Focus should be on the big wx picture

[Scott Dennstaedt, PhD]

It always amazes me when I see pilots focused way too much on the details and not the big weather picture.  I believe this is one of the absolute biggest mistakes made by pilots from a weather planning perspective. METARs and TAFs are by far the two most used weather products with the big weather pics barely used as shown in the table below.  My routine is just the opposite.  The big picture stuff is what will kill you, not the details.  It's like if someone sitting next to you on a commercial flight pulls out a bomb and you notice how shiny it is or the colorful wires are. It's a BOMB! Sure, the details are important and help to fill in what the big picture doesn't offer. But what you see below is the absolute wrong approach.  

Source: Knecht, 2007.

Edited February 7, 2020 by Scott Dennstaedt

[1964-M20E]

I think what you might be seeing is a progression of how easy a task is.  Most weather products when supplied in text format generally are very verbose and fail to keep one's attention.

My planning process is as follows for what is worth:

1. Local TV weather: easy  big picture item provides good information projected over the next few days
2. METARs: easy big picture item when you zoom out and look at the dot color on skyvector.com or other service.  This tells me what is happening right now along my route of flight.  Short flights 1 to 2 hours tells you almost everything you need to know.
3. TAF: easy detail item gives you some view of the future.  Especially important for landing conditions at your arrival.
4. NEXRAD radar.  Big picture lets you see where the weather is coming from, where it is going and how intense it is
5. AIRMET, SIGMET big picture items higher level of difficulty sometimes these seem to be crying wolf
6. Prognostication charts, weather charts etc. are very high difficulty, unless you live and breath them and understand them well they are not of much value.
7. Once my go no-go decision has been made either the day of the flight or earlier in cockpit ADSB weather services provide updated information to adapt to changing situations

 

[jlunseth]

Well, I wonder if those stats aren't affected at least to some extent by type of flight. I live in the midwest (Minneapolis).  We have to go quite a ways to run into marine conditions (advection fog, etc.) or any significant orographic effects.  My master bedroom is on the upper level, northwest corner of my house and I like to joke that there is nothing between me and Grand Forks to slow the wind down, or me an Alberta for that matter.  Statistically, most of my flights are just local flights, larking around, shooting a few approaches, doing some landings.  Being where I am, I might travel 50 miles in any direction to find and airport that is not being heavily used at the time, but I don't have to fly mountains or even foothills to do it.

When I do these "larking" trips I pick a good VFR day.  I do that because I can, why go flying for fun if I have to deal with even the threat of convection, or icing, or any of that stuff, and I am not going to go very high even though the aircraft certainly can.  Nevertheless, I generally get a briefing on Foreflight by putting in a tentative route and hitting the briefing button.  I can't say that I read all the stuff very carefully. I do look through it to see if there is anything that will interfere with a couple of hours joyriding and practicing.  I do look heavily at METARs and TAFs, but also look at Nexrad on foreflight and look at the winds aloft, among other things to see if there is anything squirelly like high low level winds even though ground winds are not too bad.  As I said, from a statistical standpoint, my focus would be on the things high on that list and I wouldn't bother with a lot of the rest of it, because I can see it with my Mark I and get back to the airport and on the ground in 15-20 minutes.

Now, if I am traveling somewhere, even a relatively short somewhere like a 100 mile trip for lunch, or out to the Black Hills, or down to the Bahamas, that is an entirely different matter and I want to know everything.  And then I have XM on the panel and will put the weather up and check ahead constantly for changes.

But as I said, from a statistical standpoint, those flights may be a lot of my hours, but not a lot of my flights. So if the stats are related to checking weather at the start of each flight, for many of my trips, the full monty is not necessary.

[smwash02]

I'd like to see a more recent version of the survey. 2008 (a year after your survey data) is when I started flying and the difference in affordability and availability of the data is staggering.

Moving from a 496 with XM radio and DUATS/ADDS to tools like Garmin Pilot/Foreflight with ADS-B and web tools like Skyvector.com make it incredibly quick and easy to get a big picture view. I still supplement with ADDS as well; however, interpreting some of its data still takes some effort.

I think with so much data available, we'll need work to prevent information overload and ensuring the right data is digested at the right times else it may just get ignored.

[ArtVandelay]

Too many variables, a 100 mile trip vs 500 mile trip vs 2000 mile trip are treated differently, in both time spent planning and tools used.

