No easy job
YMMV, but figuring out how to decode the GRIB2 files was probably faster for me than figuring out how to order an ipad 8-)
hence our app and a cheaper investment being buying an iPad for it.
But I still don't see how I could then use the winds aloft data by software, as opposed to staring at it on the screen
What do hou want The software to calculatie?
I have a list of say up to 800000 legs, each with start and end coordinates, and flight level. For each of those legs, I need the "average wind", or the midpoint wind.
Since yesterday evening I do this from raw GRIB2 files - it still needs some polishing, but is basically there.
The downside is that the complete GRIB2 dataset is over 3GB; therefore, it would be convenient to have a web service that could provide the GFS winds aloft (and temperature and QFF) forecast for a "box" of air defined by eastern and western longitude, northern and southern latitude, and top and bottom flight level, for some date and time.
(theoretically, I should adjust the timing based on how far away that leg is from the departure - but for the typical 3 hours flight, I don't think this is necessary, as the GFS sampling frequency is 1/3h only, and if the weather changes that quickly, the forecast is likely rather imprecise).
We have such a webservice/API, but don't advertise with it. It will calculate the average wind per route segment as predicted in the GFS model. Add to this things like a factor for a climb gradient, the aircraft performance, etc. and ... you have a pretty accurate estimation of Total EET but also of times until FIR boundaries.
What do you need the 800.000 legs calculated for?
tomjnx,
I think you are over working the problem. In real life, the winds at any given point and are going to be +/- 20 degrees of heading and +/-20% of forecast magnitude. Grab a point every degree of lat and lon at the GFS pressure levels (say 1000 700 500 300 200) and the GFS times and then interpolate for each leg.
As an aside, you seem to be using a lot of the information from A Boot Strap Approach to Performance. When I have used that every thing worked well accept the propeller power curves (their model for constant speed props came out way different than the propeller map for a Bonanza and hence underestimates (massively) the higher power cruise speeds (and has corresponding errors in where Vy). What are your experiences?
Grab a point every degree of lat and lon at the GFS pressure levels and the GFS times and then interpolate for each leg.
That's exactly what I do. I download all GFS layers, as they come bundled in one GRIB2 file whose structure doesn't easily allow for example cherry picking layers from a byte serving httpd. So I have way more pressure levels than just the 5 you mentioned.
The whole modelling exercise gave me quite a few insights, but the results are not accurate enough to use them for planning. Some aspects are remarkably accurate (glide performance, for example, and interestingly the prop angles match the TCDS values), but others can be pretty far off (cruise speed is up to 20MPH underestimated).
My experience is the same as yours. The prop model I have is not adequate. I even measured the activity factor of my prop, but no matter whether I entered the measured value or twice into the prop model, the overall results didn't match the POH well enough.
The POH is pretty accurate, I've measured a few cruise speed points, then corrected for the atmosphere, and the result was within 1kt or so of the POH values.
For planning, I currently use polynomials fitted to the POH values.
AeroPlus, I’m trying to file a VFR flightplan for the first time using your Flightplan iPhone app. I’m getting the following message all the time when going to file the fp after creating and saving it:
Flightplan error
Flightplan could not be sent!
Error code 400
I’ve tried to find this error code in the web with no success. What does it means? Which is the problem?
I have been using this program to file VFR FPL over the last couple of years without any problems but know I got an error message stating:
“Please sign up for a trial subscription or subscribe”.
Anyone else using this app ? Do you have the same issue ? Or this the app dead ?
@aeroplus does drop in occassionally.
Gosh, this thread is almost a decade old! We were all younger then 
And many of the pilots active then are no longer flying.
Mmflynn wrote the original FlightPlanPro which was incorporated into what is now Rocketroute.
Tomjnx wrote a unix executable for IFR route generation which was incorporated into what is now the Autorouter.
