Jump to content

Serious issue with AS 2012 winds in PFPX


Recommended Posts

The problem here, as I see it, isn't one of PFPX depicting bad winds. The data used by ActiveSky and PFPX is the same when PFPX is set to use the current_weather_snapshot.txt file. We have to remember, though, that the weather at any individual spot other than directly over a wx station is interpolated from a set of nearby data points, and there is no industry standard way of doing that interpolation. Some use the nearest reporting point, some average the three or four closest stations, and some have an even more complex method of weighting nearby datapoints more heavily than those further away. None of them is "wrong" per se, but they are different.

So the most likely way the PFPX author can produce flight planning products that will agree with the weather injected into the sim by ActiveSky during the flight is to understand and replicate the interpolation method incorporated into ActiveSky by its programmer(s). This is not purely a PFPX issue--FSBuild, which has been my planning program of choice for many years, also demonstrates sometimes significant differences between winds forecast at a given waypoint in its flight plans and the actual winds injected by AS2012 at that waypoint.

In the real world, variances between forecast and actual winds--sometimes even surprisingly large ones--can and do occur as well, hence the need to carry reserve fuel, especially on longer flights over water and through lightly populated airspace where wind reporting and forecasting is more sparse.

Ideally, I'd hope that the PFPX author might get some guidance from Damian and company at HiFiSim as to how to better match the ActiveSky interpolation in PFPX. Until then, I'll be packing some extra virtual gas as I would on a loooong overwater leg in the real world.

Regards

Bob Scott

Link to comment
Share on other sites

  • Replies 81
  • Created
  • Last Reply

I've noticed today with a small flight a big difference from as2012 winds and PFPX with as2012 winds data read. So i've tried to use the online default wheater from PFPX and Activesky 2012 for fsx. The difference is minimal so for now the better solution is to uso default weather provided by PFPX server to compute flight and not use As2012 wheater import function so the result is good.

Link to comment
Share on other sites

Ideally, I'd hope that the PFPX author might get some guidance from Damian and company at HiFiSim as to how to better match the ActiveSky interpolation in PFPX. Until then, I'll be packing some extra virtual gas as I would on a loooong overwater leg in the real world.

Regards

Bob Scott

Hi Bob.

I think what you may have missed here is that PFPX is retrieving wind data properly from ASv6.5, just not the newer versions, and that in those newer versions of AS, there was indeed a change in how the data is formatted in the snapshot file.

I totally agree with you otherwise, regarding the effects of interpolation, etc., but that's not quite what we're dealing with here. It may certainly be a factor, but it would be a minor one, if one at all. AS uses ~9000 weather stations across the globe casting a pretty large net, and in any case, from what I am seeing, the data NEVER matches, even for known physical stations. It's typically not even close.

I posted most of this above already, but as it seems we now live in a TL;DR society...

TL;DR - PFPX is not pulling wx data from newer versions of ActiveSky properly. This can be and has been compared directly with ASv6.5, which works just fine.

EDIT: From a flight I'm planning now, and as a more concrete example:

n9vh.png

ag7y.png

o3qo.png

Nothing there is correct, and if that large of a variance is an interpolation error, then something is seriously wrong. Pretty sure it's just reading the data incorrectly.

Link to comment
Share on other sites

I think we are beginning to mix the things up in this thread. Fact is:

No matter how AS collects his wind data, if it is the same or not the same as the "real weather" - If PFPX is reading the AS data, it should be able to display the same weather conditions as AS, whether they are wrong or not (compared to the real world weather in this moment). I don't care how the real weather is. All what I'm expecting is, that I can reproduce the same weather in PFPX and FSX - because I fly with the weather provided from AS. And if AS is collecting "old" or "wrong" data from their servers, so what, then I want to plan with that "old"/"wrong" data, because that is the data I will also fly with.

Another thing:

Imagine that you want to fly in a "historic weather event" (e.g. hurricane katrina). Then PFPX should very much be able to read and reproduce the historic weather data provided by AS. Currently this seems not to be the case.

Hey Chris, i read you posts again and we are on the same page, sorry for the "there is no mix" i gave you. regards !!

Link to comment
Share on other sites

Hi Bob.

I think what you may have missed here is that PFPX is retrieving wind data properly from ASv6.5, just not the newer versions, and that in those newer versions of AS, there was indeed a change in how the data is formatted in the snapshot file.

