Is America Burning?

Guest Post by Willis Eschenbach

For the last couple of decades since I started seriously studying the climate, I’ve been hearing Americans of all kinds claiming that the United States is suffering terribly from “global warming”. Over and over, from American educators, government officials, and media, the claim is made that the US is heating up fast, and that the dreaded and endlessly warned of US Thermageddon is just around the corner.

So I decided to do some research. There is a bunch of gridded surface temperature data out there, typically on a 1° latitude by 1° longitude grid. So first, I had to figure out just which gridcells are in which countries. That involved what we used to call a “SMOP”, a “small matter of programming”, which only took about a day of locating the country borders data and then actually writing the code to convert it to a usable form … but at the end of it, I knew which gridcells are in which countries.

Then I pulled up the temperature information from Berkeley Earth and from the CERES satellite data and graphed it up … here’s the result:

Figure 1. Temperature trends for what in Alaska is called the “Lower 48”, meaning the US less Alaska and Hawaii.

Cooling … not warming.

Cooling.

Now, before anyone starts yelling “But that’s just the contiguous US, not the globe!”, yes, I know that. I’m just saying that if you think the US temperature has gone up over the 21st century, you’re wrong. The US has cooled over that time.

And this reveals a deep truth, which is that if it were not for accurate thermometers, hundreds and hundreds of US surface temperature station sites, and satellites, we’d never know if the US had warmed or cooled over the last two decades … the changes are far too small and too widespread for our human senses to register. Millions of US citizens have been firmly convinced that the US has been warming in the 21st century, when in fact it has been cooling.

That’s it. Short and sweet. And no, it’s not an April Fools joke, it’s actual data.


Not only is it April Fools Day today, but it’s also my gorgeous ex-fiancee’s birthday, so we’re going for a lovely bike ride for a birthday treat.

My very best wishes to all,

w.

4.9 49 votes
Article Rating
Subscribe
Notify of
185 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Rick K
April 1, 2021 6:08 pm

Always enjoy your work, Willis. Thanks!

DHR
April 1, 2021 6:09 pm

Your thoughts on the Climate Reference System data (showing the same thing) would be appreciated.

DMacKenzie
Reply to  DHR
April 1, 2021 6:23 pm

USCRN graph here
https://www.ncdc.noaa.gov/temp-and-precip/national-temperature-index/time-series?datasets%5B%5D=uscrn&parameter=anom-tavg&time_scale=p12&begyear=2004&endyear=2020&month=12

but 2021 drops a big amount and doesn’t plot Feb. and March so dred web blues have struck the NOAA graphing routines….

Last edited 12 days ago by DMacKenzie
Weekly_rise
Reply to  DMacKenzie
April 1, 2021 6:49 pm

The USCRN data show a warming trend. The CRN starts in 2004, but is almost identical to ClimDIV, which does go back before 2000. I’ve plotted both together with their linear trends.

comment image

I would be very curious to know how Mr. Eschenbach explains the discrepancy between his results using Best and Ceres and the data from the NOAA.

Last edited 12 days ago by Weekly_rise
Weekly_rise
Reply to  Willis Eschenbach
April 1, 2021 7:24 pm

I agree with you that the various temperature products show differences in trends arising from methodological or instrumental differences. But it surprises me very much that your BEST trend is so different from ClimDIV, since BEST is, to my knowledge, using the same station network for the US. I unfortunately don’t have the coding chops to download the BEST gridded data and try to work out the differences myself in any reasonable time frame. Here’s an image from Berkeley Earth comparing BEST with USHCN (same as ClimDIV):

comment image

I’d expect recent trends to be extremely similar between both products.

Last edited 12 days ago by Weekly_rise
Reply to  Willis Eschenbach
April 1, 2021 7:40 pm

I think that the variability is so high that the [natural] noise overwhelms the puny trends which then are not significant and that it is not meaningful to discuss or compare them. Put another way: there is no real evidence for any trend.

Nelson
Reply to  Willis Eschenbach
April 2, 2021 5:10 am

Willis, I used to be a time series jock. I wish I had time to pick up R as I bet it has the tools of time series analysis built in. I did recently download R so its on my list of things to learn. Back in the day, if someone asked me to analyze the data you graphed, I would de-seasonalize the data and fit an ARIMA model to it. The autoregressive and moving average parts would pick up your persistence and the I would pick up any trend. There are much more complicated TS analysis tools, but its a start. I have always been amazed that climate scientists talk about time series data but seldom use the tools of TS analysis.

Weekly_rise
Reply to  Leif Svalgaard
April 1, 2021 7:46 pm

That is a valid point, the trend for ClimDIV since 2000 is not statistically significant.

comment image

Nicholas McGinley
Reply to  Willis Eschenbach
April 1, 2021 10:31 pm

What could possibly be all that difficult about measuring an entire planet or continent under field conditions to within hundreths of a degree?
If “climate science” did things the way real scientists did them, there would be error bars and confidence intervals and detailed explanations about exactly what was measured.

Last edited 12 days ago by Nicholas McGinley
Graemethecat
Reply to  Nicholas McGinley
April 2, 2021 5:14 am

Don’t get me started on the utter absurdity of averaging an intensive variable like temperature.

Thomas
Reply to  Graemethecat
April 2, 2021 9:43 am

Like the fact that the temperature of the atmosphere is not a measure of its heat content?

Weekly_rise
Reply to  Nicholas McGinley
April 2, 2021 8:36 am

There are error bars and uncertainties and exhaustive, peer reviewed documentation about what is measured and how from all of the orgs that produce instrumental temp records. See, e.g., here, for NASA.

Paul Penrose
Reply to  Weekly_rise
April 2, 2021 6:59 pm

And when you look into it in detail, you find that the “error bars” are just guesses. Plus, they show a complete lack of understanding the difference between accuracy and precision errors which must be handled differently.

Clyde Spencer
Reply to  Weekly_rise
April 1, 2021 7:45 pm

Wr
If something doesn’t seem right, look at your unexamined assumptions. I think in this case, the unexamined assumption is that the slopes of your trend lines are trustworthy and statistically significant.

As we have discussed previously, the slopes near zero are likely to not be statistically significant. What is your correlation coefficient? Typically, for a correlation coefficient much less than about 0.4, one has to accept the null hypothesis that there is no relationship between the dependent variable and the independent variable. That being the case, one shouldn’t be too surprised if different data sets have slopes that ‘wander’ around the nominal zero.

The take away here is that there is no strong, unequivocal warming trend that can be defended with a high correlation coefficient.

Weekly_rise
Reply to  Clyde Spencer
April 1, 2021 7:53 pm

I agree in part. As noted above, the trend is not statistically significant at the 95% level. I disagree that an absence of a statistically significant trend is evidence for the absence of a trend. In this case, the R-squared value is very small ( 0.007 for the ClimDIV trend since 2000), which indicates that the data are very noisy. Rather than saying there is no warming trend (or no cooling trend), we would more accurately say we don’t know what the trend, if there is one, might be.

