I've never been a fan of the phrase "climate normals" to describe averages of past climate. It perpetuates the myth that climate is stationary and it suggests that an average is normal when weather normally fluctuates in a range. I grimace whenever I hear someone say "normally the high today is 64˚F."
I always wondered who started using this phrase and why. I don't know the answer to that question, but according to the National Centers for Environmental Information (NCEI), about 100 years ago the World Meteorological Organization (WMO) instructed member nations to calculate climate normals using 30 year periods, beginning with 1901-1930.
So, near as I can tell, we call them climate normals and we average over a 30-year period because we've always done it that way. That's simply not a good reason, especially now that we are living in a period of accelerating climate change and we know that for some regions, a 30-year average is likely insufficient for some variables like precipitation, which exhibit slowly evolving decadal or multidecadal variations.
This issue has come to the forefront this week because NCEI issued the new 1991-2020 climate normals. Below is a series of maps produced by Jared Rennie of the North Carolina Institute for Climate Studies showing how the 30-year temperature normals have evolved compared to the 20th century average. The warming trend is apparent.
 |
| Source: https://www.noaa.gov/news/new-us-climate-normals-are-here-what-do-they-tell-us-about-climate-change |
Precipitation is more complicated, however, as there's more regional variability and variations that occur over decades. For example, 1941-1970 was relatively dry over the southwest, 1971-2000 relatively wet, and 1991-2020 relatively dry again.
 |
| Source: https://www.noaa.gov/news/new-us-climate-normals-are-here-what-do-they-tell-us-about-climate-change |
For the southwest United States, annually averaged temperatures have increased at a rate of approximately 0.7˚F per decade since 1975. Thus, the 1990-2010 temperature normals are probably already about 1˚F lower than they should be. They reflect the past climate average, not the current climate average.
 |
| Source: NCEI |
For precipitation, the situation is more complicated. Instead of there being a clear long-term trend, precipitation in the southwest has exhibited wet and dry periods, with the latter predominating in the past 20-25 years.
The climate system is changing and while knowledge of the past is helpful, it is increasingly important that those of us in Utah recognize that these climate models are averages of the past climate and that the future will be hotter and drier, even if the average annual precipitation doesn't change that much.
I listened to a briefing from the NCEI on the new normals and it sounds like the source of the 30 year interval is more arbitrary than described on their site. Evidently, backs in the 30s, the WMO asked how long countries could come up with reliable climate records and the response was 30 years...
ReplyDeleteHow might you come up with a more appropriate normal? Adding in standard deviations or percentiles might key into the variability of weather, but I am clueless as to you would find the proper time length.
As I understand, "normal" is used because it's not a pure statistical average - there are certain adjustments and smoothing that are done. But it is confusing from a layperson's understanding of the word.
ReplyDeleteThe post above is interesting, because I would have guessed 30 years is due to the magical n=30 value in statistics.
I think the high/low numbers do provide value in answering "is it hot or cold for this time of year?" I find the precipitation/snow numbers to be less useful due to such high variability. For example, the DC snow average is skewed high by Noreasters; the median is several inches lower (or was for 81-10).