Before Posting Into The Latest Model Thread - Read Here

[johnholmes]

Written by John Holmes although not now upto date as the top part is in reference to posts yesterday - the main bulk is, please take time to read it before posting into the model thread.

First of all I would not describe max temps over the following 4-5 days of 2C to 5C over the country as 'cool' Cold or rather cold might be a better description. My average max is around 5.9C so 2-4C in Met Office terms is rather cold to cold.

On to your second comment about zonality and I'll try to answer that by using the 06Z charts.

I grant that there is a degree of; I would call it, 'mobility' over part of the chart, no more than that.

On the T+168, really the limit of 'real' forecasting with any hope of success, it shows

Yes, a degree again of mobility but its hardly 'raging zonality' in my view. But I suppose you could say your zonality is there, no dispute about that.

If one looks at the 850mb temps, or indeed the surface temp mins over the UK, they show below zero over the whole country with the -5C making steady eastward progress, along with the winds swinging into the north west and surface temps of 1C to 4C for most(only the far south west being higher). If it were truly zonal then 850mb temps some 5 or 6 degrees higher would be more normal, possibly even 10C higher. Also the min temps would show higher values. This has been the situation for much of the period since late November, not really cold not really mild, quite different to previous years, in my view.

We could then look into F! which so many on this forum cannot wait to do on EVERY run. Sometimes to whoop with joy sometimes to burst into tears. What all of should be looking for are signs of a 'signal' shall we say that this winter, unlike the previous 7 or 8, maybe more, that the actual synoptics which we keep getting since mid November is very different from these recent winters.

In my earlier post I quoted a cold spell before winter (late November), and a couple of short coldish spells up to mid December. Then the forum went into mourning as it was the end of winter, then what did we get? One of the coldest spells in many years, and for quite a swathe of the country, the most snow for a while. So already we have had more 'winter' than the winter of 2004-5 gave us, yes including the late February situation. We are now almost into real time winter again over the next 4-5 days. Granted not much snow about but well below normal temperatures. So that is now 2 main cold spells with 2-3 other short ones. The facts show that December was almost average, in terms of temperature, November not far behind, January looking as if it will be not far off normal. Its too far ahead to say how much above or below, and so many on this forum still insist that its not winter and cannot be winter.

There was an excellent post by GP about connecting various things into the Pacific along with AO etc, and all SEEM to suggest that this winter will end up as the coldest since, who knows. Maybe the Met Office predictions will not be far off, or indeed our own Ian's LRF.

Why am I blathering on at such length? Because I spent my whole professional life in Meteorology, both as a trained observer and a trained forecaster, from junior into senior positions. My forecast career began before computers were used as operational tools. I remember when we began to issue reliable forecasts out to T+48 then T+72, and the issue of charts (similar to the Fax charts we have free access to) out to T+120 with huge excitement. The Met Office still does not do public forecasts beyond that time scale (often is the comment that the Sunday morning outlook only went as far as Thursday!!). Why? Because in their view to try and forecast with any degree of accuracy beyond that is not yet proven. Yet we, on this and other web sites, regularly nash our teeth and wail about the 'rubbish' that GFS or ECMWF or whatever at T+180, even to T+384 show. Come on folks, get real. NOAA use the GFS as a research tool, as do other centres with their own output, to TRY, I stress TRY, to get forecasts out on that time scale to the same accuracy as T+96 is now. I know BBC and to a lesser extent the Met Office regularly get criticised for the poor standard of their forecasts, but often quite unfairly if you looked objectively at what was actually said. As someone who spent 5 years doing public service forecasts trying to get the correct message across I can assure you it is enormously difficult. We all see and hear, to some extent, what we want to see and hear.

Finally back to GFS, I cannot stress enough, that at beyond T+168 it really is a forecast research tool, which we all get for NOTHING, so let us all be just a bit more rational in our postings. Also, do remember use just one run each day and compare it to the same time each subsequent day. Not foolproof but one heck of a lot better than chasing your tails and each run then either shouting hooray or boo.

Final end of sermon. I do hope it is read and taken, not as a major criticism of anyone, but an attempt to point you in the right direction.

Regards

John

[Paul]

This is also relevant - a statement from the NCEP about using the GFS beyond 7 days.

Statement on the Use of GFS Output Beyond Day 7

There is some concern about NCEP model variability from run to run, particularly at extended ranges beyond 7-8 days. The American Meteorological Society Statement adopted on January 14, 2001, titled "Statement on Seasonal to Interannual Climate Prediction" (Bull. Amer. Met. Soc., 82, 701), states, in part, the limit of predictability for day-to-day weather changes for deterministic forecasts is "in the range of 1-2 weeks." NCEP agrees with this statement and notes that for current state-of-the-art models, the limit of "useful skill" is about 7-8 days. Beyond that range, deterministic numerical weather predictions are characterized by little or no skill and suffer from large run to run variability. However, ensemble or probability forecasts may offer improved predictive skill.

Another AMS statement adopted January 13, 2002, titled "Enhancing Weather Information with Probability Forecasts" (Bull. Amer. Met. Soc., 83, 450), praises the value of probability forecasts and endorses ensemble model forecasts as a means for improving skill levels in the extended ranges while quantifying the level of certainty (uncertainty) in any forecast. NCEP agrees with this statement.

NCEP is working to making these ensemble forecasts products more readily available. In the meantime, we recommend using the GFS, or any global forecast system, with care beyond day 7 and recognizing the inherent variability in long range deterministic forecasts. Also recognize that probability forecasts, while providing much more flexibility and value to the user, require the user be trained to properly understand and use the new product suites based on ensemble model output. Some relevant training material can be found at http://www.hpc.ncep.noaa.gov/ensembletraining/