By Paul Homewood
https://www2.ucar.edu/atmosnews/news/1036/record-high-temperatures-far-outpace-record-lows-across-us
Every so often, the hoary old chestnut of record daily temperatures is wheeled out, as evidence of global warming. The above chart is from the NCAR study by Gerard Meehl in 2009, and the NCAR Press Release stated:
Record high temperatures far outpace record lows across U.S.
November 12, 2009
BOULDER—Spurred by a warming climate, daily record high temperatures occurred twice as often as record lows over the last decade across the continental United States, new research shows. The ratio of record highs to lows is likely to increase dramatically in coming decades if emissions of greenhouse gases continue to climb.
“Climate change is making itself felt in terms of day-to-day weather in the United States,” says Gerald Meehl, the lead author and a senior scientist at the National Center for Atmospheric Research (NCAR). “The ways these records are being broken show how our climate is already shifting.”
This graphic shows the ratio of record daily highs to record daily lows observed at about 1,800 weather stations in the 48 contiguous United States from January 1950 through September 2009. Each bar shows the proportion of record highs (red) to record lows (blue) for each decade. The 1960s and 1970s saw slightly more record daily lows than highs, but in the last 30 years record highs have increasingly predominated, with the ratio now about two-to-one for the 48 states as a whole.
The study, by authors at NCAR, Climate Central, The Weather Channel, and the National Oceanic and Atmospheric Administration (NOAA), has been accepted for publication in Geophysical Research Letters. It was funded by the National Science Foundation, NCAR’s sponsor, the Department of Energy, and Climate Central.
If temperatures were not warming, the number of record daily highs and lows being set each year would be approximately even. Instead, for the period from January 1, 2000, to September 30, 2009, the continental United States set 291,237 record highs and 142,420 record lows, as the country experienced unusually mild winter weather and intense summer heat waves.
A record daily high means that temperatures were warmer on a given day than on that same date throughout a weather station’s history. The authors used a quality control process to ensure the reliability of data from thousands of weather stations across the country, while looking at data over the past six decades to capture longer-term trends.
This decade’s warming was more pronounced in the western United States, where the ratio was more than two to one, than in the eastern United States, where the ratio was about one-and-a-half to one.
The study also found that the two-to-one ratio across the country as a whole could be attributed more to a comparatively small number of record lows than to a large number of record highs. This indicates that much of the nation’s warming is occurring at night, when temperatures are dipping less often to record lows. This finding is consistent with years of climate model research showing that higher overnight lows should be expected with climate change.
Remember that last paragraph,as we move on later. But first, a quick recap of the huge flaw in Meehl’s exercise.
The data starts in 1950, and consequently picks up all of the daily low records set during the much colder 1960’s and 70’s, but fails to pick up the daily highs during the 1930’s and 40’s. This flaw was so significant, and obvious, that I am amazed the paper ever made its way past review.
http://www.ncdc.noaa.gov/cag/time-series/us
Nevertheless, we can now move on. WUWT mentioned the other day that CDIAC, The Carbon Dioxide Information Analysis Center of the Department Energy, have introduced a new tool, which crucially uses data right the way back to 1911. It also only uses USHCN stations, regarded as high quality, and excludes any that don’t have complete, or nearly so, records back to 1911.
The data only goes up to 2010, so there will no doubt be a batch of record highs in 2012, and record lows last year, not included.
Interestingly, CDIAC introduce their new interface thus:
Like politics, you might say that all climate is local. As researchers seek to help the public better understand climate and climate change, a sensible approach would include helping people know more about changes in their own backyards. High and low temperatures are something that all of us pay attention to each day; when they are extreme (flirting with or setting records) they generate tremendous interest, largely because of the potential for significant impacts on human health, the environment, and built infrastructure.
In other words, they want to “educate” the public, who remain stubbornly sceptical! However, it appears they may have shot themselves in the foot
.
Southeast Region
Greg Kent, in his essay on WUWT, put together some graphs for the CONUS, which did not look as scary as the original NCAR study, but I wanted to delve a little deeper, so decided to take a closer look at the thing region by region, starting with the Southeast (encompassing VA,NC,SC,GA,FL & AL). Altogether, there are nine regions in the CONUS.
Within the region, CDIAC list 37 stations, out of the national total of 424. The interface offers these screens for each.
So, using the record highs and lows, I have built up a database for the whole region. (The 1910’s start in 1911, and so on).
