So What Is It with Normal?
A Wonkish Controversy Brewing between Weather and Climate
By Ralph Petersen
From Wisconsin State Journal 8/9/2020
The Weather page of a typical local newspaper often displays an Almanac such as this one from the Wisconsin State Journal. High and low temperatures are reported along with the record values for the day. Also included? The term normal. But what does normal really mean these days and why does it matter?
First, some vocabulary. The World Meteorological Organization (WMO) defined normal (in this context) as the 30-year average between the years 1961-1990. Why the WMO chose a 30-year increment is not extremely well documented, but employing this time period provides a starting place from which comparisons can go forward. It also takes into account the fact that most people can remember the weather and how it made us feel roughly 30 years back.
In the United States, the National Weather Service (NWS) updates these 30-year averages once every 10 years. So, two decades ago normal reflected the average temperature between 1971-2000, a decade ago, 1981-2010, and at the end of this year, the NWS will re-calculate normal from the average temperature between 1991-2020. The 10-year updating cycle also provides a way to include some very local changes that have occurred through urbanization, such as the phenomenon of urban heat islands.
Fine, we shrug. It’s weather, stuff changes. No big deal. But are these updated normals making it harder to spot changes to our climate?
Climate change has been progressively working its heating magic over recent decades. This means the strength of real temperature shifts can be increasingly masked every time normal is re-calculated. It’s a little like trying to monitor one’s weight with respect to one’s normal clothing size in the age of vanity sizing—if the standard keeps changing how can we detect subtle differences?
Comparing today’s temperature readings with the normal of the past 30 years lets us confirm whether today is hotter than normal or not and by how much. In the 1990s, this approach also helped detect the first signatures of climate change. However, that was serendipitous, not by design, because the 1961-1990 averages reflect conditions before climate change started to have substantial impacts.
By comparison, when the new normal is reset at the end of 2020, it will likely be 2 °F warmer than its earlier predecessor and will no longer mirror much longer-term averages. This, in turn, will reduce the perceived impact of climate change by more than 50 percent.
A better way of measuring the continued impacts of climate change would be to compare recent measurements with a long-term historical average temperature made over a period before climate change impacts accelerated. An 80 to 100-year pre-climate change average (or historical norm) would unmask the sneaky heat increases accumulating from recalculations using recent normal. For us here in Madison, this means selecting dates before 1975 based on data from annual average temperature records for the Northern Hemisphere, North American and Wisconsin, as well as Lake Ice records for Lake Mendota and local Madison temperature records shown in the figure below.
This graph shows the annual average temperatures observed at Madison from 1895-2018 along with a curved line providing a smoothed summary of the full-period trend. Horizontal lines compare the long-term historical average temperature in Madison between 1895 and 1975 with 30-year average temperatures for each of the past five 30-year periods used to recalculate Normals. The results show that the average temperature from 1951-1980 was very close to the long-term average (cooler by about 0.1 °F), but each of the subsequent tri-decadal periods averaged progressively warmer than the long-term historical average by approximately 0.2 °F, 0.6 °F, 1.4 °F and 1.65 °F. As such, these repeatedly updated averages reflect the recent changes in our climate, not the conditions that occurred before the onset of climate change. Changing Normals will continue to minimize public awareness of the strength of the Climate Crisis as we progress into new decades.
Sound scientific data is our biggest ally for assessing success or failure as we try and curb climate change. As such, bias—vanity sizing—has no place in the mix. So with clear-eyed understanding, when we look at daily temperature reports, let’s not be fooled into thinking that the increasing effects of climate change in Madison are less than they really are.
What you can do. Take a walk in a wonk’s shoes with some math. An example: in 1998, the annual averaged temperature for Madison was 47.3°F. When we use the traditional 30-year method (1961-1990 in this case) to calculate the average temperature, it shows that Madison’s average temperature was 4.9°F above normal. Now let’s calculate how abnormal the same temperature was in 2012 using the updated 30-year average (1981-2010). Voilà, the temperatures for 2012 appear to be only 3.8°F above normal. Why the discrepancy? Because the new standard or normal unwittingly incorporates the 3-decades accumulation of climate change effects, masking some added warming and inadvertently reducing people’s perceptions of the strength of the Climate Crisis.
For more on computer models, climate data, and the politics of global warming, be sure to check out A Vast Machine by Paul Edwards.
Splash photo by Victor Rodriguez on Unsplash
Would you like to be notified by email when the latest weekly Swinging for the Fences blog post is available? Sign up here! Subscribers to Climate Corner are already signed up.
If you receive blog posts by email, our system automatically inserts “by Brook Soltvedt.” Brook is the webmaster, not the author of the blog.