Foreflight Winds Aloft Wrong? or Misunderstood?

brcase

brcase

En-Route
Joined
Jun 11, 2008
Messages
2,821
Location
Boise, Idaho
Display Name
Display name:
Brian
I don't normally use Foreflight (I am to cheap) But when training people that do I have I noticed the winds aloft data they come up with often does not agree with other official sources (Like 1-800-WX-Brief).

In the attached photo the verified data on the right shows the winds aloft at 24K to be 240@29 for Boise Between 2000-0300Z (0100pm-0800pm Local).

Foreflight is showing the winds at 24K to 246@8. That is quite a difference? Why is foreflight showing us such different winds than the official sources?

Brian
CFIIG/ASEL
upload_2022-2-22_7-40-49.png
 
I have the same question, Brian, and I do use Foreflight.

I was easy chair flying the other day, and pulling up a previously entered flight had winds different from the winds at the airports. I thought it likely had something to do with Foreflight not picking up the latest data when you pull up a previously planned flight. Anxious to read what the experts here think.
 
Forecasted vs observed is my guess. The NWS source is a forecast, and I believe the Foreflight numbers are based on actual data gathered by airline aircraft that are reporting actual data through their ACARS or other automated reporting systems.
 
you don't have these issues with garmin pilot, just sayin






:)
 
Well, I booted up ForeFlight again a few minutes ago. I called up a flight, did the briefing via ForeFlight, then looked at the winds. Even though I sent the data to the map, the winds shown on the flightplan and the briefing page were different, and neither of them matched the winds from METARS at airports. I have about lost confidence in the program. There must be a glitch somewhere. I used to run Garmin Pilot and may have to switch back! :) I have a couple of months to make up my mind.
 
brcase said:
I don't normally use Foreflight (I am to cheap) But when training people that do I have I noticed the winds aloft data they come up with often does not agree with other official sources (Like 1-800-WX-Brief).

In the attached photo the verified data on the right shows the winds aloft at 24K to be 240@29 for Boise Between 2000-0300Z (0100pm-0800pm Local).

Foreflight is showing the winds at 24K to 246@8. That is quite a difference? Why is foreflight showing us such different winds than the official sources?

Brian
CFIIG/ASEL
View attachment 104828
Click to expand...
Sounds like your students are not to (sic) cheap to use foreflight and they can still afford to fly planes that operate in the flight levels. LOL :cheerswine:
 
Sounds like a great lesson for a student about the inherent inaccuracies of forecasts and the importance of checking and adjusting for actual conditions.
 
I see it a lot when checking different sources. Some, like Windy.com have shown 40-60kts while PIREPS on AWC indicate 8kts. same altitude.
 
Skyrys62 said:
I see it a lot when checking different sources. Some, like Windy.com have shown 40-60kts while PIREPS on AWC indicate 8kts. same altitude.
Click to expand...

I will be using 1800wxbrief to see if that works better. I know ForeFlight just made some changes in the wind functions so maybe there is a glitch or two.
 
I use ForeFlight. I have found winds aloft forecasts to be anything but accurate in western Colorado.
 
Archer Jack said:
Sounds like your students are not to (sic) cheap to use foreflight and they can still afford to fly planes that operate in the flight levels. LOL :cheerswine:
Click to expand...

Students don’t know any better and drink the FF marketing KoolAid.
 
brcase said:
In the attached photo the verified data on the right shows the winds aloft at 24K to be 240@29 for Boise Between 2000-0300Z (0100pm-0800pm Local).

Foreflight is showing the winds at 24K to 246@8. That is quite a difference?
Click to expand...

Maybe it has to do with variations in where the jet stream is forecast to be?

I don't have an old forecast for Monday, to correspond to your photo, but below is the ADDS forecast from aviationweather.gov for about 21 hours later. Note that there's a strong gradient in wind speed in northern Utah for example. A hundred miles makes a big difference. If two different models predict slightly different locations for the jet stream, then they might predict very different wind speeds for a location near the edge of the jet stream.