Figure 1
Figure 2
As far as daily highs go, there are 17730 records/ties, an average of 479 per station (i.e suggesting 113 ties). The total number of records set between 2001 & 2010 was 1222, much lower than the average of 1773. Daily lows were also below average in the last decade, with 1094, against an average of 1604.
We can also look at the ratio of highs and lows, in Figure 3, showing that they have been pretty much in balance since 1991. The imbalance during the warm 1920’s to 50’s, and cold 1960’s to 80’s, is also evident.
Figure 3
I don’t want to make too big an issue of these figures, because the Southeast is generally accepted to have shown the least warming of any regions during the last century. We can only get the full picture when I have worked my way through the other regions.
But, here things started to get a bit interesting.
Urban Heat Island Effect
As I was transferring the data, I noticed that some stations had much higher ratios of highs to lows. I only really cottoned on when I noticed it for Charleston, SC, which you may have realised is a rather large city!
So, I backtracked, and entered the population data, using the GISS database. GISS either show <10,000, or the actual population, so I have split the Southeast database into two, one for “rural”, or less than 10,000, and the other half for urban, or more. The numbers split almost equally, with 18 rural, and 19 urban.
The difference between the two sets in Figures 4 & 5 is startling.
Rural stations now show more lows than highs, with a ratio of 0.86 highs to lows during the last decade. In contrast, urban stations show a ratio of 1.61.
Figure 4
Figure 5
It is clear from analysing the numbers that it is record lows, and not highs, which are making the urban and rural ratios so different, as the Table below shows.
| 2001-10 | Decadal
Average 1911 to 2010 |
% of average | |
| Record Highs | |||
| Urban | 593 | 911 | 65 |
| Rural | 629 | 862 | 73 |
| Total | 1222 | 1773 | 69 |
| Record Lows | |||
| Urban | 367 | 813 | 45 |
| Rural | 727 | 791 | 92 |
| Total | 1094 | 1604 | 68 |
Rural stations are actually showing a slightly higher number of record highs than urban ones, relative to the average. But, with daily lows, rural sites have posted double the number compared with urban.
It is hard to find clearer evidence, that night time temperatures have been biased upwards by the Urban Heat Island effect in recent years.
OK, these are just daily records for one region. What does any of this tell us about annual trends for the country as a whole?
The second part of the question will be answered in due course, once I have run through the other eight regions. But the following comment in the NCAR press release, introducing Meehl’s work is significant:
The study also found that the two-to-one ratio across the country as a whole could be attributed more to a comparatively small number of record lows than to a large number of record highs. This indicates that much of the nation’s warming is occurring at night, when temperatures are dipping less often to record lows. This finding is consistent with years of climate model research showing that higher overnight lows should be expected with climate change.
So NCAR admit that much of the warming has occurred at night, and believe that this is caused by “climate change”. The evidence from the Southeast suggests that this is not the case, and that night time warming is largely the result of UHI.
It is also worth bearing in mind, that many of the stations, which I have labelled as rural, could still be small towns with their own UHI effect, given that GISS lump all sites with populations of less than 10,000 together. Lexington, VA is a good example of this. It has a population of 7000, and a ratio of highs to lows during the last decade of 2.82.
Clearly, a lot more work needs doing, but this exercise suggests a lot of serious questions need to be asked about the true impact of the Urban Heat Island effect.
References
The CDIAC interface is here. Go play!
http://cdiac.ornl.gov/climate/temp/us_recordtemps/dayrec.html
Isn’t it the case that models suggest most of the ‘global warming’ should take place during the day rather than the night?
Very interesting Paul.
One thing that struck me straight away is that there seems to be a twenty year lag between the highs and the lows. This is even clearer in the rural data in fig4, less to in urban.
No idea why that should be or if it’s a pattern that holds up for different regions , just an observation.
Whenever I get into a discussion with an alarmist about topics like this, I always say “Let’s take the starting point back to where the long term curve changed slope. I choose the year 1650 or anywhere near the bottom of the Little Ice Age. Typically the climate alarmist calls foul because they know without a good cherry picked starting point, their arguments do not hold up.
“I am amazed the paper ever made its way past review.”
The buddy system that the climate “scientists” use instead of a real Peer Review, means that any old crap will be published.
If you go to an ordinary table of state high temperature records, just about one half of them were set in a single decade — the 1930s. Outside of the 1930s, they are pretty uniformly distributed, with nothing interesting about the 2000s compared to the 1910s (for example). This alone is an enormously damning statistic, as it suggests that all or most of the alleged warming of the CONUS is an artifact of the computational process. Lack of accurate UHI correction is the least of it.