I totally agree with you otherwise, regarding the effects of interpolation, etc., but that's not quite what we're dealing with here. It may certainly be a factor, but it would be a minor one, if one at all. AS uses ~9000 weather stations across the globe casting a pretty large net, and in any case, from what I am seeing, the data NEVER matches, even for known physical stations. It's typically not even close.

Brian--As I recall some previous discussions of this issue vis-a-vis FSBuild, AS2012 uses a different (i.e. weighted values) interpolation method than AS6.5 did. PFPX does appear to interpolate wind values very closely to the values produced by AS6.5 (which I believe uses the easy-to-replicate average of the closest four stations), but I'm still thinking the mismatch between PFPX and AS2012 is an issue of different interpolation methods and not the data import itself. Even overhead a reporting station, there would presumably be a skew to the interpolated data, as nearby stations would still enter into the calculation.

In reality, 9000 stations spread across the 200 million square miles of the earth's surface isn't a very dense network of data points, especially considering that the majority of those are r/w reporting stations concentrated in areas other than the oceans and vast unpopulated regions, where the synthetic reporting stations are much sparser than an even distribution might produce.

Regards

Bob Scott

Link to comment
Share on other sites

I was trying to avoid any more comparisons to FSBuild, but since you've mentioned it, FSBuild has no issue with either source, at least no more than it ever has.

Neither of them will ever be able to provide perfect data, everywhere, and that is totally acceptable, imho. Perfection is not at all what I am looking for.

I guess the issue as I see it though, is that when I look at the raw snapshot and it says X; when I ask ASE through the interface for that same data and it agrees with the raw data; when I use FSBuild to plan and it ALSO agrees with the raw data; and finally when I load the sim and see a very close approximation of that data presented, it's difficult to not single out PFPX as the outlier.

Having said that, more often than not PFPX is close enough, and it's many other features are arguably more important/valuable than 90% accurate wind data. I just wonder if it couldn't be better.

Link to comment
Share on other sites

Hi Bob.

I think what you may have missed here is that PFPX is retrieving wind data properly from ASv6.5, just not the newer versions, and that in those newer versions of AS, there was indeed a change in how the data is formatted in the snapshot file.

I totally agree with you otherwise, regarding the effects of interpolation, etc., but that's not quite what we're dealing with here. It may certainly be a factor, but it would be a minor one, if one at all. AS uses ~9000 weather stations across the globe casting a pretty large net, and in any case, from what I am seeing, the data NEVER matches, even for known physical stations. It's typically not even close.

I posted most of this above already, but as it seems we now live in a TL;DR society...

TL;DR - PFPX is not pulling wx data from newer versions of ActiveSky properly. This can be and has been compared directly with ASv6.5, which works just fine.

EDIT: From a flight I'm planning now, and as a more concrete example:

n9vh.png

ag7y.png

o3qo.png

Nothing there is correct, and if that large of a variance is an interpolation error, then something is seriously wrong. Pretty sure it's just reading the data incorrectly.

Is correct that works with Active Sky 6.5 but now most of people use ASE or AS2012 and other programs like EFB are full functional with AS2012/ASE wheater. So i wish that in a future update there will be an extra options to import meteo from ASE/AS2012 with correct value import or the possibility to manually input the wheater into the flight plan but it would be terrible in case of transoceanic flights

Link to comment
Share on other sites

Here it goes..... all the files related to weather and a route LOWW KJFK.....

so just a side question.... you are obviously reading the current_wx_snapshot and wx_stations_list files, and interpret them.... whereas most people load the route into AS2012 and compare the winds in AS2012 route briefing to the OFP from PFPX..... so that might be the problem.

your interpolation and AS2012's interpolation are probably not the same...... would there be a way to import the AS2012 winds forecast into PFPX? or would it maybe desirable to actually alter the programm to sort of force you to load the flightplan to AS2012 and then read the forcast from there?

Just some thoughts....... asspecially since I think AS2012 will produce a readable file at least for the PMDG777...

Regards

Richard

ACH928 LOWW-KJFK (07 Sep 1335 UTC).pdf

document1.pdf

current_wx_snapshot.txt

wx_station_list.txt

Link to comment
Share on other sites

Hello,

I'm also with the very same issue. I tried to compute a L/R flight this morning after the .06 update and notice that the wind intensity was way off the AS2012 depiction and also PFPX weather server.

I had also noted maximum 35 degrees variation in the direction.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Guest
This topic is now closed to further replies.



×
×
  • Create New...