20220222_22_F01_rap_400_wind.gif
 
midwestpa24 said:
Forecasted vs observed is my guess. The NWS source is a forecast, and I believe the Foreflight numbers are based on actual data gathered by airline aircraft that are reporting actual data through their ACARS or other automated reporting systems.
Click to expand...

No, ForeFlight uses forecasts.
 
John Collins said:
No, ForeFlight uses forecasts.
Click to expand...
What I would really like to know is which forecast or combination of forecasts/interpolation is foreflight using?

This is a fairly common problem with most EFB's, They give you an answer but don't tell you how they got the answer. When the answer is obviously wrong and/or unverifiable how do we find out why?
This isn't just winds aloft, but is often the case when trying figure out what winds were used in the Nav/log or ETA computations.
This can be problematic when you sitting with an examiner on a checkride trying to explain where the winds you used for your calculations came from.

At least the nice thing about most other EFB's is they usually just echo what is reported by Flight Service.

scottd said:
I don't like using the FBWinds for the reasons I state in this article I wrote a long time ago
Click to expand...
Nice Article, Thanks for Posting.

Brian
CFIIG/ASEL
 
I’ve had FF winds be fairly accurate, and I’ve had them be way off. It’s strange
 
Foreflight uses a combination of NWS Forecasts and worldwide wind models to interpolate winds aloft data throughout your flight. In my area this ends up being very accurate the majority of the time.

https://support.foreflight.com/hc/e...-are-winds-aloft-used-in-NavLog-calculations-

I went down the rabbit hole working out the details of how their process affects flight planning, and the specifics of what you are viewing on the NavLog for any flights you've planned. It does follow the normal rules but because they don't explicitly outline some things, it can be kind of confusing - but really only for those of us who are intent on drilling down to fully understand how the back-ground calculations are made in Foreflight.

Fwiw, here's my quick rundown. Winds aloft are given in True (just like always) and in knots (again, just like always). This only gets slightly confusing when you are doing manual conversions to compare with and realize their Variation numbers may also be slightly off from what you find on a sectional. I'm guessing this is due to their data source being more frequently updated than sectionals as well as their interpolation when between two Variation lines drawn on our sectionals. It's basically immaterial (usually a max of one degree difference) - but makes digging into it a little confusing because the numbers never match precisely.

Another minor hiccup is that the wind speeds on the NavLog are always given in knots even if all your settings are switched to MPH for the aircraft you fly. Again, it's one more minor item you have to know if you are drilling into the numbers to figure out how it all works.

So, the main issue to understand is the fact that they use algorithm based interpolations of NWS (and other) wind data. If you go "old school" and use a good old telephone briefing wind forecast, your numbers will inevitably be different. The rest of the variance can be traced to the minor issues above. Those are only confusing when you are trying to make a perfect comparison between e6b calculations and Foreflights output.

In the end, like everyone, the end result is some very accurate ETEs and headings for any given flight. I generally am within 5 minutes (and often less) on a long cross country flight.

I hope that helps.
 
"All models are wrong, but some are useful." - George Box
"It's only a model." - Patsy
 
Last edited:
brcase said:
In the attached photo the verified data on the right shows the winds aloft at 24K to be 240@29 for Boise Between 2000-0300Z (0100pm-0800pm Local).
Foreflight is showing the winds at 24K to 246@8. That is quite a difference? Why is foreflight showing us such different winds than the official sources?
Click to expand...

