[Dave K]

Tried the search function couln't find earlier thread where this has been discussed though obviously it has probably been done to death ;)

It's all down to the WMO apparently, according to this article in Norwegian at http://www.yr.no/nyheter/1.7378341  

Here is a translation, of course it still rasies many questions

---

When meteorologists talk about normal weather, it is to the average weather in the period 1961-1990.  This seems odd to yr.no reader Terje Kristiansen. He asks:

- Why is the average temperature based on what I would call historical data (1961 - 1990) and not the actual average past 30 years?

 

A normal is an expression of the average value over a specific 30-year period. It is internationally determined that the normal period for which we use today is 1961-1990. This average alone is used as references to look at changes in temperature and precipitation. To say something about the changes from year to year  it requires that the reference value will be the same.

It is the World Meteorological Organization which determines the normal periods. The previous normal period was 1931-1960. If we continue to follow this system in 2021, we will adopt a new normal, namely, the average values for the period 1991-2020.

Since the normal period is determined internationally as agreed by most meteorological institutes around the world they use the same normal period. It makes it easier to compare precipitation and temperature changes across borders.

Although the period 1961-1990 has been selected as the default period, there is nothing to prevent calculation of the average values for other periods. For example, the seNorge [ data services division of Norway Meteorological Institute  ] also provide normal values for the period 1971-2000.

[PK2]

I've read much the same on the Ilmatieteenlaitos.fi (Finnish Met Service) website. (my underlining)

Quote

A new period 1971-2000 is useful in climatological issues
The Finnish Meteorological Institute among many other meteorological institutes throughout Europe has published a new climatological 30-year summary concerning the period of 1971-2000. The need of updated climatological 30-year statistics was obvious already in the end on 20th century. However, the former period 1961-1990 will still remain the official normal period defined by WMO. The next official normal period is 1991-2020 but until then we probably see the statistics for the period 1981-2010 as well as we now see for the period 1971-2000.

[skanky]

UKMO:

Quote

The World Meteorological Organization (WMO) requires the calculation of averages for consecutive periods of 30 years, with the latest covering the 1961–1990 period. However, many WMO members, including the UK, update their averages at the completion of each decade. Thirty years was chosen as a period long enough to eliminate year-to-year variations.

8.4 here describes it: http://www.wmo.int/p..._practices.html

It was probably decided when the climate was (perceived to be) stationary (see http://www.aip.org/h...mate/index.html for that).

[Andy Mayhew]

In climatological terms it doesn't really matter what period is used so long as it is consistent.

In terms of perception of day to day weather, more recent periods might be more appropriate - ie comparing 2011 temps with the 1981-2010 average since that what people better remember.

[Ian Williams]

UKMO Also give some 1971-2000 averages

http://www.metoffice.gov.uk/climate/uk/averages/19712000/

http://www.metoffice.gov.uk/climate/uk/about/methods.html

[John Robert Mellor]

Where I can, I always base my comparisons with figures I have extrapolated from the official 1971-2000 on my website. I also compare to more recent timescale of 1991-2005, again against figures that I have extrapolated from the official returns. Yep I know the last set wouldn't be considered a long enough series, but it is more up to date.

[Dave K]

Quote

The Finnish Meteorological Institute to adopt a new normal period for climate statistics

8.2.2012 13:20

The Finnish Meteorological Institute has calculated climate statistics for a new normal period based on weather observations made between 1981 and 2010.

The recent statistics describe the prevailing climatic conditions and form an important base for interpreting the current weather conditions and for monitoring the weather. Climate statistics are also used in research and decision-making, especially in sectors dependent on the weather and the climate.

Climate statistics are calculated globally using weather observations ranging over a 30-year period. The official normal period set by the World Meteorological Organization (WMO) is 1961–1990. However, it is recommended that the statistics be updated, whenever possible, once every ten years. In Finland, the previous update took place ten years ago.

Full article in English: http://en.ilmatietee...-release/476066