I think that accurate treatment of the UHI correction is around the corner, however. There are systematic studies underway (that still do not utilize the available resources particularly well) that offer strong evidence that the UHI correction to GISS is terribly implemented, more often than not having the wrong sign (which should nearly always be a cooling of the more urban present compared to the past, but somehow ends up the other way more than half the time in GISS) and of course there is no UHI correction to HADCRUT4, so it is absolutely certain (in my opinion) that there is a systematic warming error across the entire range it covers to where it is at least a few tenths of a degree C by the last two or three decades (probably larger than their claimed method error for that time period, that is. Since the entire warming indicated by HADCRUT4 from the 1850’s to the present is itself considerably less than 1 C, knocking 0.2 C off of the right hand side reduces it to perhaps 0.5 C plus or minus a couple of tenths.
The pattern of warming from 1910 to 1945 greatly resembled the pattern of warming observed from the late 1970s to the present, right down to press reports that the Arctic was melting and Arctic sea ice all but disappeared. I am deeply skeptical of any claim that we can accurately know global average temperatures and even more skeptical that we can know global average temperature “anomalies” relative to the present from that era, but it is reasonably probable that at least some fraction of the global warming claimed for the last 35 years is really local warming, specifically warming driven primarily by UHI effects as the world’s population more than doubled over the interval. For better or worse, people measure(d) temperatures primarily where they live(d), and wherever they live(d) in the past, far more people live there in the present. Then there is the utter imprecision of global estimates of sea surface temperature for most of the climate record. Even with ARGO, HADCRUT4 only claims 0.15C accuracy (probably egregiously). Pre-ARGO? Who knows? Only that the error (systematic and otherwise) is certainly much larger, especially prior to perhaps 1950 after world war II and the advent of modern aviation and navies opened up oceanic travel worldwide.
rgb
rgb
I think it’s easy to miss the point here, I would expect more record highs than record lows to be recorded, always. It’s just a question of thinking where would you expect record highs to be recorded and where would you expect record lows to be recorded, then ask yourself where the recording instruments are most concentrated. In cities.
Well, we can add “they’ve taken into account UHI” to the list that also contains “they’ve taken into account the sun”, and “the Little Ice Age was an extremely localized phenomenon” that warmists will believe with all their hearts.
Anything else?
Powerful analysis, simply stated. Such detailed, long-term, thorough-going analysis belongs in permanent archives as baseline refutation of the contentious over-generalizations typical of loose-leaf AGW catastrophists.
I question if comparing record high/low temps by decade would even have any real meaning. Imagine creating a database for any stable system over time. The first year a 100% percent of all measurements would be records because there would be the nothing to compare it too. In the second year the number of records would decrease from the previous year [because] the amount of data to compare it to would increase and the same would happen for the fourth year and fifth year and so on. Eventually the full range values of the system would have been experienced and the number of records would reach some stable value (0 if we are not considering ties). From looking at your charts that seems to be whats happening here, although the 80’s seem to throw a wrench in my observation. Any thoughts?
It would be entertaining to re-issue the press release of Meehl with the same reasoning using the recent decadal numbers – his certainty of global warming becomes NCAR’s reasoning for global cooling. Hey, use the same logo with some small changes.
On a mathematical note, one should understand that new records should get fewer with time. If the variations were absolutely random, the number of records one can expect is approximated by Ln N where “N” is the number of years in the series and the first year in the series is counted as a record. This works very well for major floods of rivers. In a hundred years the number of record highs at a station would be Ln 100= 4.6, about 5. So how many should we expect in 500 years? Ln 500=6.2 !!!! only one more. Now to do this properly, when you have a heat wave and progressive days increase a little, you should really only count this as one record. Perhaps select the highest within a week of highs surrounding the high. It is also a better measure of what they are trying to communicate.
If random and we are counting all the records (with my caveate) the NCAR figure for highs in the 60 year period would be Ln 60*1800 stations. Ln 60 is 4, so we should have expected 7,200 record highs and not 291,000. This is 40 times my figure. If we were to triple my figure to allow for records within days of each other at a given station (picture a week’s heat wave) then we would have ~22,000. Lets triple this figure to allow for fiddled records that pushed a number of whole years from below 1937 which still held the US record high as late as 1998 until it was rejigged by Hansen at GISS precisely to deep six 1937 temperatures. That’s 65,000 and we could add more because of pushing the pre 1945 temps down to make the recent rise steeper and thereby moving a bunch of those records up into the 1950s + years. Okay, that still doesn’t do it. Random falsified.