Weekly_rise
Reply to  Willis Eschenbach
April 1, 2021 8:12 pm

Roasting is relative. We can say with confidence that we have warmed since, say, the mid 90s – the trend since 1995 is positive and statistically significant. We Just can’t say whether we are warmer in the US today than in 2000. Probably in a year or two we will be able to make that determination.

comment image

Jeff Alberts
Reply to  Weekly_rise
April 1, 2021 8:24 pm

We Just can’t say whether we are warmer in the US today than in 2000. Probably in a year or two we will be able to make that determination.”

We can say with fair certainty that we are not as warm as other epochs in this interglacial. We also can’t say that we’ve experienced a faster rise in temp than at any other time. The proxies just aren’t good enough to tell us much of anything with reasonable certainty.

And if we don’t know the past, the present has no context.

Thomas
Reply to  Jeff Alberts
April 2, 2021 9:49 am

Jeff, “We also can’t say that we’ve experienced a faster rise in temp than at any other time.” Proxies give temperature averaged over ten decades or more. If you average the temp anomaly of the last hundred years, we cannot say we have experienced a faster temp rise than any other time.

Weekly_rise
Reply to  Willis Eschenbach
April 2, 2021 7:26 am

Edit: I think a p-value of 0.08 is not statistically significant at the 95% level, but it’s close enough for government work. How far back do we need to go to identify a trend that is significant?

Last edited 11 days ago by Weekly_rise
Clyde Spencer
Reply to  Weekly_rise
April 2, 2021 9:10 am

The alpha threshold is used as a sharp cutoff, not a fuzzy boundary. You are the one who has been emphasizing statistical significance. Use it properly, or not at all.

The alpha threshold of 0.05 (‘Nickle’ threshold) is not a magic number and is probably popular because social sciences so rarely see correlation coefficients above 0.5! The ‘Mensa’ threshold (0.02) is probably more defensible, but would end up also rejecting a lot of social science papers. The ‘Intertel’ threshold (0.01) would be even better as it means only 1 out of 100 would be wrong, instead of the more common 1 out of 20, which you are using. Even your P-value of 0.015 would call for acceptance (non-significant) of the null hypothesis at the ‘Intertel’ threshold.

Have you noticed that the Excel package sometimes reports an adjusted R^2 value that is negative? That is physically meaningless as squaring any number should always result in a positive number. Furthermore, a negative variance for a positive trend is meaningless. As I have remarked before, there are some problems with Excel, and one should be cautious about accepting at face value what it reports.

Graemethecat
Reply to  Weekly_rise
April 2, 2021 5:15 am

You genuinely see a trend in noisy data like that?!

Clyde Spencer
Reply to  Weekly_rise
April 2, 2021 8:44 am

While it may be strictly statistically significant at the 0.05 threshold, meaning that we can’t reject the null hypothesis that there is no correlation, it is functionally useless because the independent variable (time) only predicts 0.0000130% of the variance in the dependent variable (temperature). Even casinos couldn’t make a profit on such a low margin!

pHil R
Reply to  Weekly_rise
April 2, 2021 9:36 am

Can someone say, Pinatubo? How much of that rise since 1995 can be attributed to warming after the global cooling effects of the Pinatubo eruption?

Davidf
Reply to  Willis Eschenbach
April 2, 2021 1:19 pm

Thats because people are constantly informed, by media, politicians, activists, teachers, academics, and vast numbers of busybodies, that temperature is soaring, and catastrophe is just around the corner. And most have neither the time or the interest to become independently informed from varied sources, and just assume that what they are being told from the institutions must be so.

Lrp
Reply to  Weekly_rise
April 2, 2021 2:18 am

Semantics; what would you know about something that doesn’t exist?

Clyde Spencer
Reply to  Weekly_rise
April 2, 2021 8:23 am

we would more accurately say we don’t know what the trend, if there is one, might be.

We can say what the trend isn’t! It isn’t strong. Indeed, it is so close to zero that is it statistically indistinguishable from zero.

Weekly_rise
Reply to  Clyde Spencer
April 2, 2021 9:36 am

I disagree with this. We cannot say what the trend isn’t, particularly for very noisy data. “Strong” is not quantifiable. Trends that are far from zero can be not statistically significant in very noisy data.

As I’ve said in the past, absence of a statistically significant trend is not equal to the presence of a statistically significant non-trend. This is perhaps the most common misuse of statistical significance one encounters.

Clyde Spencer
Reply to  Weekly_rise
April 2, 2021 3:03 pm

“Strong” is not quantifiable.

Then why do we bother with Correlation Coefficients? I would say that a “strong” correlation is greater than ≈0.71 because the independent variable is then explaining more than 50% of the variance in the dependent variable.

Trends that are far from zero can be not statistically significant in very noisy data.

How do you get a high Correlation Coefficient (0 << CC < |±1|) with very noisy data?

Part of the problem here is that time series are being treated as actual relationships between a dependent variable and an independent variable, with a presumed cause and effect. Time series tell one when something happened, not how or why. Time series actually provide one with a spurious correlation because time is not causing the change in the ‘dependent variable.’ Something else that also changes with time is the actual driver of the change in the dependent variable. Plotting the dependent variable against the driver causing the change, without regard to time, will provide a real correlation. If the relationship between the two variables is 1:1, then one will get a Correlation Coefficient of ±1; as one or both variables become noisy, the Correlation Coefficient will decline. With a time series, if the noise is growing over time, the slope may increase while the dependent variable may be constant.

However, this all started with Willis demonstrating that we are not “burning up,” and he is right that there is no evidence to support the idea that there is significant warming in the recent past.

Tim Gorman
Reply to  Clyde Spencer
April 2, 2021 3:35 pm

If the function being used is close to a sine wave, as the typical daily temperature profile is, then the slope of that temperature profile changes continuously over time. If one of the functions is A * sin(t) and the other is A * sin(t + d), i.e. an offset in time as in east coast vs west coast, then you should never find a correlation between the two functions, the value of their derivative will always be different at time “t”. What each function is actually most correlated to is the earth’s orbit with confounding factors like weather, altitude, and geography. Only by ignoring time can you develop any kind of correlation.

Tim Gorman
Reply to  Tim Gorman
April 3, 2021 5:55 am

ROFL!!! I get two downchecks without a single reply showing where my post is wrong! Sometimes the truth hurts folks!

Clyde Spencer
Reply to  Tim Gorman
April 3, 2021 9:16 am

Tim
I just gave you an up-vote, but strangely the number didn’t change. I have been trying to make sense of the voting and I conclude that most of the votes are a popularity contest. That is, if the readers who are inclined to vote like your position, then you get an up-vote, and vice versa.

In an ideal world, the votes would reflect what the reader(s) think of your argument and supporting facts. However, this isn’t an ideal world. Yes, sometimes the truth hurts!

TonyG
Reply to  Clyde Spencer
April 3, 2021 3:13 pm

That’s what I don’t like about the up/down vote system – it makes it too easy to simply downvote if you don’t like something and move on, instead of actually engaging with what you disagree with. IMO it doesn’t really have a place on a site like WUWT. Personally, I just don’t use it. I can get all of that I want on twitbook.