[steingar]

I start with my phone's weather app for both my home base and my destination.  If it's a longish or long flight I program the phone to watch some waypoints as well.  I start really obsessing about the weather a couple days before, and usually hit the NOAA site to see how things are shaping  up.  Day before I start with Foreflight, seeing how things are and what the avails forecasts look like.  If all looks good I launch, and still check for weather en route (dam I like ADSB).  Once in awhile the weather is what's forecast, but not often.

[Scott Dennstaedt, PhD]

On 2/10/2020 at 3:09 PM, ArtVandelay said:

Too many variables, a 100 mile trip vs 500 mile trip vs 2000 mile trip are treated differently, in both time spent planning and tools used.

Perhaps, but from my perspective, my procedure isn't much different.  Yes, there's more ground to cover and potentially more weather to deal with...but if I think back to some of the most intense weather planning I've done in the last 20+ years, it's been on very short flights (less than 100 miles).   For me, short or long flights, that big picture is the key to making a safe flight. My decisions to go or stay typically come from this.

[Scott Dennstaedt, PhD]

On 2/10/2020 at 1:44 PM, smwash02 said:

I'd like to see a more recent version of the survey. 2008 (a year after your survey data) is when I started flying and the difference in affordability and availability of the data is staggering.

Moving from a 496 with XM radio and DUATS/ADDS to tools like Garmin Pilot/Foreflight with ADS-B and web tools like Skyvector.com make it incredibly quick and easy to get a big picture view. I still supplement with ADDS as well; however, interpreting some of its data still takes some effort.

I think with so much data available, we'll need work to prevent information overload and ensuring the right data is digested at the right times else it may just get ignored.

I'm actually doing my own study right now as I'm working to finish my PhD.  And it turns out that it doesn't really matter.  In fact, it's amazing how little pilots are taught these days about how to look at the big picture. CFIs still tend to gravitate to the METARs, TAFs, PIREPs...all of those "details" and less focus on the big picture.  So pilots don't really know any better.  Yes, the availability is there for all of these big picture products, but the right approach to using the big picture is never taught. 

I'm writing a journal paper right now and for VFR into IMC accidents (the most deadly of any causal factor) a majority of these accidents are during the cruise phase of flight.  Recent studies have shown that pilots tend to focus on the wx at the departure and destination airports and less on the en route phase largely because they fail to spend enough time on the big wx picture.  

[Scott Dennstaedt, PhD]

On 2/10/2020 at 1:29 PM, 1964-M20E said:

I think what you might be seeing is a progression of how easy a task is.  Most weather products when supplied in text format generally are very verbose and fail to keep one's attention.

My planning process is as follows for what is worth:

1. Local TV weather: easy  big picture item provides good information projected over the next few days
2. METARs: easy big picture item when you zoom out and look at the dot color on skyvector.com or other service.  This tells me what is happening right now along my route of flight.  Short flights 1 to 2 hours tells you almost everything you need to know.
3. TAF: easy detail item gives you some view of the future.  Especially important for landing conditions at your arrival.
4. NEXRAD radar.  Big picture lets you see where the weather is coming from, where it is going and how intense it is
5. AIRMET, SIGMET big picture items higher level of difficulty sometimes these seem to be crying wolf
6. Prognostication charts, weather charts etc. are very high difficulty, unless you live and breath them and understand them well they are not of much value.
7. Once my go no-go decision has been made either the day of the flight or earlier in cockpit ADSB weather services provide updated information to adapt to changing situations

 

Yes, a METAR gives you some concrete info so it's consumed easily.  I agree that textual forecasts are certainly not the best approach.  

It's always interesting to see how pilots view certain products.  I don't see NEXRAD as a big picture thing.  It may help to explain the big picture (for the past), but it's not something I utilize heavily to understand what might happen over the next two to six hours.  I like to focus on those prog charts...but I tend to focus on not just the surface, but use constant pressure charts to define where the real energy is in the atmosphere. How those upper-level features evolve gives me the information I typically use for a go or stay decision.  I love the G-AIRMETs.  I don't see them as crying wolf at all - but you understand their limitations...in fact, they tend to drive my decisions about route and/or altitude more than anything else.   

For TAFs and METARs I rarely spend more than a few minutes looking at those.  I draw some important "key words" from those and apply that to the big picture. I tend to use TAFs more for alternate selection (plan B or C).    In fact, when I do a 1-on-1 training session with pilots we spend about an hour going over the trip and rarely get time to discuss these.  Most of the time, they end up being less important once you see all of the big picture details.       

