June July August Prognosis

[Michael]

Tony T > "So, now we don’t expect a cooler than normal winter, but we suggest the predictability is low, and therefore we don’t have any confident prediction. In the absence of that, we may as well plump for near normal, on the basis that normal is what we see most often (in theory – another digression could examine the idea that we hardly ever see a normal season, and the normal is just a statistical mean of two rather extreme types of season). I am not sufficiently convinced that the current SST pattern around NZ is especially warm, or especially likely to stay so through the winter. "

Before 1982-83 we never heard of El Nino or La Nina or the Southern Oscillation-of course it wouldve been mentioned earlier but since that event we either positive or negative but very little at neutral.
Perhaps its at times like now we get more variable weather rather than the constant SW of last May to December (Elnino) or in 1988-89 it was almost solidly E-NE for that summer but perhaps apart from the big ones ie 82/83 86/87 92/93 97/98 and last year Elninos and 88/89 96/97 La nina the rest are normal years.In the early 90's there was lots of high winter rainfall in the country and these were supposdly normal years and it looks at this stage for that now too.

[NZstorm]

I guess we rely on NIWA to come up with accurate predictions (if we wish to look ahead). But I am enjoying the mild weather this winter and hope it lasts. Last winter was very mild for June and then turned cold from early July with a bumper snow season on Ruapehu.

Freezing level over the NI today was 3000m.....no good for snow making at Ruapehu.

[Michael]

This June has so far been mild like June 1998

[TonyT]

I guess we rely on NIWA to come up with accurate predictions (if we wish to look ahead).

Not just NIWA. There are others having a crack at seasonal forecasting.

And NIWAs predictions are not that accurate. Nobodies are. My own feeling is that NIWA spend too much time predicting around the average and not enough time predicting away from the average.

Which I think is what Micheal was getting at, picking up on my comments about "normal" seasons. In some parts of the world, I dare say some parts of NZ, a climate average is meaningful in that the conditions are often near the average, so it represents a type of weather which occurs quite often. But for many parts of NZ the average is more statistical than real - rainfall in Canterbury is a very clear example. Few months experience rainfall totals near the long term average - most are either significantly wetter than average or drier. In other words the standard deviation for monthly rainfall in Canterbury is high. I suspect this is also the case for many other parts of the country, and not just with rainfall either.

Which means that many 3 month periods are near the long term average, and make NIWA look good in the analysis of their predictions, but which contain months which might be far from average.

[Manukau heads obs]

winter rain up this way is usualy reliable, and not much variation, but then again it has varied from 35mm for july to 340mm for july here in the past!
spring rain is quite reliable (i.e not too much variation), but autumn can be all the over the place here (auckland area), and so average rain for say april is a little bit meaningless

[GraemeWi]

Working with averages especially with as many variables as the weather would be very difficult. My daytime job is an analyst working with forecasts (not weather!). For a start the exceptions would have to be pruned off using simga smoothing - but then you would miss the big rain events or even the periods of very fine weather... then you would have to put some kind of seasonality into it, and then decide how many historical periods would be used...at this point I'm now sitting down fumbling for a packet of neurofen!

Phew I'm glad I'm not a weather forecaster!

G

[tich]

Seasonal averages can disguise monthly variations(remember the very cold July 2001 followed by a mild August), and monthly averages often disguise wide variations within a month. For example March was very dry for most of the month throughout Canterbury, yet parts of the area received normal or more than average monthly rainfall. Why? Good rain fell at the end of the month. Last winter was milder than normal for NZ as a whole, yet I recall it was also quite a snowy winter, with significant cold outbreaks, and a good ski season. (winter of 1999 was similar)

BTW the Unisys models (yes they look a fair way ahead, but are still useful in suggesting general weather patterns) predict a change to cold southwesterlies for NZ in about a week's time.

[NZ Thunderstorm Soc]

Yes Ben, I saw that too, but I didn't want to say anything. Wednesday next week for a bet.

JohnGaul
NZTS

[Michael]

BTW the Unisys models (yes they look a fair way ahead, but are still useful in suggesting general weather patterns) predict a change to cold southwesterlies for NZ in about a week's time.

We have those today

[ScottyD]

where can i find these unisys models?

[Andy]

http://weather.unisys.com/mrf/9panel/mr ... el_aus.gif

http://www.atmos.albany.edu/weather/gop ... ly_mrf.gif

http://www.weatherzone.com.au/security/processLogin.jsp

Here is few, should keep you busy for a while.

[Andy]

Some great and interesting links can be found from here


Steve Rosie home page - http://met.net.nz/index.html

[ScottyD]

Cheers, Thanks guys.