Jump to content


  • Content Count

  • Joined

  • Last visited

Posts posted by WhiteFox

  1. 40.2 mm. Looking at the lastest radar, may get up to 45mm before midnight, after which it looks like fizzling out here. Very impressive event nonetheless, and that represents almost 25% of my rainfall for the year to date!


    Flood alerts now issued for the entire length of the  Loddon as expected.

  2. Flood warning out for Emm Brook in Wokingham; property flooding is expected. Not surprised really as 25mm is generally enough to cause it to flood ( although not property), and there must have been more than double that there today, falling on some very dry ground. Would also expect the River Loddon to flood over the next day or so as well.

  3. Just now, kold weather said:

    Its a pretty impressive summer rainfall event, especially for the SE and I'd be interested to see the return rate for this much rainfall in the SE section, especially for a summer month. From a brief scout its rare to go above 25mm for a daily average, but I'd guess we have a fair shot at going above that, especially as we still have a good 90 minutes left of the day.

    I'm thinking that possibly July 2007 would be the closest?

  4. 7 minutes ago, Daniel* said:

    Crikey for some of north Kent some very heavy rates chucked in there, moving at a snail pace and backbuilding severe flooding potential high.

    Yes. Quite a serious convergence line seems to be occurring over Kent between Rochester and Maidstone heading West. Showing rates of 75mm or so per hour. That's pretty serious!

  5. 32 minutes ago, Eastnorthwest said:

    Unreliable model madness will kill positive debate if things don't improve in the future ,what images shows is what I've been saying all winter ,anything on the models beyond 120 hours can't be relied on is now sometimes down to 96 hours which makes the forecasters job a nightmare ,I look forward to the day when the model performance improve s

    I disagree! What you need to do is tune out the noise. The longwave pattern is identical in the two charts that @West is Best posted. The surface analysis is different because it always will be at that range; smaller scale features are enormously complex to forecast.

    As far as forecasting goes, IMHO, you can only reliably predict PPN amounts and temperature within a range that broadens as you move further out in time (yes, there are times when the range narrows or widens, generally when a pattern change is occurring). Beyond that, you are simply looking at likelihoods and probabilities; no scientific forecaster will be willing to say if it will rain and what the temperature will be at a particular time on a date more than six days in the future. They may forecast above or below average conditions in a range of categories for a period of time, but I don't think that models going past 6 days were ever intended to be taken literally, but rather as a guide.

    Forecasts do bust, with this winter being a prime example. But, on the whole, NWP models are pretty good when it comes to forecasting PPN (not type!), wind direction and overall conditions out to the six or seven day range. Going further out, I only consider a forecast to be a bust if a predicted pattern change hasn't happened or an unforeseen change did happen (e.g. mobile to blocked) and anything beyond day 10 is simply a possibility rather than a likelihood.

    A forecast hasn't failed just because a secondary low appeared on one operational output and then does not on the next: that is what I mean by noise.

    • Like 2
  6. 50 minutes ago, feb1991blizzard said:

    Your never going to be that accurate that far out - its a strongly +AO / NAO setup which it got right.

    Yes. It's often said, but look at the overall pattern: the 500mb forecast was largely correct, but it's incredibly difficult to forecast exact conditions closer to the surface at a longer range. Given the range involved, it's pretty good. If you'd looked at the 500mb forecasts you'd have said that the likelihood for the weather in ten days time was unsettled with a high probability of rain at times. And so it transpires.

    • Like 3
  7. 17 minutes ago, prolongedSnowLover said:

    It’s a parallel universe the GFS-P runs on ?

    I suspect that the GFSp needs more than a little tweaking. I've been following some of the recent storms in the US and its predictions for snow amounts in the north east have been utterly woeful: consistently far too high. To be honest, I haven't paid it much attention, but from the few charts I've seen in here it has a consistent bias towards producing cold and blocked output for Europe. That could be because people only post it when it shows cold, but I get the feeling it does this far more frequently than the live GFS.

    I've seen a couple of forecasters in the US say that the whole model needs to be deleted and to start again!

    • Like 1
  8. 2 hours ago, Nick F said:

    Thanks, I've been guilty on a number of occasions this winter of getting carried away and jumping on the SSW and/or MJO going through 'colder' phases being a definite force for good and WILL change the patterns without addressing the caveats too. Some do of course attach the caveats of caution to the possible benefits of the 'positive' signals, something I will certainly do from now on, after those positive signals have clearly not manifested in the troposphere (such as SSW) or have been overridden by other factors (MJO). Worth pointing out the drivers and teleconnections are still a useful tool, but being too confident in their use to predict more than 10 days ahead can easily end up with too many bust forecasts. This winter has been a good test for those longer range predictions.

    Some very good points and the one highlighted is the key. Making predictions based on signals is fine, but it's based on what little we know and always has caveats applied to highlight that they are based on probability and are not definite. I fear that sometimes people choose to ignore the probability side and choose instead to make snide remarks, but that should not out anyone off from making a forecast if they base it on sound analysis. 

    We still have a huge amount to learn and have very little data to work with; onwards and upwards!

    Just to add, it has seemed likely for quite a while now that high pressure setting close to the UK seems the most likely outcome and nothing has changed my view in that respect. The MJO moving into phase 8 means nothing on its own as I posted a few days ago; until the Pacific pattern changes from the current setup there is little to no chance of anything popping up to our north IMHO. I'm interested to know just why the Pacific has been so locked into the same pattern this winter.

    • Like 4
  • Create New...