[Jerry 5TJ]

18 minutes ago, Scott Dennstaedt said:

For me, short or long flights, that big picture is the key to making a safe flight. My decisions to go or stay typically come from this.

Thanks, Scott for your perspective.  
 

I start flight planning with a look at the National Wx maps: Location of fronts, areas of L and H pressure.  NEXRAD and the AWC convective and icing graphic forecasts are next.  By this point I’ve sometimes already decided against a flight just based on that initial overview.  

[Scott Dennstaedt, PhD]

12 minutes ago, Jerry 5TJ said:

Thanks, Scott for your perspective.  
 

I start flight planning with a look at the National Wx maps: Location of fronts, areas of L and H pressure.  NEXRAD and the AWC convective and icing graphic forecasts are next.  By this point I’ve sometimes already decided against a flight just based on that initial overview.  

You are welcome.  I typically start with the surface analysis chart and then work my way to the prog charts, and spend most of my time on the 500 mb chart...then, depending on the time of year and location of the flight, I compare this upper-level picture to the convective outlooks/forecasts, precipitation forecasts and G-AIRMETs.  Often with that overview I will typically begin to look for other times to fly...

[steingar]

What pilots need is sufficient meteorological knowhow to look at a weather chart and be able to predict what's going to happen based on what they're seeing.  While we get some of this during our training phase, we don't get enough and it doesn't stick that well.  Then again, meteorologists get degrees in this stuff, so I suppose I can't blame busy professionals too much for a lack of insight in an unrelated discipline.  But it does come in handy.

[0TreeLemur]

13 hours ago, Scott Dennstaedt said:

You are welcome.  I typically start with the surface analysis chart and then work my way to the prog charts, and spend most of my time on the 500 mb chart...then, depending on the time of year and location of the flight, I compare this upper-level picture to the convective outlooks/forecasts, precipitation forecasts and G-AIRMETs.  Often with that overview I will typically begin to look for other times to fly...

Great thread Scott!

In my opinion, it would be possible to provide a more consistent and complete presentation of available data.   The Foreflight briefing app provides an example that helps to guide the way, but I still augment it with other information.   I wish NOAA would add more explicit representation of freezing level.  It might be there in some kludgy javascript flight path tool on aviationweather.gov that nobody uses because the graphics are slow and look like an 80's computer game.  They could also show GFS model output for the time of flight because it usually shows areas of convection.

Advances made by the aviation industry in avionics have not been matched by the presentation of actionable info.   This time of year, I want to know where the freezing level is along my route of flight, and am I going to be in clouds or precip.  The flight I planned last Thursday from AL to CO was a planning nightmare because of a descending freezing lvl along my route of flight and multiple cloud layers.   In the summer, I want to know the risk of thunderstorms, which they already do a decent job of showing using green/yellow/red blobs.  That could be improved using short term model output which is available, but most pilots are not aware of it.

Just my personal thoughts.  While affiliated with NOAA, I do not work with NOAA's aviation division.

[PaulM]

Scott, 

  Part of this is that you are a Meteorologist, and will analyze those charts.   Most pilots are tactical and want the post analysis data.

I usually start about a week out with windy.com and the ECMWF model.  I'm looking for the projection of fronts and low clouds/IFR.  on my targeted travel days, to see if I need to move the travel date +/- 1 day.     I then watch this solidify over the week as the timing of the frontal movement gets clearer. 

Once we are within 48 hours I'm pretty set with what we will get, and look then only for problems with T-storms and icing, IFR minimums and alternates for the flight.  I'll use windy and the MOS product. 

The FAA/NOAA forecast products are very vague... if they show no precipitation, you can be sure there is nothing there.. but you will have large areas of forecast rain, and possible rain, that are over broad.. 

Within 24 hours the TAF's now give you local knowledge for the terminal forecast. Once we are within 12 hours, the ICP products are valid, and the GFA, and Flight path tools can be used.   This allows me to finalize the icing decision, look at cloud layers and skew-t.   I'll finalize the route and optimize the height for winds, cloud layers etc.. and execute. 

Examples:

  I was in Denver for Christmas, and there was a snow storm forecast, by Wednesday  I knew that we had to get out a day earlier (Friday) to beat the storm and picked a destination (Ohio) that would be beyond the affected area.   The  12 hour forecast showed freezing fog, and possible light icing in the lower clouds.. So I planed for right at that minimum (FIKI), and would be out soon to the east.   Weather was clear, we were above a layer and had no issues. 