ATheoK
Reply to  Weekly_rise
April 1, 2021 7:51 pm

Graphed anomaly…

Calculating anomalies require history.
USCRN has no history before 2004.

Where did they get the history from to calculate anomalies??
Fudged graph.

Weekly_rise
Reply to  ATheoK
April 1, 2021 7:57 pm

This is explained in the USCRN FAQ. (Apparently, it ‘s the only frequently asked question.)

Independent George
Reply to  Weekly_rise
April 1, 2021 8:40 pm

Those trendlines don’t look correct to me. That graph shows cooling.

Right-Handed Shark
Reply to  Independent George
April 2, 2021 12:38 am

Me too

Weekly_rise
Reply to  Independent George
April 2, 2021 3:17 am

You’ll have to take that up with Microsoft, I’m afraid. Can you provide your trend estimates?

Thomas
Reply to  Weekly_rise
April 2, 2021 9:38 am

Short-term trends in data that has random variations are meaningless. There is a cooling trend in the last ten years of the record. Do you think that means we’re headed for an ice age? Neither do I.

A set of 1000 random numbers between 1 and 10 will almost always show a trend, either up or down, but the values are constrained between 1 and 10, so no real trend can exist. The odds that the first series of the numbers in the set will have about the same value as the last series of numbers are very slim.

February was the coldest month in the 15 year record, by nearly 1 °F.

DHR
Reply to  DMacKenzie
April 1, 2021 6:50 pm

Feb 2021 is there, just have to reset the date and year to Feb 2021. And yes, February drops about 4 degrees, a very large amount. March should show in about a week.

Weekly_rise
Reply to  Willis Eschenbach
April 1, 2021 7:34 pm

The NCDC page I downloaded the data from says “Contiguous Average US Temperature Anomaly,” so I’m not sure that that is the explanation here.

Weekly_rise
Reply to  Willis Eschenbach
April 1, 2021 8:05 pm

As far as I’m aware, the aim of the adjustments applied to the USHCN is essentially to bring it in line with the USCRN (USCRN is, after all, meant to be a reference network), so if there were a large difference between them that would be a bigger surprise.

Jeff Alberts
Reply to  Weekly_rise
April 1, 2021 8:28 pm

How exactly do you adjust 1200 to match 112?

Weekly_rise
Reply to  Jeff Alberts
April 1, 2021 8:46 pm

The CRN is meant to be a pristine network free of any siting issues, and is designed to provide a representative sample of US surface temperature. If the larger ClimDIV network is consistent with the CRN, it would give us confidence that the adjustments have made it look more like a “pristine” network. That is, we’ve adequately accounted for changes in network composition, siting issues, etc.

Trying to Play Nice
Reply to  Weekly_rise
April 2, 2021 4:39 am

It gives me confidence that the data is made up and is not an indication of what is happening in the real world.

Jim Gorman
Reply to  Weekly_rise
April 2, 2021 6:18 am

it would give us confidence that the adjustments have made it look more like a “pristine” network. 

Look closely at what you just said. Adjusting a data record to “make” it agree with another IS NOT SCIENCE. Corrections to data is only done when you have gone through a comprehensive analysis of each measuring device and scientifically determined systematic errors and their value. That is called a correction factor.

If you have two sets of data that don’t agree, YOU need to scientifically investigate the reasons for that and make conclusions based upon the differences.

Simply making measurements agree is not a valid practice in any field. If you were in a certified lab and get different values from two different machines would you allow me to decide which one is correct and just change the reading from the other to agree?

Weekly_rise
Reply to  Jim Gorman
April 2, 2021 7:52 am

The adjustments are not corrections – there are some corrections applied to individual station data if an actual error is identified, but overall the adjustments aren’t being done to correct errors. The temperature that a station read on a given day is the temperature that the station read on that day. Easy. The problem is that when you want to see how the area around a station is changing through time you now have to worry about whether the station was moved, or whether something about the way the station measures temperature has changed (did they start taking readings in the morning after taking them in the afternoon? Did they swap out the instruments in the station?). And then if you want to combine multiple station records to see how the wider region might have changed, you have even more to worry about. Were there broad, systematic shifts in the way measurements were taken (was there a trend in volunteers switching their measurement time?)? Are some station records longer than others? Are some stations oversampling what might be a very localized trend (i.e. in an urban area)?

This is why you need adjustments. Unless you just happen to know that your station network is completely pristine – stations have never been moved around, they’re all well sited, the instrumentation is maintained and consistent, all the records are the same length, stations drop in and out, etc. Then you don’t need to make adjustments. For the Climate Reference Network, we happily know that this is the case because it was designed that way. Unfortunately, nobody put together the Climate Reference Network until 2001. And we want to know what temperatures were doing alllll the way back to the earliest days of the instrumental record. And we know that that broader, longer station network has experienced changes in composition through time. So we know we have to make adjustments to account for all that messiness.

We can make adjustments and then compare the full network to our pristine reference network to see if they look the same. If they do, great! The adjustments are working well. If they diverge, then we need to go back and figure out what’s going wrong with the adjustments. We’d scientifically investigate the reasons for that and make conclusions based upon the differences.

Is this clearer for you?

Lrp
Reply to  Weekly_rise
April 2, 2021 10:09 am

You’re still fudging data according to subjective criteria. That should be clear to you.

Thomas
Reply to  Weekly_rise
April 2, 2021 10:11 am

Weekly. That’s a reasonable explanation. But you forgot to explain why the adjustments always cool the past.

Last edited 11 days ago by Thomas
Weekly_rise
Reply to  Thomas
April 2, 2021 12:54 pm

That’s a good question. Globally, the adjustments do not always cool the past. Here is a graph fro the NOAA showing the distribution of trend changes resulting from their adjustments for the globe:

comment image

As you can see, the adjustments are as likely to lower the trend as they are to raise it. We can see that when land temperatures are combined with ocean temperatures, the sum of the adjustments actually lowers the warming trend:

comment image

For the contiguous US, the situation is different, the adjustments do increase the warming trend:

comment image

I can’t speak to the specific reasons why this is the case (there are other commenters here who can, though), but the overarching reason is that the US has a “weird” station network. It is entirely manned by volunteers, for one, and there has been a systematic shift in the time of day that volunteers were taking readings (there was a change in guidance about when to take reading sometime in the mid-century, and it took decades for that guidance to percolate across all the station locations), leading to an overcounting of warm days in the past, and an undercounting of warm days in the present.

The US network is also very dense, but at the same time very transient. Stations come online or are taken out of the network all the time, stations are moved around with great frequency, etc. In fact (and I can’t locate the source for this), I understand that there has been a systematic shift in station density over time by latitude, so there are fewer stations at low latitudes now than there were in the past, which imparts a cooling bias.

The only other final point I can make is that people are often confused because it is always the past that is seemingly adjusted. People ask why there isn’t a need to adjust modern temps. The reason is simply that the present is the point of reference. The adjustments are essentially telling us, “what would the station network look like if all the stations were located exactly where they are today?” So you’re always adjusting the past relative to the present.

Hopefully this helps clarify a bit more.

