Of these, the RocketRoute pack is usually the most conservative
A very mild statement. One could also say their icing prediction tool is total BS…
I don’t know how anybody can forecast icing conditions in any way useful to non-deiced light GA.
If you could forecast the vertical extent of IMC i.e. cloud (and nobody has yet achieved this – for stratus cloud – with anything you could call “reliability”) then knowing temperatures aloft (which is fairly reliably forecasted) you could plot out regions of probable icing.
Otherwise, icing forecasts seem to all just forecast where convective activity is expected, and then throwing in the (fairly well forecasteable) temperatures aloft gives you an icing forecast in convective weather. The problem with that is that those convective regions would be obvious to a blind man feeling an MSLP chart made of clay – this one would do nicely
My worst by far icing encounter was in stratus cloud, base 1500ft, tops 4000ft, smooth bottom and a smooth top, April, SE UK, and at 3200ft (the EGMD ILS platform altitude 
) I collected 3cm in about 5 minutes… Obviously I had to break off the ILS, full power (TB20 with a TKS de-iced prop), Vs was about 100-110kt, and could not maintain level flight until I got down to about 2000ft. +1C was found about 2000ft and half an hour later the stuff all came off. Those conditions (NS cloud I guess) are pretty rare otherwise we would all be dead, but it illustrates how careful you have to be. Forecasting solid IMC between 0C and say -15C would be a good start but currently nobody can do even that.
My worst by far icing encounter was in stratus cloud, base 1500ft, tops 4000ft, smooth bottom and a smooth top, April, SE UK, and at 3200ft (the EGMD ILS platform altitude ) I collected 3cm in about 5 minutes
This is even quite common. The worst ice we pick up is usually when flying level to intercept the ILS just as Peter describes it.
Otherwise, icing forecasts seem to all just forecast where convective activity is expected, and then throwing in the (fairly well forecasteable) temperatures aloft gives you an icing forecast in convective weather.
I don’t have sufficient recent experience to vouch for its accuracy, but the Swedish MET office does not limit its ice forecasts to convective weather. Far from it.
That “probably some low cloud” thing was actually from ~3000ft AMSL to FL200, and dropped 30cm of snow
Otherwise, icing forecasts seem to all just forecast where convective activity is expected
This is not true for DWD ADWICE, they even tell you what type of icing (convective, stratiform or freezing) they predict. That said, they always predict light stratiform icing wherever the temperature is in the right range and the atmosphere is not totally dry, so IMO pretty useless and I ignore that.
Right now, pretty much all icing they forecast for today in central to north europe is of the stratiform type.
tomjnx:
This is not true for DWD ADWICE, they even tell you what type of icing (convective, stratiform or freezing) they predict. That said, they always predict light stratiform icing wherever the temperature is in the right range and the atmosphere is not totally dry, so IMO pretty useless and I ignore that.
That’s what I assumed.
In the GRAMET what is the criteria to depict an area as being an icing area?
In the GRAMET what is the criteria to depict an area as being an icing area?
Most of jets that I have flown which have ‘anti ice’ (as opposed to ‘de ice’) specify that it should be activated between +10 C and -30 C when in ‘visible moisture’.
It’s also worth making sure that your OAT probe is accurate.
For some reason best known to Cirrus, there are two probes on my SR22 G2 which read anything from 2 to 5 deg C apart. The fact that one of them is mounted on the side of the cowling above the exhausts may have something to do with it.
Edit: I missed the last bit of Peter’s post, but would confirm he’s right!