Last year on the way to Florida, there was a front forecast stretching from WV to the coast with freezing levels to the MEA and other things in the mountains, but what looked like the 100miles to the coast were clear in the morning, with the freezing level above 6000' and just a stretch of IFR weather.   This worried me, but it was over 12 hours so I didn't have the ICP products yet.   On Friday morning, I ran the OGIMET analysis for the next day, and it showed me what I was missing, it was snow above 6000'.. descending into warmer weather, and therefor freezing rain.    I moved up the departure to that afternoon, and flew down to Grand Strand which would be south of the front, and completely bypassed the icing threat.   An easy IFR flight to FL the next morning. 

I'm planning this year's flight down to FL... 1 week out it was showing a front from the SE across the Carolinas.   By T-5 that forecast front had moved on beyond the coast for Saturday, and I'm seeing easy VFR until south Florida and perhaps a simple IFR letdown at my destination.    With that forecast there isn't much more to do until the final morning to adjust for winds, and morning fog, and looking at the cloud layers for the final IFR planning. 

 

[Ibra]

2 hours ago, Scott Dennstaedt said:

In fact, it's amazing how little pilots are taught these days about how to look at the big picture. CFIs still tend to gravitate to the METARs, TAFs, PIREPs...all of those "details" and less focus on the big picture.  So pilots don't really know any better.  Yes, the availability is there for all of these big picture products, but the right approach to using the big picture is never taught. 

Forecast models and weather resources are not very optimized to aviation aspects (say accuracy of predicting VMC/IMC between 2000ft and 6000ft at specific 100nm square in the next 3 days), the useful tools I found to get that global picture for short-term planning (6h to 3 days) are "data visualization tools" (opposed to forecasting numerical models or data point measurements), they just show you a lot of interpolated raw data, aggregate/average some of the raw data yourself, then check against your weather minima: yes go and fly !

Example is Windy, "Low Cloud" (average cloud % bellow 6000ft), "visibility" (30km average) and "cloudbase" layers are good average indicators of VMC/IMC at my GA flying height, the rest is probably just too much details for me and did cause me lot of disappointments: arriving at wrong METAR without seeing it coming in TAF/METAR when I departed

One is not that wrong on global picture using space/time averages, even when wrong you will know how much room you have and where good weather sits
Obviously, I need TAF/METAR to guess weather at A & B but the actual weather will be on ATIS, real-time data and naked eye     

 

Edited February 12, 2020 by Ibra

[Bob - S50]

I agree with Paul.

I start looking at Windy and know it changes every day so I don't plan much based on what it tells me.  It's just a trend indicator.  I watch it for celings, temps at cruise altitudes if I'll need IFR, and winds (for turbulence here out west).

Three days prior I start looking at the route using your weatherspork looking at cloud layers, winds, and freezing level.

Day of I look at everything that 1800wxbrief shows me.  Airmets. I take the turbulence airmets with a grain of salt and rely more on wind velocity and direction at mountain top levels.  Convective forecast (not usually an issue in Seattle).   I look at all the METARs and TAFs along the route of flight looking not only at the forecast but the trends in those forecasts.  If weather is even a bit iffy, I compare current weather with the forecast for the same time period to see if it is better or worse than forecast.

Good enough for me.

[Scott Dennstaedt, PhD]

9 hours ago, PaulM said:

Scott, 

  Part of this is that you are a Meteorologist, and will analyze those charts.   Most pilots are tactical and want the post analysis data.

I usually start about a week out with windy.com and the ECMWF model.  

 

Paul,

It's funny that you say that I'm a meteorologist and will analyze those charts...and then you launch into the fact that you look at windy and the ECMWF model.  We are actually looking at the same thing...the big difference is that I'm likely squeezing more out of this than most, but that doesn't mean pilots can't learn how to do this.  It's not all that difficult....and those folks that I've taught over the years will swear by it.  Learning what a negative tilt to a trough means...or what a jet streak in the atmosphere means is quite important to explaining the weather you might face.  The glance value of these charts are incredible.  I've always thought the FAA missed the boat here not teaching pilots how to use these upper level charts.  

[Scott Dennstaedt, PhD]

9 hours ago, Bob - S50 said:

Airmets. I take the turbulence airmets with a grain of salt and rely more on wind velocity and direction at mountain top levels. 

I don't pay a lot of attention to the wind velocity.  Strong winds don't equate to turbulence. Wind speed (and direction) is just one ingredient in the turbulence recipe.  I've have flown over the Rockies dozens of times where the winds at the peaks were greater than 50 knots and the air was glassy smooth.  The more significant bumps were near the surface where thermal turbulence takes over.  