Also to be considered is how past record highs/lows have been changed, not broken, changed. (Adjusted if you prefer.)
Find an online list of your area’s record highs/lows then copy the address into the search part of TheWayBackMachine (http://archive.org/web/web.php) and compare.
For example, where I live the 2007 list says there were 0 new record highs or lows set between 2002 and 2007. Yet the 2009 list says 12 new records were set in that same time period.
PS The list for 2007 for my area is not listed but I copied it “live” in 2007.
if the earth is in an inter glacial warming period then one would expect warming. The co2ers have to show how much extra co2 induced warming they think there is. Why all warming HAS to be co2 is very odd assumption during an inter glacial warming period.
decontextualising any study from the ice age cycles to scream ‘it co2 warming’ is just seeking headlines in a media hungry for disaster movie stories.
OOPS!
That should read:
“For example, where I live the 2007 list says there were 0 new record highs or lows set between 2002 and 2006.”
Thank goodness there is this double standard in science that holds medical research to a much higher standard for peer review and product or procedure review. Otherwise the life expectancy would be about like that of isolated hunter gatherer groups. Climate scientists with defective methodologies in turn would have to churn out papers faster, before they all croak.
I see a different sort of error:
“If temperatures were not warming, the number of record daily highs and lows being set each year would be approximately even.”
Prove it. There is no definite or even logical connection between the number of record highs or lows and general warming or cooling. The climate might only become more variable with time in preparation to make a step up or down (something that happens repeatedly, as records show). The number of record single day highs and lows tells us nothing about the average temperature for a month or year.
To support a claim of general warming one would have to first calculate the average temperature for the period for which the claim was made. “Temperatures warming” is a very different thing from number of ‘record high temperatures for a given day’.
For a proper (useful) metric one would want a time-weighted record of daily temperatures, say by the hour, and if you want to make a engineer-quality assertion it should be temperature and absolute humidity-corrected – i,e, heat content, and demonstrate a net change in enthalpy.
“This decade’s warming was more pronounced in the western United States, where the ratio was more than two to one, than in the eastern United States, where the ratio was about one-and-a-half to one.” (italics added)
There was no general “warming”, there was an increase in the number of record high’s recorded (so runs the claim). No warming trend was demonstrated so why the claim? It is an unsupported leap over a large canyon of required proofs to jump from a record high or low to change in average surface temperature.
Should not more records be set earlier in any series than later, since they all start at zero? As time goes by the time between extremes should grow, unless there was a trend. Then there would be fewer extremes on one side than the other.
As soon as I saw the first graph (which started at 1950) I immediately asked myself where are the 1930s?
And so on…………….
Resourceguy says:
March 14, 2014 at 1:45 pm
Thank goodness there is this double standard in science that holds medical research to a much higher standard for peer review and product or procedure review.
===============
” Thousands of medical papers cite Wikipedia “, study says
http://news.nationalpost.com/2014/03/12/its-against-all-principles-of-scientific-reporting-thousands-of-medical-papers-cite-wikipedia-study-says/
Heat and wild weather.
There are so many things that can potentially affect record lows that the statistic is practically useless on its own. For example increase cloudiness or haze at night, increased wind at night and so on.
I personally enjoy the fact that every time we have a cold snap in winter in Perth I can sit back and listen to the climate people talk about how we haven’t had sufficient rain. Well duh, the reason we had a cold snap was because of cloudless nights. Needless to say whenever it rains a lot in winter, the same people go on about how it’s so warm.
The 1930s was a bad decade for warm weather.
======================================================================
That’s funny. Since they changed their format for “Weather on the 8’s” back in November I haven’t noticed them even once list my area’s record highs and lows for that day. I wonder why?
You know what? I turn 70 this year, and like most people my age, I’m heavier than I was in high school, about 50 pounds heavier. Now my weight goes up and down some, but the record highest I ever weighed in my life record is in my last decade. Wow I bet you’re all surprised by that factoid right? OK, you’re not surprised. Well then no one should be surprised, and it certainly isn’t news that there are more high temperature records during this last decade where the average temperature of the world is warmer (+0.8°C) than it was 165 years ago.
During the mid 1930’s in Topeka, Kansas there were periods of several consecutive days when the temperature remained above 100 degrees all day and all night. I have tried to locate these records without any luck. I can remember a couple of days in the 1950’s or 1960’s that the low temperature for the day was above 100. Has that type of record been recorded anywhere since the 1930’s?