability VS Sampling Procedure and/or the observingProcedure

[sgrellet]

Do we describe the Sampling Procedure and/or the observingProcedure?
=> May be worthwhile setting up a “virtual sample” for the ObsCaps, including the planned Sampling Procedure
At the WQ station level you have a notion of which procedures are authorized for a given observable property but one cannot anticipate the procedure that will be implemented by the Lab
=> How do we deal with lab analyses where we don’t know the procedure before the measurement is performed? Is this still one Datastream?
=> Can we set up a Datastream with different Procedures? Provide details in the Observation parameter?

[ksonda]

Imo datastream/sensor should be defined by things known a priori (eg observedproperty and sampling method at a high level, and then the realized specific sampling procedures and observing procedures could be encoded in parameters at the observation level