What you designated as "official sources" is generated from the same NWS forecast product that ForeFlight gets its data from. The GFS runs 4 times daily (0000Z,0600Z, 1200Z,and 1800Z) and produces much more granularity than is published on "official sources" or ForeFlight in terms of time, place, and altitude (I believe the GFS granularity is every 0.25 degree latitude/longitude, hourly forecast out to 120 hours and every 3 hours for the next 240 hours and vertical in 500 foot increments to FL1050). I checked and updated the details where I was wrong, but the main point is the raw forecast product is much more granular than the "official source". The "official source" displays forecasts at a limited set of specific airports, with Flight levels of 3, 6, 9, 12, skipping 15, 18, 24, skipping 27, 30, 34, and 39. The "official source" reports publishes three time frames for the forecast that are 7 hours, 9 hours, and 12 hours long, each with only one time within each forecast period specified as valid time.

ForeFlight provides the winds aloft data in three hour increments on the airport view with altitudes in 3000 foot flight level increments to FL 540. The graphic Map product provides the winds each degree of latitude/longitude. The performance calculations use even more granularity in 500 foot increments and hourly for hours 0 to 36. So same data source, but grouped in different time frames, locations, and altitudes.

The comparison app screenshot shows the times as 2000Z to 0300Z, 0300Z to 1200Z, and 1200Z to 0000Z, so this is from the 1800Z run of the GFS and the only times these three forecast time frames are valid is at 0000Z for the 2000Z to 0300Z time frame, 0600Z for the 0300Z to 1200Z time frame, and 1800Z for the 1200Z to 0000Z time frame. So the time frame that ForeFlight should come closest to matching the "official source" is for the period before the one in the screenshot. So I would not expect the two to be an exact match, especially if there are wind changes during the time periods of the "official forecast" that will not show up on the "official forecast" but may show up on the ForeFlight winds aloft forecast due to the increased granularity of the forecast as presented by ForeFlight.

If you get the winds aloft from ADS-B FISB, it is even less granular in that it is only updated every 12 hours, even though the forecast is updated every 6 hours and is also much less granular than the ForeFlight internet product in terms of locations and altitudes. So I believe that if you had the winds aloft downloaded into the ForeFlight cache before a flight departure, it is very likely that it is a later forecast update than what you get during flight via FISB and is much more granular.

All that said, the winds aloft product is a forecast and we all know weather forecasts are 100% accurate. :)
 
Last edited:
I am unable to receive in-flight winds aloft via foreflight, if over a city at lower altitudes and cellular data becomes available then it’s possible winds aloft would display, so I’m guessing nothing comes from adsb-in. It also appears as though Foreflight does not save any data either for winds aloft when “packing” data for your trip.

Garmin displays winds in-flight but unsure if this is at altitude or ground.

I find that Foreflight ETA times en route can be accurate, but their performance profiles are ultimately too optimistic. I am unable to reach a destination in the total time that Foreflight typically estimates. First, on longer journeys, typically I plan point A to B, takeoff to landing, usually direct. This could be 500+ nm trip, when going into Foreflight I select the altitude to find winds aloft and fuel consumption. Typically there is an optimal level, if there’s a headwind at one level and a tailwind at another, I’d usually go for the tailwind. With the consideration of terrain in mind, if I am crossing mountains then I’d go higher.

It would be nice if Foreflight can suggest optimal legs, first 100nm at this altitude, second 100nm at this altitude and so forth, to take advantage of the winds (especially with the higher level Foreflight packages!). I did have a tailwind the entire journey, but it was minimal and ranged from about 0kts to about 5-10kts. I’m sure some calculation could optimize this even further.

Now I am unsure to the accuracy of Foreflights performance profiles and calculations, climb and descent rates, fuel burn, etc. After looking at some trips I have to say Foreflight is not accurate, especially when climbing to levels where you engine performance starts to decrease.

For example on a Piper Arrow 180hp, Foreflight believes we climb 100mph at 500fpm. The Arrow does not climb that well in my experience, at 25”/2500rpm, maybe there is a more optimal setting? You would be lucky to keep airspeed at or above 100mph, typically would need to reduce airspeed to 80-90mph to get a climb going otherwise 300fpm at 100mph is very possible.