[Scott Dennstaedt, PhD]

10 hours ago, Ibra said:

Forecast models and weather resources are not very optimized to aviation aspects (say accuracy of predicting VMC/IMC between 2000ft and 6000ft at specific 100nm square in the next 3 days), the useful tools I found to get that global picture for short-term planning (6h to 3 days) are "data visualization tools" (opposed to forecasting numerical models or data point measurements), they just show you a lot of interpolated raw data, aggregate/average some of the raw data yourself, then check against your weather minima: yes go and fly !

 

I like to see both.  I want to know the story...that comes from looking at the forecast models (e.g., constant pressure charts) and learning what story Mother Nature is telling.  Then, when I look at the "visualization" tools, then and only then can I understand why I should believe them.  To look at them without that big picture is like driving down the road looking through a straw.  Nobody would do that...so why should we do that with our weather planning?  

[whiskytango]

This is a great thread.  One thing I have been pleasantly surprised by is the accuracy of the 3-day NWS GFS MOS forecasting.  I use the usairnet.com Aviation Weather Forecast product as part of long range planning / trend analysis, and have found that it is often more accurate at 3 days out than the 24 hour TAF.  On a recent flight to Florida the TAF was predicting good VFR at the destination airport, but the 3-day MOS was predicting marginal VFR with lower ceilings than the TAF.  Based on the MOS forecast, I was not surprised when ATIS advised to expect a GPS approach.  I had leisurely briefed the approach the day before.

[carusoam]

Scott,

it is great to have you here.

We have a couple of experts in their area(s)...

And a few... where we all have interest in what you are doing!     Like Mechanics, and flight instructors... and weather guy!

...then there are the other experts... the kind of expertise you don’t want to need... doctor, and lawyer... 

I learn something every time you start a thread...

Good luck with your studies.  It would be interesting if there is a way MSers can collectively help you with a project.

Best regards,

-a-

[Ibra]

@Scott Dennstaedt, I will be interested in reading any white paper on "weather and flying"

The take-away IMO is not about accuracy of weather forecasting or getting swamped with lot of data, it is just about simple flying strategy on a decent global picture

Say VFR-in-IMC or CFIT, the problem of going A to B without "hitting clouds" or "hitting terrain" is really hard when you include tactical changes of height/heading
It is hard even with complete static/deterministic weather picture, obviously weather uncertainty and dynamics makes it more difficult
The core challenge: optimize a global A to B flight path while looking through a straw (getting stuck in local minima or hit boundaries)

The safe strategy involve simple routes with enough margins that is built around a global picture (one or two headings/altitudes to fly and one or two early tactical changes on few data)
Everybody get stuck in dead ends driving on GPS with no magenta line if they keep looking at it very often rather than just driving west

There is also aircraft performance, even with good 20km visibility I just went into clouds trying to climb on top
I have a trick now: only climb on top of things at 10deg above horizon anything above circle (5deg in hot summer)
Luckily tried this on clouds not mountain peaks  

 

[ArtVandelay]

Perhaps, but from my perspective, my procedure isn't much different.  Yes, there's more ground to cover and potentially more weather to deal with...but if I think back to some of the most intense weather planning I've done in the last 20+ years, it's been on very short flights (less than 100 miles).   For me, short or long flights, that big picture is the key to making a safe flight. My decisions to go or stay typically come from this.

On short flights, there is always the option of turning back.
Weather forecasts are much better today, but they still have problems predicting the timing. I think every trip I have canceled was because a system didn’t move when it was forecasted to, or move earlier/quicker than was forecasted.

[Scott Dennstaedt, PhD]

On 2/13/2020 at 7:09 AM, ArtVandelay said:

On short flights, there is always the option of turning back.
Weather forecasts are much better today, but they still have problems predicting the timing. I think every trip I have canceled was because a system didn’t move when it was forecasted to, or move earlier/quicker than was forecasted.

Correct on timing.  However, for these short flights it's not as much about forecasts as it is about observational data.  Learning to assess the upstream trends.  The first few hours of a TAF, for example, are often more related to these trends than any short-term model forecast.  Short flights you lose one of my favorite options...altitude.  Climbing on top is usually not an option.  

[steingar]

25 minutes ago, Scott Dennstaedt said:

 Climbing on top is usually not an option.  

Yeah, but if the flight is that short you're not going to get beat up for long.