Peter wrote:
What is not being mentioned is the cost of satellite data.
For my trip to Norway in the end of August took out a 28 day trip kit for 39 € and on top I paid another 35 € for data traffic (excl VAT) When you perform auto updates, you can count between 1,50€ – 2 € every 15 minutes depending on the number of packets required, which depends on the quality of the downloaded image. I had my share of poor weather from Malmø to Kristiansand and from Ålesund to Copenhagen, as well as Copenhagen to Köln 
For my Austria trip around Christmas and New Year I took a 2 week trip kit 25 € and 6 € data traffic. Weather was pretty good and I never used the auto-update.
I find this very good value.
There is a good online course called “IFR Insights: Cockpit Weather” available for free (after registration) on the website of the AOPA Air Safety Institute. It contains a good explanation of the limits of datalink weather.
This article matches my experience.
Peter wrote:
Charts and wx models are no good for showing where convective activity will actually happen.
No, but they show where the atmosphere is unstable, so convection can develop…
I’ve hacked code some months ago that colours the routers GFS derived Skew-T diagrams red where the atmosphere is unstable, and green where inhibiting layers exist – I’m curious how that works out in practice, let’s see during the next thunderstorm season…
Apparently, radial spread is still a big issue with current GA systems, if the WX-500 Users Guide is to be believed. Page 4-2 states:
These phenomena are examples of radial spread. Discharge points in radial spread do not necessarily indicate the exact location of atmospheric electrical discharges. To counteract radial spread, L-3 Avionics Systems applied its extensive research in lightning detection to develop enhanced lightning positioning algorithms. These algorithms (used only in the cell display mode) greatly reduce radial spread and improve the depiction of thunderstorms on the display.
So to put it bluntly, the measurements are crap, but their “super secret sauce” turns it into gold.
Except that this cannot work, the Data Processing Theorem states that you can only loose information by processing it, therefore, while such a clustering algorithm can make the picture more pretty, it’s even more fiction than before…
The remainder of chapter 4 is instructive, the right side shows the actual measurements (strike mode), the left side shows what their “secret sauce” makes out of it (cell mode). Looking at the strike mode pictures, that factor of 25 spread I claimed actually looks quite realistic.
You may be reading the manual, Tom, but I have been flying with a WX500 since 2002, about 1700 airborne hours, IFR since 2005, and I can you that isn’t true.
Peter wrote:
You may be reading the manual
Guilty as charged, I know real men don’t read manuals, and manufacturers are known to portray their products badly in their Marcom materials 
So why don’t you post a few photos of the device in strike mode next time you fly in convective weather?
I can’t count the times I have flown thru rain where ATC has been recommending a reroute, but I have followed the Stormscope advice. After I get thru and am asked about the ride. which is usually wet and smooth, all the big radar equipped aircraft twins follow me straight thru.
Tomjnx, you can believe what you want, but my 35 years of experience using this is what I use as guidance.
tomjnx wrote:
Except that this cannot work, the Data Processing Theorem states that you can only loose information by processing it, therefore, while such a clustering algorithm can make the picture more pretty, it’s even more fiction than before…
Sure it can work. We’re talking about the amount of information, not the quality. If your interpretation were true, then no process which “refines” data would work.
This has turned into a Stormscope thread, may be worth splitting?
A few questions on Stormscope:
Is it better at detecting early convective activity? Weather radar by definition picks up cells better in the mature phase of TS.
Robert Buck uses 20 miles as an ideal avoiding distance for TS with weather radar (he has different distances depending on OAT and altitude), would you use more with a Stormscope?
Am I correct you also get strikes from high tension pylons/lines?
I think 20 miles upwind avoiding distance is a good rule of thumb, which sort of implies a forecast for scattered area TS, as against isolated, might give pause for thought.
Airborne_Again wrote:
If your interpretation were true, then no process which “refines” data would work.
“Data refining” is usually interpolation or extrapolation, so it’s just a guess. It can be rough or it can be better and more realistic if it’s based on some external knowledge.