Tim Gorman
Reply to  Weekly_rise
April 3, 2021 6:36 am

The adjustments are not corrections”

OF COURSE ADJUSTMENTS ARE CORRECTIONS! You are trying to rationalize to yourself why it is ok to make subjective changes to data!

If there are uncertainties associated with measurements then that uncertainty should be documented and handled with standard methods of propagating uncertainties. It truly is that simple! When you have to *decide* what the adjustment/correction should be then it *is* screwing with the data!

Unless you just happen to know that your station network is completely pristine – stations have never been moved around, they’re all well sited, the instrumentation is maintained and consistent, all the records are the same length, stations drop in and out, etc. Then you don’t need to make adjustments.”

Again, you do *NOT* make subjective adjustments/corrections to the data. You note the uncertainties and handle them in standard ways!

“For the Climate Reference Network”

Even the CRN stations have uncertainties associated with them!

Why do you never mention the uncertainties associated with altitude and geography of each station? Why is it assumed that each of these stations is sited on a perfectly flat plain extending for hundreds of miles around it and they are all at the same altitude?

“We can make adjustments and then compare the full network to our pristine reference network to see if they look the same.”

Again, you are assuming that the “pristine reference network” *is* pristine and has no in-built uncertainties. You are exhibiting the mindset of a mathematician or computer programmer and not a physical scientist.

Weekly_rise
Reply to  Tim Gorman
April 3, 2021 9:22 am

OF COURSE ADJUSTMENTS ARE CORRECTIONS! You are trying to rationalize to yourself why it is ok to make subjective changes to data!

That of course is untrue, and I carefully explained why in the comment you’re responding to. The changes are not subjective, they are both well justified and necessary. The adjustments are not corrections, since the adjustments are intended to make the network more cohesive, not fix errors in specific datasets.

A simple illustration: I want to take the average height of people in a room. I have a dataset consisting of a measurement from the floor to the top of each person’s head. I see that half of the people were standing on .5 meter tall boxes while being measured. If I subtract .5 meters from the measurement for those individuals, am I correcting an error in the data? Or making the dataset more cohesive so that I’m averaging together the same things? Or perhaps you’d argue that I shouldn’t even subtract the height of the boxes from those people, because I’m tampering with raw data, and my average should include the boxes?

Last edited 10 days ago by Weekly_rise
Tim Gorman
Reply to  Weekly_rise
April 3, 2021 10:41 am

The changes are not subjective, they are both well justified and necessary. “

Of course they are subjective! Even deciding to make an adjustment is subjective!

You are still trying to differentiate between “corrections” and “adjustments”. They both do the same exact thing to the data and they are both equally subjective, even using your subjective definition of each.

“Cohesive” is one of those squishy terms that means what you want it to mean and nothing more – it can’t be objectively defined.

tpg – “Well, this station was moved so we need to “adjust” its readings by adding 1C to it in order to make it more cohesive with the rest of the data”

If you don’t consider that to be a subjective adjustment/correction then you are only fooling yourself!

What you describe with the height of the people is a SYSTEMIC ERROR that is known and that can be corrected for BEFORE you actually take the measurements. That simply isn’t possible for a temperature station that has made measurements in the past. You cannot know what systemic errors are involved in the past measurements, the measurements are in the past and cannot be redone. In fact, the systemic errors may not even be what you think they might be – they are simply unknown. You can make a SUBJECTIVE guess at what should be done but that is nothing more than manipulating the actual data to make it look like you want it to look.

Again, physical science practice would say that you define what the uncertainty interval would be, even if that is a subjective “guess”, and then propagate that uncertainty using standard practice. You do *not* manipulate the data to make it into what you subjectively think it should be.

The only other alternative is to start a new data sequence totally unrelated to the older data sequences. THAT is what all the data manipulators *should* have done from the word go. Use the stations that have actual records over the time period being studied and let the rest go hang, their data is useless and trying to manipulate older data to be “cohesive” with newer data is subjective as all get out! It’s creating data out of thin air based on guesses and assumptions.

Let’s say you’ve lived in Kansas City for ten years and kept track of the daily miles/gallon used for your car. Now you move to Denver and all of a sudden the new data doesn’t match your old data. Do you “adjust’ all the old data to match the new data? How much do you adjust it? And *why* do you adjust it in the first place? It’s two entirely different data sets so why do you insist on jamming them together? Are you going to learn something from the adjusted data set?

Weekly_rise
Reply to  Tim Gorman
April 3, 2021 11:15 am

Of course they are subjective! Even deciding to make an adjustment is subjective!

comment image

What you describe with the height of the people is a SYSTEMIC ERROR that is known and that can be corrected for BEFORE you actually take the measurements.

It is not a systematic error, because in fact it’s not an error at all. The measuring device accurately measured the height of the people from the floor to the top of their head. The person originally taking the measurement of the person on a box had a use case for it. But now we want to use these measurements for something different, and our needs require everyone’s feet to be on the floor, so we make an adjustment.

Let’s say you’ve lived in Kansas City for ten years and kept track of the daily miles/gallon used for your car. Now you move to Denver and all of a sudden the new data doesn’t match your old data. Do you “adjust’ all the old data to match the new data? How much do you adjust it? And *why* do you adjust it in the first place? It’s two entirely different data sets so why do you insist on jamming them together? Are you going to learn something from the adjusted data set?

Of course there are cases where adjusting the data would make sense. What if you want to know if your fuel efficiency is degrading over time? That could tell you something is wearing out in the car’s engine. But if you combine the Kansas data with the Denver data there will be a spurious trend that is actually resulting from terrain differences.

Last edited 10 days ago by Weekly_rise
Tom Halla
April 1, 2021 6:13 pm

Wish her a happy birthday. Besides, given the UHI effects, the cooling is probably greater than shown.

Anon
April 1, 2021 6:21 pm

Some of the most educated people I know insist that we are living in times worse than and hotter than the 1930s Dust Bowl Era. Even as a child, I remember living through epic heat waves that triggered regional black outs, so when I hear that 2019 was the hottest yet, I sctatch my head.

I have no idea why I never fell under the CAGW spell and others have. It is an interesting social phenomenon.

Last edited 12 days ago by Anon
KT66
Reply to  Anon
April 1, 2021 6:57 pm

I had a conversation with a person once and she said the climate is certainly warmer than what she remembers growing up, so it can’t be normal.

I told her of course, because when you were growing up in the 50’s, 60’s, and 70’s, it was a period of cooling. Why should you think that was normal? Maybe it was abnormal?

Last edited 12 days ago by KT66
Jeff Alberts
Reply to  KT66
April 1, 2021 8:29 pm

Should have asked her, should nothing ever change? Or, would it be worse if it was colder?

Bill Toland
Reply to  KT66
April 2, 2021 1:27 am

That’s odd. I have the definite impression that the weather is colder now than when I was a boy. I assume that I am just feeling the cold more now that I am older.

Graemethecat
Reply to  KT66
April 2, 2021 5:39 am

I had a conversation with an old friend who, regrettably, cleaves to the AGW dogma. He asked me when the climate would go back to “normal”. I replied that there is no such thing as a “normal” climate, and never has been. Climate is inherently unstable and ever-changing.

rah
Reply to  KT66
April 2, 2021 9:46 am

So I assume she will retire to Alaska and avoid Florida?