Based upon my observations, fuel flow based on the manifold pressure gauge and speed is GS without considerable winds. En route lower altitudes under 5000’, I usually see:
135-140kts 12gph
130kts 10gph

When going higher 7500-9500’, I usually see:
120kts 7gph

And above 10,000’
110kts 6gph or so

Now if I look at my last two trips, Foreflight seems to calculate full performance at higher altitudes, showing low fuel burn, and faster arrival times. But in actual the time to get there was much longer.

I updated Foreflight to a custom profile of (generally this is for planned altitudes >5000’ and I have another profile for planned altitudes <5000’):
Climb 100mph @ 300fpm @ 12gph
Cruise 140mph @ 7gph
Descent 500fpm 160mph @ 7gph

And both of my previous flights actual time in flight and Foreflight prediction with my custom profile were spot on time wise. The interesting tidbit was that Foreflight also shows higher fuel consumption at higher altitudes and lower fuel consumption at lower altitudes. Going back to Foreflight’s profile it was the reverse.

Any thoughts? Would fuel consumption be higher at higher altitudes, is this because of an older plane or not climbing at correct power settings? (Too long in climb power settings is burning a lot of fuel too)

My custom profile
DF6E4583-A839-4693-997C-4BE9081D0314.jpeg

Foreflights performance profile
A6CE1264-FBAD-4D18-8A68-1767151347D8.jpeg
 
Just to be clear, do you have a performance subscription? If so, which profile are you using? There are three cruise profiles, 75%, 65%, and 55%. They should be based directly on the POH performance tables which will have a specific set of conditions, example RPM, MP, temp, pressure altitude and leaning technique. For climb, there are also specific conditions specified. So if your aircraft does not match the POH performance, it will not match the performance profile provided by ForeFlight. For each profile, there are bias adjustments the pilot can make. I don't have a copy of the 180 HP Arrow POH, but that is the source for performance data. In my Bonanza, the POH climb performance is based on 25 inches or full throttle MP, 2500 RPM, mixture leaned to appropriate fuel flow for altitude, cowl flaps closed, 107 Kts climb speed. What I actually do is full throttle, 2700 RPM, leaned to appropriate fuel flow for altitude, cowl flaps open, and 120 Kts. I could make a bias adjustment to the climb to account for some of the differences, but haven't.

In the altitude advisor you show, the headwind/tailwind is based on the average block GS for the trip with winds factored in vs the GS in a no wind situation. So you may have a head wind for part of the route and a tail wind for a different portion, if the total time is less than a no wind situation, then a tailwind will be shown and if the total time is greater, then a headwind will be shown.

I would probably climb at full throttle, 2700 RPM to altitude rather than reducing power to 25 square so to speak.
 
Yes, I have the three profiles as mentioned 75%/65%/55%. I was always taught to climb using 25”/2500RPM, would full throttle for 10-15 minutes be ok with the engine?
 
mandm said:
Yes, I have the three profiles as mentioned 75%/65%/55%. I was always taught to climb using 25”/2500RPM, would full throttle for 10-15 minutes be ok with the engine?
Click to expand...

