Weather Forecasters!

[tacpot]

I wanted to go flying this evening. All the signs we good, but the TAF for 18:00 was ??G27 and SCT 035. The school said no way!

The current METAR:
"30007KT CAVOK M01/M04 Q1020" - right down the runway.

Not a cloud in sight!

Gits!
http://secure.metoffice.com/lib/template/spacer.gif

[Gertrude the Wombat]

Pick a different school? - there are those who will say "yes, we know the forecast says G27, but it hasn't done that all morning so off you go".

[FlyingForFun]

Pick a different school? - there are those who will say "yes, we know the forecast says G27, but it hasn't done that all morning so off you go".
Been there, done that. I remember it well.

One day last winter, the wind was forecast at something like 180/20G30, and the closest runway to that was runway 13. But the wind had been more like 150 all day, and had never gone above 15-20 knots.

Last lesson for the day was a night flight - my student's first night flight, in fact. I briefed him for the entire flight, and at the end of the briefing I explained that I'd let him have a go at the landing, and explained the essential points, but since landings weren't the main objective of the lesson we weren't going to spend any time looking at it in detail.

The weather was perfect for the flight - a clear, moonless night, perfectly smooth. We did everything we needed to do, then headed back home. Imagine our surpise(!) when ATC told us the wind was 180/25G35!

I remember, very clearly, turning to my student and saying "Do you fancy giving it a go?" He looked back at me and "I think I'll leave this one to you." Damn! At the time, that was the most cross-wind I'd ever landing in, in any aircraft. But it certainly concentrates the mind - I managed an absolute greaser, although it was a bit further down the runway than I'd have liked!

You never can tell.....

FFF
-----------

[scooter boy]

FFF - copy that.

As regards cheeky x-winds:
Yesterday, case in point I popped over from Plymouth to Jersey to gorge my little Mooney's capacious long range tanks with el cheapo fuel and also pick up a couple of 20L deicing fluid drums.

Plymouth was dead calm on departure and the metar still gave light winds when I left Jersey and took off at close to MTOW. The Plymouth TAF was for it to get pretty gusty but there was no sign of this 30 mins after the beginning of the BECMG period.

45 mins later the wind was pretty lumpy on the approach (30-40 kts at 1000') gusting well into the twenties right across the runway and although I managed a smooth landing my pulse took a while to settle down to normal!

On the subject of forecasters:
Although we can all think of examples of wildly inaccurate predictions (e:g Michael Fish's hurricane) the forecasts have an uncanny habit of being spot on.

My personal favourite at present is Carol Kirkwood - sadly though every time I call the met office to "talk to a forecaster" she is unavailable - must be washing her hair again!!

SB

[Gertrude the Wombat]

Although we can all think of examples of wildly inaccurate predictions (e:g Michael Fish's hurricane)
That's unfair. He knew perfectly well there was a not-quite-hurricane coming, he just thought it would pass a few tens of miles further to the south than it actually did, thereby only affecting France and not part of the British weather forecast.

[scooter boy]

Gertrude,
What really sticks in my mind about "the night of the storm" is the fact that Fishie went out of his way to categorically deny the possibility of the storm occuring and even to half jokingly mention the lady who telephoned to predict a "hurricane"!

Not mentioning it is one thing but to say that it will definitely not happen is a little different.

Don't get me wrong, we all make mistakes in our professional lives, MF was just unlucky that his was recorded for posterity and is replayed very frequently on "TV's greatest ballsups programmes".
SB

PS
you didn't happen to be a beardy balding weather forecaster called Michael Fish in a previous life did you?
;)

[IO540]

Firstly 1987 is a long way back in terms of computer weather models.

Secondly I don't think the Met Office consists of one man (or woman) who stands up in front of the BBC TV camera. In fact some of the present-day BBC presenters look like they might be more comfortable selling lottery tickets. There is a lot of work behind the scenes involving a lot of people. They also look at models produced abroad to see how they compare.

Fast moving / extreme weather is very hard to forecast. One can occassionally find big differences in the data from the UKMO, and the U.S. run GFS model (e.g. here (http://www.arl.noaa.gov/ready/cmet.html)) which most of the world runs on.

As to the conditions (actual or forecast) which a school will fly on (dual assumed) one could debate this both ways. An honest instructor might not wish to waste the student's money flying in bad weather. I also know many instructors do take people up in bad weather; one sees this every day when flying. IMHO, if you take up a fresh student for circuits in 15kt crosswinds his money will be largely wasted, but again this could be argued both ways.

For solo flight when self fly hiring, it's up to the school where (in terms of weather) they set the rules. If pre-PPL solo, the instructor doesn't want a blot on his license by sending up a student against a bad TAF which looks unreal but which later turns out to have been correct; he won't have a leg to stand on if that happens.

[robin]

I agree with IO540.

But there has been a noticeable trend over the past year or so of the forecast timings being out by a long way, and that some of the events turn out to be less severe than predicted. (have they never heard the fable about the boy who cried wolf?)

Personally, I now never cancel a booking on the basis of a morning forecast, but usually go to the field and review the situation there. I reckon last year I flew on a lot of my planned days, which the forecast said would be poor, and vice versa.