Thomas
Reply to  KT66
April 2, 2021 9:59 am

Children tolerate both hot and cold better than old folks. I know this because I have been both.

lee riffee
Reply to  KT66
April 2, 2021 12:22 pm

My husband (who is a few years older than I am) keeps saying it was way colder when he was a child (in the 1960’s). But for some odd reason, as I have pointed out to him, he has failed to account for the fact that he grew up in the Appalachian mountains (West Virginia pan handle)! I asked him if he knew the elevation of his home town and he had no idea, not even a guess. He looked it up, and that town is about 3,200 feet above sea level. Which would account for memories of more snow in his childhood. He said it always snowed before or around Thanksgiving, and I told him that growing up in the 70’s it almost never snowed then (even in a cool period). But the town I grew up in is nearer the east coast and is only about 100 feet above sea level.
My mother in law is also like this – but she was born in the early 1930’s, having just missed (or was too young to recall) the blistering heat of that time. She had no altitude change but then again, the 1940’s was a cold period.

I wonder how many people fail to account for altitude (they grew up in the hills or mountains and later moved to lowlands) in remembering it being “much colder” years ago….

Clyde Spencer
Reply to  Anon
April 1, 2021 7:51 pm
Abolition Man
Reply to  Clyde Spencer
April 1, 2021 10:32 pm

Interesting point, Clyde.
It’s fascinating how often the cases occur among young women and girls! I think I’ll stick with women more like Princess Yuki of the Hidden Fortress!

yirgach
Reply to  Abolition Man
April 2, 2021 8:48 am
Clyde Spencer
Reply to  Abolition Man
April 2, 2021 9:18 am

Historically, that is where the term “hysteria” came from, as in hysterectomy. I’m surprised that the militant feminists haven’t yet banned the word “hysteria.”

Abolition Man
Reply to  Anon
April 1, 2021 10:23 pm

Anon,
You, like most here at WUWT, still retain the ability to think critically! The tools of logic and critical thinking have been almost entirely removed from the American education system over the last few generations; leaving a population that is both more pliable and less able work out solutions to problems and puzzles!
As students have been trained to accept most “authority” without question, the media have grown ever more ideological! We now have de facto state propaganda organs posing as media outlets, and our schools teach religious beliefs like Climastrology and Progressivism as basic curriculum! Those who question are to be shunned or attacked; auto da fe and guillotines are beginning to enter the conversation!

Joseph Zorzin
Reply to  Anon
April 2, 2021 2:22 am

“It is an interesting social phenomenon.”
More like a dangerous religion.

Patrick B
Reply to  Joseph Zorzin
April 2, 2021 7:29 am

And I think that’s part of the answer. It seems that a common element in the human psyche is an attraction to doomsday scenarios. My guess is that when a large part of the population practiced religion, this drive was satisfied by the religious beliefs. In the absence of traditional religious beliefs, people’s psyche have been drawn to other beliefs that offer a doomsday scenario. Because it’s impossible for the large majority to have sufficient knowledge to scientifically analyze the data, the support for CAGW is nothing more than a new religion.

Joseph Zorzin
Reply to  Patrick B
April 2, 2021 8:05 am

The new Satan is “carbon pollution”. And, now we must be “saved” by the purification of becoming “clean and green”. The new religion has priests and prophets and saints and holy scripture and martyrs and crusades. Somebody needs to write a book on this new religion- pulling no punches- slam these new fanatics with all the sarcasm and satire possible. I’d do it but I’m too lazy. There are a lot of good climate skeptic books out there- I have quite a few- but I don’t think any have focused on the theme that it’s a new religion.

Mr.
Reply to  Joseph Zorzin
April 2, 2021 8:27 am

So many people have eschewed the traditional religions now, they don’t even recognize a new one when it appears.

TonyG
Reply to  Mr.
April 2, 2021 10:50 am

An interesting observation there…

MPassey
Reply to  Patrick B
April 2, 2021 8:37 am

This doomsday psychology even has it’s own horseman of the apocalypse, the Tipping Point.

Anon
Reply to  Patrick B
April 2, 2021 9:53 am

Yeah, it strangely does have a lot of that:

1] Invisible Deities: Carbon Dioxide

2] Bible: IPPC Report

3] Priest Class: Unchallengeable scientists who interpret “the signs”

4] Saints: Greta Thunberg

5] Heretics: Skeptics

6] Indulgences: Carbon Offsets.

And if you go further back than the Judeo-Christian Era, many primitive religions were about pleading with the Gods to avert droughts and floods and to deliver the harvest.

So, it could be like a primordial “socket on a motherboard” that was “missing a chip” and the Climate Change Module was inserted to fill it?

Last edited 11 days ago by Anon
beng135
Reply to  Anon
April 2, 2021 8:57 am

I remember as a child having to sleep in the basement during the hot summers of the mid-1960s in west MD (but the winters then were indeed cold). Hasn’t been that hot & dry since. Record summer heat for Maryland like many states was in the 1930s.

Last edited 11 days ago by beng135
Michael E McHenry
April 1, 2021 6:25 pm

I just posted it to my FB family and friends list LOL

Latitude
Reply to  Michael E McHenry
April 1, 2021 6:31 pm

and FB gives you a time out in…3….2….1

Michael E McHenry
Reply to  Latitude
April 1, 2021 6:38 pm

I posted it as an image LOL

Jeff Alberts
Reply to  Michael E McHenry
April 1, 2021 8:30 pm

Lol

Why are we laughing?

beng135
Reply to  Jeff Alberts
April 2, 2021 10:13 am

Maybe he thinks posting just an image might escape attention of the censors.

Last edited 11 days ago by beng135
Doonman
April 1, 2021 6:28 pm

So the data says our work is done. All our policies in reducing emisssions have been effective and we’ve turned the corner on global warming. Clearly, the 20 year trend is going down.

That means we no longer need to take draconian measures to reduce our CO2 emissions. What we have eliminated so far is working just fine.

nickc
Reply to  Doonman
April 2, 2021 1:28 pm

When do we stop reducing emissions so that the declining temperature does not over shoot the “normal” temperature?

Craig James
April 1, 2021 6:29 pm

Shouldn’t the graph say 21st century?

Michael E McHenry
April 1, 2021 6:31 pm

Willis you have it marked 20th century when it should be 21st.

Dennis
April 1, 2021 6:41 pm

Burning, climate emergency, hottest day/week/month/year, and all the other climate hoaxer’s nonsense, crafted to divert people’s attention away from the Marxist Globalist plot to ruin the free market capitalist system that has brought prosperity to the developed world conveniently uses fires as a scare tactic to gain the attention of the gullible average member of the public.

In Australia we have droughts and severe droughts from time to time spanning back beyond colonisation by Great Britain in 1788, East Coast discovered by them 1770. The Australian Aborigines developed from watching nature at work their seasonal burning tradition, burning small patches at a time when the weather conditions are suitable including wind direction.