Actually, it is easier on the engine. In the olden days, we taught to reduce the throttle to 25 square, but then when engine analyzers showed that the engine runs cooler with full throttle, we changed how we recommend operating engines that are rated for continuous operation at full power. It turns out the reason the engine runs hotter when you retard the throttle is because what the linkage is doing is leaning the engine first and not really reducing the power. Engines are setup to be extra rich at full power. Have you noticed that when you make a power reduction from full throttle, not much happens for the initial reduction and the MP stays at the top of the arc. Only after a substantial reduction do you see the MP start to come down. That is because the initial travel on the throttle mostly reduces the fuel flow, so the leaner mixture raises the EGT. Also climb rate in a piston engine is determine by the excess power available over that power needed to remain level at that speed. So although you have 180 HP at sea level ISA, only about half that power is excess and available for climb. I would estimate that reducing the power to 25 square, reduces total power about 10%, but the climb rate is reduced more like 20%, so it takes longer to climb to altitude. As you climb, each 1000 feet the MP drops about an inch, so if you start near 29 inches at sea level, without any changes in the full throttle setting, you will be down to 25 inches MP by the time you reach 4000 feet. That is a matter of 4 or 5 minutes. I am in no hurry to reduce the RPM, but I usually don't touch it until at least 1000 feet above the runway or my first level off. As I said the engine is rated for continuous use at full power. To keep the best power as you climb, you should lean the engine for the altitude. Many aircraft have a climb altitude vs fuel chart or table to know the desired fuel flow or an EGT/Engine Monitor. If you have the latter, note the EGT on the initial climb and lean to keep it at the same temperature. By the time you reach 7000 feet, even at FT and 2700 RPM, the total power is down to around 75%.

mandm said:
I am unable to receive in-flight winds aloft via foreflight, if over a city at lower altitudes and cellular data becomes available then it’s possible winds aloft would display, so I’m guessing nothing comes from adsb-in. It also appears as though Foreflight does not save any data either for winds aloft when “packing” data for your trip.
Click to expand...

Packing should update the Winds Aloft graphical forecast on the Map view from the internet. However, the Winds(Speeds) and Winds(Temps) are only available when you have an internet connection. Also the Winds Aloft on the airport view should get updated when you are on the internet and they should persist in flight. ForeFlight does not recommend using internet via cellular while in flight as it can cause errors. Remember that winds aloft is a forecast product updated only 4 times a day, whereas the winds aloft from FISB is only updated twice a day and has much lower resolution.
 
Did it change from winds aloft being a forecast to a report? Forecasts are always hit or miss.
 
John Collins said:
Actually, it is easier on the engine. In the olden days, we taught to reduce the throttle to 25 square, but then when engine analyzers showed that the engine runs cooler with full throttle, we changed how we recommend operating engines that are rated for continuous operation at full power. It turns out the reason the engine runs hotter when you retard the throttle is because what the linkage is doing is leaning the engine first and not really reducing the power. Engines are setup to be extra rich at full power. Have you noticed that when you make a power reduction from full throttle, not much happens for the initial reduction and the MP stays at the top of the arc. Only after a substantial reduction do you see the MP start to come down. That is because the initial travel on the throttle mostly reduces the fuel flow, so the leaner mixture raises the EGT. Also climb rate in a piston engine is determine by the excess power available over that power needed to remain level at that speed. So although you have 180 HP at sea level ISA, only about half that power is excess and available for climb. I would estimate that reducing the power to 25 square, reduces total power about 10%, but the climb rate is reduced more like 20%, so it takes longer to climb to altitude. As you climb, each 1000 feet the MP drops about an inch, so if you start near 29 inches at sea level, without any changes in the full throttle setting, you will be down to 25 inches MP by the time you reach 4000 feet. That is a matter of 4 or 5 minutes. I am in no hurry to reduce the RPM, but I usually don't touch it until at least 1000 feet above the runway or my first level off. As I said the engine is rated for continuous use at full power. To keep the best power as you climb, you should lean the engine for the altitude. Many aircraft have a climb altitude vs fuel chart or table to know the desired fuel flow or an EGT/Engine Monitor. If you have the latter, note the EGT on the initial climb and lean to keep it at the same temperature. By the time you reach 7000 feet, even at FT and 2700 RPM, the total power is down to around 75%.



Packing should update the Winds Aloft graphical forecast on the Map view from the internet. However, the Winds(Speeds) and Winds(Temps) are only available when you have an internet connection. Also the Winds Aloft on the airport view should get updated when you are on the internet and they should persist in flight. ForeFlight does not recommend using internet via cellular while in flight as it can cause errors. Remember that winds aloft is a forecast product updated only 4 times a day, whereas the winds aloft from FISB is only updated twice a day and has much lower resolution.
Click to expand...

