This subject was briefly discussed recently, in connection with the "de-iced prop" thing.
It is more straightforward in the USA, where the FAA will (or won't) certify a plane for flight into a specified category of known icing conditions. Moreover, the FAA has ruled that a
forecast of icing amounts to known icing.
My view is that it's pretty obvious that the FAA was referring to forecasts done by
US weather services in the USA.
It is stretching things to stretch this "forecast icing = known icing" to the rest of the world (as one would do by default, if the plane is N-reg), where one gets local forecasts of widely varying quality, ranging from none at all, to UK's well known permanent (F215) forecast of icing in
all cloud, which is patently ludicrous unless the Met Office has discovered something new about physics of liquids and solids
However, the
CAA also will (or won't) certify a plane for flight into known icing. So, what does the
CAA use to define what constitutes known ice? Or does the CAA simply use the FAA categories? Or do they simply use the FAA test results and rubber stamp them (or not)?
One example which I believe is current is a TB20 with full TKS. Under G this is OK for known ice but under N it isn't. This is a rare example of the CAA being more generous than the FAA - quite reasonably so in this case because the full TB20 TKS is reportedly exceedingly effective.
A known ice certification must involve two things
a) it must be done by reference to specified icing conditions, and
b) since the certification status is a legal thing, none of it makes sense unless there is a specified and
legally referenced category of weather which one can (or can't) fly into. For example, the respective national CAA would state that a particular forecast would constitute known ice if it says so.
The b) bit is missing in the UK and
AFAIK in Europe.