Fires need fuel, leaving fallen leaves and other material on the ground below trees and allowing other vegetation to mass and tangle at ground level invites fires, a lightning strike for example. Drought dry conditions add to the danger of a fire starting and spreading quickly.

But Australia also has floods from flooding rains at other times and unless land management is conducted the floods promote growth of fire hazard material. And of course CO2 also boosts growth, and without it there would be no life.

Left alone to become overgrown bushland will eventually burn out of control, wild fires, hot fires that destroy trees and wildlife, people and property.

In Australia Greens get themselves and comrades into positions of influence, local government councils for example where land clearing and controlled burning to remove hazards are issued, or in most cases not issued or very few are issued. After all bushfires are good for Green politics, unless we point out the destruction and loss of wildlife that might have been avoided with more care. But time moves on and the Greens turn to their other political games, in Australia claiming extinction of native animals because of climate change hoax, one example.

Richard (the cynical one)
April 1, 2021 6:51 pm

All the conflagrations are political. I guess when the energies are so polarized, the sparks fly.

rbabcock
April 1, 2021 7:06 pm

All the temperature datasets are different. Some are more corrupted than others. I like the Japanese JRA-55 Dataset which seems to be more honest in fixing the past. I know you are showing the US data, but looking at the last 30 years globally it’s moving within a really narrow range. Here is the global anomaly compiled by Ryan Maue.
http://climatlas.com/temperature/jra55/jra55_globe_t2m_1990_2021.png

Mike
Reply to  rbabcock
April 1, 2021 7:29 pm

I think that graph is…………………………………..garbage.

Jeff Alberts
Reply to  Mike
April 1, 2021 8:37 pm

Why?

philincalifornia
Reply to  Mike
April 2, 2021 3:22 am

The data doesn’t care what you think Mike.

Mike
Reply to  philincalifornia
April 2, 2021 7:03 pm

Ok I must be missing something. Why is the 2016 anomaly 2.4 C higher than 98?

philincalifornia
Reply to  Mike
April 2, 2021 8:06 pm

I’ve looked at the graph again to try to figure out what you’re talking about but I conclude that indeed you must be missing something and I can’t figure out what it is.

2016 anomaly is super high because of the strong El nino.

Mike
Reply to  philincalifornia
April 3, 2021 5:56 pm

No stronger than 97/98

philincalifornia
Reply to  Mike
April 4, 2021 7:51 am

If you’re still here, I honestly want to figure out what you’re on about. The graph does not say what you say it says. Are we looking at different graphs ??

rah
April 1, 2021 7:49 pm

“Cooling … not warming.
Cooling.
Now, before anyone starts yelling “But that’s just the US, not the globe!”, yes, I know that. I’m just saying that if you think the US temperature has gone up over the 21st century, you’re wrong. The US has cooled over that it’s”

Somethings about the alarmists are always consistent.

  1. “Global Warming” is always “Global” until it’s not! Then what ever data conflicts with their strongly held belief in some area of the world is “cherry picking”.
  2. As in all realms of leftist politics their beliefs rely very heavily on what they do not know, or claim to not know. For example last year the only Basin in Dr. Ryan Mau’s ACE index that had an above average ACE was the North Atlantic basin. The fact that the incidence and strength of tropical cyclones in every other basin was well below average and that the NW Pacific basin where the most powerful storms form had a historically low ACE, was ignored. Global Tropical Cyclone Activity | Ryan Maue (climatlas.com)
Chris Hanley
Reply to  rah
April 1, 2021 10:19 pm

“Global Warming” is always “Global” until it’s not!
CO2 is a ‘well mixed gas’, the atmospheric concentrations on Mauna Loa, at Cape Grim, American Samoa and South Pole average about the same over an annual cycle.
From season to season or year to year the troposphere T anomalies may vary over the globe due to local factors but over the forty or so years’ record, although there is a general warming trend, it is hard to attribute the anomaly pattern to ‘human influence [CO2 emissions] as the dominant cause’ (IPCC 2014).
comment image

rah
Reply to  Chris Hanley
April 2, 2021 1:22 am

From what I see and read there is no significant warming trend anywhere but at the poles. Even NASA admits that the poles have warmed faster than the temperate and tropic zones, but your map above does not show that.

philincalifornia
Reply to  rah
April 2, 2021 3:24 am

The Arctic yes, but not that pole which is the Antarctic.

Frank from NoVA
Reply to  rah
April 2, 2021 6:08 am

Water vapor swamps CO2 so if CO2 has any impact it will be most apparent in dry atmospheres, e.g. at the poles and over deserts. A couple of degrees warming that takes place mostly in the coldest air masses on earth obviously isn’t going to be a problem.

rah
Reply to  Willis Eschenbach
April 2, 2021 10:35 am

BTW I would like to thank you very much Mr. Eschenbach for all your excellent posts. You have done much to edify this truck driver on the subject of climate and climate change.

philincalifornia
Reply to  rah
April 2, 2021 8:13 pm

Do we have a second?

Yes ….. me

rah
Reply to  rah
April 2, 2021 5:03 am

“Even NASA admits that the poles have warmed faster than the temperate and tropic zones, but your map above does not show that.”

While I do not necessarily disagree with your statement that Antarctica is not warming, I was referring to what NASA says.

philincalifornia
Reply to  rah
April 2, 2021 8:27 am

Yes, I got that, I was correcting them not you, for people passing by here and taking a look.

Tim Gorman
Reply to  Chris Hanley
April 2, 2021 5:32 am

CO2 is a ‘well mixed gas’”

I don’t know where this canard started but it just isn’t true.

Go here: https://climate.nasa.gov/vital-signs/carbon-dioxide/

NOAA has a graphic on that page that shows global CO2 concentrations from 2002 to 2016. CO2 concentrations are *not* well-mixed. They vary greatly from point-to-point and time-to-time.

Tim Gorman
Reply to  Willis Eschenbach
April 2, 2021 10:19 am

The NOAA animated graph shows CO2 varying by more than 5% at the same latitude around the globe in the same year.

Tim Gorman
Reply to  Willis Eschenbach
April 3, 2021 6:40 am

Did you actually go look at the graph? The variation is *not* just seasonal, it is GEOGRAPHIC as well. The differences in concentration at the same latitude show up in every month, January to December! The difference between the east coast of the US and the west coast shows up in July as well as in December.So how can it be said that CO2 is well mixed?

Thomas
Reply to  Willis Eschenbach
April 2, 2021 10:21 am

Willis. Some of the highest absolute humidity levels ever measured have occurred in desert climates. Around the Persian Gulf and the Red Sea, for example.

Clyde Spencer
Reply to  Tim Gorman
April 2, 2021 3:15 pm

Tim,
Part of the problem is that while the term “well mixed” is used commonly, I haven’t seen any agency formally define what they mean by the term. However, as Willis points out, it is not unreasonable to use the term when the spatial variation is less than a couple percent.

Without someone putting a stake in the ground and declaring “This is my definition of ‘well-mixed,'” we are arguing about how many CO2 molecules can dance on the head of a pin.