Thank you, everything you said makes sense. I flew again today and kept full throttle in longer, and climb performance was there. I did lean the mixture slightly, and with the climb, although I thought I read somewhere that mixture should only be leaned under 75% power?

I did have the same reaction, when bringing back the throttle, the MP would remain and require quite a bit of reduction before MP would drop. You can get significant fuel flow reduction by finding the right combo which seems more of an art than following the Lycoming engine setting chart.

Speed wise, under 6000’ MSL, delivered the best airspeed. 24”/2400RPM 130-135kts at 9.5-9.75gph fuel flow. Anytime I went higher, I couldn’t get the desired performance. Unfortunately I had to go higher today for a bit due to clouds, winds and turbulence. Was a miserable 80-100kts GS until I descended and picked up 140-160kts GS.
 
In my Bonanza, the fuel flow gage has a climb fuel flow marking based on altitudes of 3000, 5000, and 9000. Some POH manuals will specify a fuel flow table for climb vs altitude. Without markings or a table, you can accomplish the same by using an EGT. Assuming you start out at full rich on the ground and have an engine analyzer, note the EGT for the initial climb. Say it is 1200. This is with a full rich mixture position and is provides an extra rich mixture for climb. Every few thousand feet, you can lean the engine to the same EGT value (example 1200) you had at takeoff. This will not over lean the engine in the climb at full power. Note that as you climb, the percent power the engine develops continues to decline, even with full throttle and 2700 RPM. By the time you get to 7000, you are roughly only producing 75% power. I believe the Arrow has an analog fuel flow indicator, note the fuel flow for each altitude when you lean the mixture in the climb using the same EGT method and make a fuel flow vs altitude chart. You should be able to lean in the future using your chart or your EGT. If you don't lean in the climb, the engine will go from very rich initially to over rich as you climb and you will lose some climb performance. Leaning properly restores the mixture from over rich to just very rich.
 
TommyG said:
Did it change from winds aloft being a forecast to a report? Forecasts are always hit or miss.
Click to expand...

No, nothing has changed, Winds Aloft is and has always been a forecast. Some aircraft systems with an ADC and HRS can calculate the actual wind speed and direction, but ForeFlight can't do this type of calculation because the aircraft heading and air data information is not available. All ForeFlight knows is GS and GPS track and some form of altitude.
 
John Collins said:
No, nothing has changed, Winds Aloft is and has always been a forecast. Some aircraft systems with an ADC and HRS can calculate the actual wind speed and direction, but ForeFlight can't do this type of calculation because the aircraft heading and air data information is not available. All ForeFlight knows is GS and GPS track and some form of altitude.
Click to expand...

Ok that’s where my mind was.
 
Fore Flight allows you to enter your own performance tables. I wasn’t happy with the planning from their own tables, so I took the time to enter a custom one from the POH, and get almost exactly the same time and fuel used. :D

For climb, unless you have a time limited take off setting, full throttle and full RPM and full Rich are best.

The hard core LOP guys do the Big MIxture Pull at WOT. They quickly pull the mixture back to a given fuel flow that is known to be well into LOP. And it works. You are exchanging cooling with air instead of cooling with excess fuel.
 
somorris said:
Well, I booted up ForeFlight again a few minutes ago. I called up a flight, did the briefing via ForeFlight, then looked at the winds. Even though I sent the data to the map, the winds shown on the flightplan and the briefing page were different, and neither of them matched the winds from METARS at airports. I have about lost confidence in the program. There must be a glitch somewhere. I used to run Garmin Pilot and may have to switch back! :) I have a couple of months to make up my mind.
Click to expand...

Of course, you must know that METARS only show surface winds, so winds aloft would rarely match.
 
You must log in or register to reply here.
Back
Top