What has annoyed me more, though, is reviewing the weather the night before, from a collection of sources, and waking up to find something totally different by the 08.00hrs report.

In my view, either the weather we get now is more unpredictable than in the 70s, 80s and 90s, or the methods they use now are more prone to changes. I'd like to know which.

[IO540]

I am certain that the UKMO errs on the side of caution. They got a lot of stick for 1987.

Some of it is silly, like the permanent moderate icing forecast for any cloud (F215).

The best forecasts one is going to get are the TAFs and I find they are usually there or thereabouts.

An instrument qualification and capability is priceless, in terms of being able to fly on a planned date/time.

[robin]

The best forecasts one is going to get are the TAFs and I find they are usually there or thereabouts.

If only that were true. The classic last year was a series of TAFs forecasting at each update a weather change was scheduled to arrive during the period.

The first TAF expected it around 10.00, this went back to 14.00, then 18.00hrs, then 22.00hrs. In the end it didn't arrive at all, and the day had been completely flyable. All this at an airfield 2 miles from the Met Office!!!!!

[IO540]

The point here, Robin, is whether some other forecast was any better?

If weather is moving along at 200kt and the fronts are spaced by 20nm then you can forget weather forecasting :) It just can't be done. You can sit there and point at the MSLP chart and all the muck coming this way from the US east coast and you can see it will come but you cannot tell when exactly, or even approximately until a few hrs beforehand.

The other extreme is say the summer of 2003 when I flew from UK to Malaga and the QNH was 1013 the whole route, and flying back 3 days later I didn't even need to bother getting weather; the QNH was still about 1013 and the picture was the same.

Localised fog is also impossible to forecast well because its formation depends on very local conditions.

TAFs remain the best you will get. For VFR, one always should get TAFs along the whole route. Well, a METAR is the best :) but that's only good for an immediate flight, having first checked on the MSLP chart that the situation over the whole area is pretty static.

The bit where the UKMO does pilots a dis-service is in longer range data, e.g. "can I fly back home 3 days from now?" They give you only the MSLP charts, and they flog the rest to commercial weather repackagers. That's where the GFS link I gave above comes in - use the detail from it in conjunction with the timing of fronts from the UKMO MSLP charts because the UKMO model is better for the UK typical SW airflow.

[scooter boy]

The bit where the UKMO does pilots a dis-service is in longer range data, e.g. "can I fly back home 3 days from now?" They give you only the MSLP charts, and they flog the rest to commercial weather repackagers.

Couldn't agree more.

With patchy internet access when you are in your hotel room wherever that may be getting good met data enabling you to optimise return route and timing can be a challenge.

I now have a mobile card for my laptop (as of last month), I had previously used a sony ericsson P910i phone for weather whilst away but the data was neither fast nor as detailed as I needed it to be and rather expensive too.

I have also noticed the BBC interactive weather forecast on the TV becoming limited to 24 hrs (max 48hrs) in advance rather than the 5 days we used to get this time last year.

An inane picture of a babbling brook or a snowy motorway takes the place of a weather map - anything to cut costs and reduce risks of inaccuracy.

How about a 5-day aviation forecast free of charge?
Do you think the general public would buy that?
SB

[IO540]

Go to the NOAA URL I gave earlier. You can get a 84-hr (or longer) forecast as detailed as you want.

Whether it will be accurate is another matter; even the best website doesn't abolish randomness... :)

[FlyingForFun]

The best forecasts one is going to get are the TAFs and I find they are usually there or thereabouts
I disagree.

On the Met Office website, although there might not be a single source which is better than a TAF, there are several sources which, when combined together, can give a much more accurate picture.

A typical day's weather briefing, for me, might start with looking at the TAF. The TAF will forecast a large period of Prob40 Tempo bad weather.

Next, I'll look at the Metform 215 to figure out what's causing the weather. Perhaps a front, for example. Do the timings on the TAF co-incide with when the front is forecast to arrive? Very frequently not. Which direction is the front coming from? Are there any airfields in the region the front is coming from which publish METARs? If so, check the METARs from these fields, and find out where the weather is up to right now. Also find out what the weather is like in front of and behind any precipitation.

Last step is to look at the weather radar. This shows where the precipitation is right now, where it's going and how fast it's moving. From this, you can work out if it's likely to pass overhead your area at all (very often it's not), or if it is, when. Combine this with your knowledge (from the METARs) of what the weather's like in front of the precipitation, and you've got a good idea of exactly which parts of the day are likely to be flyable for VFR and which aren't.

This is far more accurate than relying on any one single source. And it's all free on the Met Office website (although I have to admit that the subscription weather radar is far better than the free one).

FFF
----------------

[stray10level]

http://www.pprune.org/forums/showthread.php?t=255047 Check out the thread for a very simmilar discussion.!

[IO540]

there are several sources which, when combined together, can give a much more accurate picture.


I agree!

This stuff should be taught in the PPL, but a) it isn't; b) the internet is not in the syllabus; c) many airfields have no public/accessible internet access; d) the CAA doesn't hand out laptops with GPRS to each new PPL :)