Tim Gorman
Reply to  Clyde Spencer
April 3, 2021 6:48 am

If the differences were small I might agree. But the difference between the US east and west coasts is at least 7% and in Siberia is almost 8% when looking at NASA’s animated graph. Those are significant differences. And they persist no matter the month. If CO2 *is* a temperature control knob then differences that large should be encoded into climate models. The assumption that “CO2 is well mixed” is a crutch allowing these differences to be ignored in the models. A global parameterization can be substituted instead.

Clyde Spencer
Reply to  Tim Gorman
April 3, 2021 9:25 am

Well, they are, after all, doing a lot of homogenization with both the data and modeling! 🙂

But, it would still help if the groups processing the data and exercising the models provided their definition of “well mixed” and then we could point out violations of the term or argue about how appropriate the definition is. As things stand, we are stuck with a Humpty Dumpty definition of the term, meaning whatever they want when they want. They can’t be pinned down if they have the flexibility of terms that aren’t formally defined.

RickWill
April 1, 2021 7:55 pm

The single chart shows why those who perpetuate the climate scam deal almost exclusively in anomalies. To believe a 1 degree change in temperature is going to create hell on Earth, it is imperative that the 50 degree swing over a typical year must be eliminated.

If anyone claims the globe is warming, ask them what the current global temperature actually is and what was it 50 years ago. Best to pin them down to a monthly average because the monthly average across the globe swings from about 6C to 20C. So a 1C rise in that range makes it hard to sustain any notion of it being catastrophic.

Rory Forbes
Reply to  RickWill
April 1, 2021 10:05 pm

Most people can barely feel a whole degree F over the course of an hour, never mind a century. They certainly can’t compare the relative temperature difference between childhood and old age. This entire AGW fraud is so preposterous it’s a wonder they’ve managed to maintain it for so long.

PhilipA
Reply to  Rory Forbes
April 1, 2021 10:30 pm

I know it is fashionable to think that aboriginals burned small areas to mange grass growth, but my impression has always been that aboriginals burned areas to flush out animals to make it easier to hunt them, and they still do that to this day.

In some cases like extensive fires in the Kimberley the fires get “out of control” or more correctly do not die out naturally and burn very LARGE areas.

To me its like what boomerangs were for. They were to scare ducks or whatever into the air so that they could throw spears at them. There are some rock formations in Arnhemland where youths practiced throwing spears and they are embedded in rock fissures. Spears with woomeras were always the main weapon .

Tim Gorman
Reply to  RickWill
April 2, 2021 5:40 am

The other factor to consider is that the uncertainty associated with these independent, uncorrelated, measurements of different things ( local temperatures), even with only 112 stations, is over +/- 5C. And this is assuming that the average temps and mid-range temps used to determine these anomalies have an uncertainty of +/- 0.5C. The true uncertainty of these is undoubtedly larger than this.

That means the entire set of data being graphed falls within the uncertainty interval. Meaning the actual trend is not only unknown but is unknowable!

Jim Gorman
Reply to  RickWill
April 2, 2021 6:30 am

Anomalies by their very nature scale down the variance in the data. If the climate scientists using anomalies were credible, they would include the variance (or standard deviation) of the original absolute temperatures in their graphs and studies. It would look a lot less threatening that way.

It is also noteworthy that temperatures have not followed CO2 level increases. This alone makes the theory that CO2 is a “control knob” a vacuous conclusion.

Tim Gorman
Reply to  Willis Eschenbach
April 2, 2021 3:25 pm

The anomalies are calculated from daily mid-range temperatures which themselves hide the daily variance of the temperature profile.

Averages and mid-ranges hide variances. If I give you a mid-range temperature value of 56F can you tell me what the variance of the temperature profile was that produced that mid-range value? If I give you an anomaly of +0.2C can you tell me the variance in the temperature profile that produced that anomaly?

Abolition Man
April 1, 2021 10:52 pm

Willis,
Happy Birthday to your better half!
Most people are sadly ignorant of your emergent phenomena and Jim Steele’s concept of climate cycles as explanations for almost all of the climate variations we experience. Trying to push that particular rock up the hill must feel very Sisyphean sometimes!
I just wanted to thank you for your trove of easy to understand and interesting posts; I may have wasted most of my college education but I still enjoy the deep thoughts and conversations your writings provoke!

griff
April 2, 2021 12:17 am

Well looking at it as one huge lump conveniently obscures the trends…

Look instead at the boom/bust drought/heavy precipitation cycle in California, the temperature trends in Alaska, the western US winter temp trends…

I suspect much of the interior US is not affected such that residents would notice… absolutely not the case in the UK, where the change in UK climate is obvious (and irksome)

MrT
Reply to  griff
April 2, 2021 8:33 am

I’ve lived in the South-East of England for a few decades now. Pray tell what are these irksome and obvious climate changes I seem to have missed?

Tom Abbott
Reply to  griff
April 2, 2021 9:19 am

Griff, the U.S. has been in a temperature downtrend since the 1930’s.

rah
Reply to  griff
April 2, 2021 9:38 am

California is enjoying a long term period of greater than average precip so I don’t get your claim. The ski resorts in the west were supposed to be going out of business by now due to lack of snow. Instead they have had great ski seasons of late. Last year, Squaw Valley, one of the oldest ski resorts in the US, extended it’s ski season by a month for the first time in it’s history. And the thing is that much of the rocky and Teton ranges are forecast to get slammed with more snow in the near future.

Take you BS elsewhere Griff. Nobody here is buying it.

Andrew Wilkins
Reply to  griff
April 2, 2021 12:20 pm

Absolute twaddle. I live in the UK and have done for nearly 50 years. There has been no obvious change in the climate. Some years have been rather hot, some have been rather cold, and most have been somewhere in between. The perfectly benign UK climate is not in any way “irksome”.

Gerald Machnee
Reply to  griff
April 2, 2021 2:07 pm

Just a reminder, Griff, that stuff is called WEATHER, spelled W-E-A-T-H-E-R

Reply to  Gerald Machnee
April 2, 2021 5:26 pm

GM:
Exactly!
There is always some weather somewhere to talk (or complain) about.
But here in central Arizona we are having another beautiful day, perfectly consistent with what we would expect from ‘climate change’.

Peta of Newark
April 2, 2021 12:29 am

and my tuppence ha’penny is from the wee brood of English Wundergrounds I watch over.
(They’re all nearly 20 year old now. I have 4 on the west but can only find 3 with a decent length record in the east)

Most noticeable from each individually is that the ones on the Western side have a falling temp trend and the Eastern ones are rising,

Most especially crazy and especially especially if you’ve ever driven around there or, best of all, seen it from an aircraft at night, is the Wunderground Personal Weather Station at Salford, Greater Manchester.
About 2 miles due west of Manchester City Centre;
Greater Manchester has got to be THE most humongous built-up area you ever saw and getting ever more built-up all the times. It just goes on for ever. evah and evah
They might even have to start taking down a few Traffic Speed Cameras so as to squeeze in more houses. Yes, Greater Manchester is afflicted by Global Yellowing. Has been for quite some time – I suspect the Local (left-wing socialist) Council is bankrupt.

Because Salford has the steepest falling trend of the whole lot with:
.. a monthly trend line of y = -0.0053x + 11.173
So much for UHI eh?
The more you think about that, the worse it gets. So much for population increases, aerosols, clean air, pollution controls, diesel particle filters etc etc
The average temperature of Manchester City Centre has fallen 1.2 Celsius in the last 18 years. = the length of the record of that one station. No moves. No changes.
Same station. Same back-garden. Same guy running it, I guess

All England rolled into 2 Trend Lines
Eastern England trend line says: y = 0.0231x + 10.358
Western England trend line says y = -0.0269x + 10.772

x refers to ‘years’ = that’s the annual trend for the last 20 years
(as recorded and uploaded live by Wunderground at 5 minute intervals)

Last edited 11 days ago by Peta of Newark
Phil S
April 2, 2021 12:35 am

Thanks for all your excellent work. Many many thanks for showing actual Celsius and Fahrenheit temperatures rather than temperature anomalies which, not being a meteorologist I have never liked!

M Courtney
April 2, 2021 1:58 am

Most people have experienced warming over their lifetimes because:

1) Urbanisation.
2) Improved housing and insulation (they experience the jump to double-glazing from their childhood).
3) When people can they move to nicer places which are warmer. Who retires to the cold?

So their experience is counter to the facts for most of the place. People live in the few places where people live by definition. And they warm those places.

Alternatively the USA should thank Greta Thunberg. She has stopped AGW in the US just by being born.

Anthony
April 2, 2021 2:04 am

Ah, good old climate trends. You know also coming from Manchester UK we know a bit about the climate. We laugh at people who say, since we started the Industrial Revolution around 1760, that temps have risen one degree centigrade. What they don’t do is ask the right questions such as what was life like in 1750 to cause a revolution and what was the climate. Two good questions, well life was crap then because we had just had over 200 years of freezing weather and yes you can call them mini ice-ages if you want. So, people are measuring from the middle of a series of mini ice-ages, how sweet of them. In Manchester’s case we developed indoor work ( to get out of the cold, dummies) and most of you now call it industry. You can, if you want to, go back to those times with the rivers freezing in winter and snow being deeply on the ground in London, from November to the end of April. You could also go back to the time of the battle of Valley Forge in 1776, where the ground was frozen nearly three foot deep but I guess most people will pass on that one.

So if there is the odd climate warmest reading this, let’s try a similar experiment. Put your thermometer in the freezer for five minutes then bring it out, you will find, magically that the temperature in your room has increased without putting the heat on. Now that really is magic.

Joseph Zorzin
April 2, 2021 2:19 am

“the claim is made that the US is heating up fast”
Not warming enough for me- here in cold/damp Massachusetts.

beng135
Reply to  Joseph Zorzin
April 2, 2021 10:23 am

Same here in mid-Appalachians — nasty cold blast for April. Barely above freezing this afternoon w/strong winds & Arctic dew-points.

Last edited 11 days ago by beng135
Bruce Cobb
April 2, 2021 4:41 am

You need to understand the physics of CO2 energy. Sometimes it manifests itself as heat, and others as wild and crazy extreme weather. Additionally, CO2 heat will sometimes hide deep in the oceans, to come out at a later time when we least expect it, saying Boo! Then temperatures will really skyrocket. Science.

Anthony
Reply to  Bruce Cobb
April 2, 2021 5:48 am

Nah

beng135
Reply to  Anthony
April 2, 2021 10:28 am

His post is obvious sarcasm….

griff
Reply to  Bruce Cobb
April 2, 2021 9:13 am

Absolutely the case.

Joseph Zorzin
Reply to  griff
April 2, 2021 10:55 am

mental case- religion

Phil
Reply to  griff
April 2, 2021 11:20 am

Absolutely the case? That CO2 says boo? 🙂

Gerald Machnee
Reply to  griff
April 2, 2021 2:09 pm

it is April 2 already and Griff still missed it.

Sara
April 2, 2021 4:57 am

Willis E., you simply cannot do THAT!!!

You are destroying their fantasies that the USA/World/Planet is on fire and we’re all doomed!!!! This must be stopped!!! STOPPED, I tell you!

Unfortunately for the Doomsayers and Gloommongers, we’re having a rather average Spring here in the Midwest, but we could use some rain. I have been out photographing Canada geese, red mergansers, blue-winged teals, mallards and am looking for the buffleheads, and I can confirm it’s pretty normal, some surface water still standing in the marshy areas (normal) and Madame Spring is creeping in on little cat feet, and frankly, everything looks to be quite N-O-R-M-A-L. I know that word is anathema to the ecohippies, but that’s what’s going on.

Okay, I’ll go back to my corner now.

John Phillps
April 2, 2021 5:16 am

For anyone wishing to check against the source data, Berkeley Earth themselves publish trends for North America, the Contiguous US and the United States. Since 1990, their estimates in degrees C per century are: Contiguous USA: +2.42 (± 0.36), North America: +4.72 (± 0.39)
 
It seems those claiming a warming USA are backed up by the actual data providers. I find the Berkeley numbers rather more plausible for a variety of reasons. A 20 year trend is liable to get lost in natural variability.

http://berkeleyearth.lbl.gov/regions/contiguous-united-states

Doonman
Reply to  John Phillps
April 2, 2021 11:04 am

Which century are you referring to? Willis looked at the 21 century data. Clearly, temperature has been going down. Which makes perfect sense since that’s when emission controls and carbon footprint reductions were implemented. Since America is getting colder in the 21st century, our worries are over. No sense crying wolf when there is none around.

Doonman
Reply to  Willis Eschenbach
April 2, 2021 11:45 am

Oh dear. Now you’ve gone and spoiled the “CO2 is the earths temperature control knob” theory too. I’m guessing that 97% of all climate scientists who I stupidly agreed with will have to apologize soon.

John Phillips
Reply to  Doonman
April 2, 2021 2:44 pm

” Clearly, temperature has been going down. ”
 
Did you take a look at the link?

Willis took the Berkeley Earth data and applied a hand-rolled USA mask to the gridcells and came up with a 20 year cooling trend. As noted above the statistical significance of such a short trend is not exactly compelling. Also BE themselves publish trends for North America and various subregions and their number for the Contiguous USA since 1990 is +0.278 degrees C per decade.
 
So, Willis’s preferred data curator (and all the others, actually) seem to come to the opposite conclusion to him, and provide support to those who claim the USA is warming rapidly. 

Doonman
Reply to  John Phillips
April 2, 2021 6:08 pm

1990-2000 is in the 20th century and we are not discussing that. Re-read the opening sentence again. “For the last couple of decades…” Why you insist on bringing 20th century data up as pertinent to this discussion is beyond me.

Clearly, the temperature has been going down in the contiguous US for the last couple of decades because as shown on the chart made for this discussion, it has.

You need to show the chart is incorrect to make a valid argument in this discussion, but so far, you haven’t done that. If Willis’ “handrolled” US gridding method is incorrect, then show us where. Or if the data used is corrupt or not fit for purpose, then show us where.

John Phillips
Reply to  Doonman
April 3, 2021 2:34 am

“1990-2000 is in the 20th century and we are not discussing that. