A Forecast Extreme New Zealand Weather Cold and Snow Event
By Dr. Roger Pielke Sr.
While it is common to state that weather is not climatology, the reality is that climatalogy is composed of a collection of weather events over some time period. 30-year average temperatures and precipitation, for example, are two examples. NCDC has recently released its new climatological averages; e.g. see
Anthony Arguez, Russell S. Vose, 2011: The Definition of the Standard WMO Climate Normal: The Key to Deriving Alternative Climate Normals Bulletin of the American Meteorological Society Volume 92, Issue 6 (June 2011) pp. 699-704. doi: 10.1175/2010BAMS2955.1
It is also important to recognize, however, that extreme weather events are themselves part of climatology. It is such occurrences that often cause the most significant societal events. It is also useful to identify thes extreme events as there are often claims that extreme events, such as drought and heat waves, will become more common (e.g. see), or less common such as snowstorms (e.g. see).
The extreme snow event in New Zealand that is forecast this weekend is noteworthy in the context of climatology since, according to the IPCC-type predictions, such events should be becoming less common. The forecasts for this event are quite serious. The news agency TVNZ just released the article
Much of NZ braced for a polar blast
The text reads
Snow to sea level and blizzard conditions are set to hit New Zealand’s deep south, with snowfalls also spreading north.
MetService is warning of a polar outbreak in the deep south overnight tonight and tomorrow morning.
An extremely cold southerly outbreak is expected to bring snow to sea level over the south of the South Island early Sunday morning, the forecaster says.
A heavy snowfall warning has been issued for Fiordland south of Te Anau, Southland and the south and east of Otago including Dunedin.
Snow is forecast to spread to many other parts of the South Island and the lower North Island later on Sunday, it says.
Significant accumulations are likely in Fiordland south of Te Anau, Southland and the south and east of Otago.
The snow is expected to continue on Monday and into Tuesday.
The heavy snow is likely to cause major disruptions to traffic and make driving conditions very difficult, MetService warns.
Strong southerlies, gale-force on exposed coasts, with the cold temperatures will make the wind feel bitterly cold and create blizzard like conditions in some places, it says.
Farmers are being advised that stock may need shelter.
Road workers at the ready
Roading contractors are preparing to work around the clock this weekend clearing snow and laying grit.
The Transport Agency says it’s inevitable restrictions and some closures will be needed during the polar blast predicted.
Spokesman Andy Knackstedt says the number one concern is ensuring people’s safety.
He says people need to plan ahead, check the latest information, and think carefully about whether the journey is necessary or not.
This quite likely will be an historic extreme event for New Zealand, and is not in the direction of expected extreme events forecast such as presented in the news article in Cosmos by Oliver Chan titled
No snow, more drought, climate report warns
that I posted on yesterday in
source of the two images ECMWF
While it is common to state that weather is not climatology, the reality is that climatalogy is composed of a collection of weather events over some time period. 30-year average temperatures and precipitation, for example, are two examples.
=====================================================================
serious question…..
How did we get stuck on this 30 year thing?
Other than it being convenient………it has no real relation to anything
I live here in snowey Ottawa, Canada, where we are always ready for the worst that winter can, and does, throw at us. However, my sympathies to the people of New Zealand, who are probably simply not ready for this sort of thing. Please dont travel unless it is necessary. Stay safe and warm at home.
http://dictionary.reference.com/browse/Generation
Generation: The term of years, roughly 30 among human beings, accepted as the average period between the birth of parents and the birth of their offspring.
Dr. Pielke,
Like Latitude, I question the 30 year convention. With the global impact of climate modes like the PDO, it seems that 60 years should be the minimum. Is there any literature on the issue that could be used to confront the IPCC authors on the issue? — thanx
Latitude asks:
“How did we get stuck on this 30 year thing?”
Somewhere in the distant past in one of my probabiliity & statistics course, I remember something to the effect that to compute an average, at least 31 data values are needed in order to meet some criteria which I have forgotten. Can someone speak to this?
This quite likely will be an historic extreme event for New Zealand,
It’s historic in that if it will reach as far north as predicted areas that haven’t seen snow fall for many years will get a covering. But it isn’t all that extreme.
It’s winter, we’ve had heavy dumps of snow before, sometimes as late as October in the South Island. Hunker down, keep warm, make sure the stock are moved off the mountains and be prepared to accept loss of livestock.
Situation normal, nothing to see here 🙂
They say it is better to be safe than sorry when admitting their models could be off the mark. In other words the old warning from the warming crowd, can we take the risk that we are wrong.
When a group of people are consistently wrong in their work while supporting a political agenda at what point is the risk they are warning against ridiculous?
http://www.nzherald.co.nz/weather/news/video.cfm?c_id=10&gal_cid=10&gallery_id=120801
Good description of what is happening.
As a Kiwi, this weather report does not sound all that unusual – weather happens. Historically, NZ experiences it’s most severe lamb losses due to Summer snow and rain storms.
“Climatology” above is the wrong word; it should have been “climate.”
I have always thought we should use as long a period of data as is possible, so presumably, for the US, it could extend to 1900 or so. At the very least, we should endeavour to get it as long as 60-80 yrs to cover a full AMO cycle — PDO cycles apparently being shorter.
In fact, given the computer resources we have today, why shouldn’t the length vary with the station/location? If it has sufficient data for 30 yrs, let’s use 30 yrs by all means, but if it has data for 90 yrs, let’s use 90. So the weather guys
Above comment is in response to Latitude’s question on August 13, 2011 at 9:03 am.
Latitude says:
August 13, 2011 at 9:03 am
serious question…..
How did we get stuck on this 30 year thing?
Other than it being convenient………it has no real relation to anything
______________________________________________________
I asked the same thing on another thread a few weeks ago. The silence was deafening.
As far as I can tell, 30 years has no particular scientific significance whatsoever – it was merely a compromise figure that the WMO got everybody to agree to back in the day.
@ur momisugly noaaprogrammer says:
August 13, 2011 at 9:13 am
Latitude asks:
“How did we get stuck on this 30 year thing?”
Somewhere in the distant past in one of my probabiliity & statistics course, I remember something to the effect that to compute an average, at least 31 data values are needed in order to meet some criteria which I have forgotten. Can someone speak to this?
It varies, but there is a minimum recommended data set in order to fit a frequency distribution. The fewer data points available, the less accurate the fit (mean, median, mode, etc. ). Typically it is at least 50 points, and 500 or more is desirable. For a brief description of common distributions and fitting thereof: http://www.statsoft.com/textbook/distribution-fitting/?button=1 . There are many more than just the Normal distribution, and basing predictions on the wrong one can result in huge errors.
I think there’s little doubt that with 30 year flips in oceanic modulations that the 30 years has some usefulness, once you have several 60 year cycles of data available.
That 30 year average will be what you might expect within a warm/cold PDO for example. But it won’t necessarily be relevant to the other mode.
As an example, when as a young man I started reading about alpine snow patterns, all that was written was, in fact, for cold PDO. My experience for 20 years was that the ‘traditional’ patterns simply didn’t happen. Now that the pattern has flipped back, we are seeing ‘traditional’ patterns more often again. Earlier start to winter, less snowy springs.
The 30 year average has some relevance, but you need the start and finish dates right to know what it actually means. I suggest people look at the following 3 30 year periods for their own region and see what they find:
1. Cold PDO or AMO cycle.
2. 2nd half cold/1st half warm PDO/AMO cycle.
3. Warm PDO/AMO cycle.
That will indicate the susceptibility to decadal change, how much difference there is between the two modes and how sudden the change is.
A long way still to go before either the data records are long enough or human understanding great enough for ‘averages’ to have much meaning I’m afraid.
I agree that 60 years should be the minimum if you have the data. I noticed years ago in data for PDX before I had heard of the of the PDO or AMO that there was a 60 year cycle, 30 years of warming, 30 years of cooling. If we use only a 30 yr cycle someone will always ‘find’ that the climate is warming or it is cooling and run around in circles saying either “an Ice Age is coming! The Government needs to interviene NOW before we all die!” or “the world is over heating! The Government needs to interviene NOW before we all die!” I’ve lived long enough to see both.
What is it happening in the southern seas?, the south pacific has been agitated the last weeks, with waves bigger than normal along the coasts of south america.
My short-form understanding of climate change (the proportions of which are “natural” and “man-made” to be put to one side for the moment) is that the added heat in the system, if actually present, can be thought of as energy available to drive weather. In a nutshell, while the climate trends generally upwards in terms of temperature, the weather becomes more extreme, either in the cold or hot directions; the “spread” gets wider and the customary norms lessen. “Seasonal” weather becomes less expected, and historical norms (or at least of the last eye-blink of a century!) are exceeded.
Of course, it’s a mug’s game to base single weather events to predict climate, but if there is more energy in the system, one might expect wilder oscillations in weather events and unseasonable spells.
As I recall, in a standard distribution the standard deviation increases with sample size then starts to level off at around 45 samples at which point it becomes stable. The World Meteorological Organization (WMO) settled on achieving 2/3 of reaching a stable standard deviation because 45 years was too long to wait and 66% accuracy in such a chaotic system was seen as good enough to differentiate a trend and eliminate most year to year variation in averaging the data.
………we use 30 years because of satellites
more than that, and you would have to use some “trick” to blend the data
30 years is just convenient…and has no real relation to anything
Because that’s was once the time frame best suited for showing the most warming?
I’ve been watching this weather develop over the past couple of days. It’s nothing that we haven’t seen before, the only unusual part is that we had something similar only a couple of weeks ago.
Alchemy, as an interested weather observer, I think I can respond to your note in a small way. The AGW proponents have again been barking up the wrong tree by saying increasing temperatures result in more energy in the system, whereas in my experience, weather is driven in large extent by temperature differentials. In other words, if temperatures go up or down across the board, then not much will occur. If temperatures rise or fall markedly in one area compared to another area, then that causes extreme weather. As an example, storms and tornadoes at a cold front could be exacerbated either by warmer (than normal, whatever that is) air ahead of the front, or colder (than normal) air behind the front. Same applies for cyclones/hurricanes. It’s the method the atmosphere has of moving energy from one area to another to try and achieve equilibrium.
Latitude says:
August 13, 2011 at 9:03 am
> serious question…
> How did we get stuck on this 30 year thing?
I don’t know. I recall my “Climate of the US” (or whatever) book from the early 1960s which had 30 year averages, so the’ve been around for a while. The Coop program started in 1890, it may be that in the 1920s someone wanted weather averages and someone else said “Hey, we have this 30 year record.” Or maybe someone in the 1950s wanted weather averages and someone else said “Hey we have this 60 year record, but you only gave us enough money to average the last 30 years’ data. Hope that’s okay!”
The old paper weather record books were pages with 10 years of records with a month per page, with the days of the month from left to right, so they would fill out the decade long book then along the bottom of the page was a couple rows for the 10 year average for each date, The furthest right column was the averages for the month for each of the ten years, and the lower right block was the two average sums from each month that should match the totals for each day averages for the decade.
When questions arose about how it compared to the past, they just got out the last three decade long books and the current partially filled out book of records, laid them out together side by side on the desk and turned the pages simultaneously, it is easier to get an average from three sets of data than just two, and there was no more room for more books to be open at the same time, on a 3 X6 foot table. Simples.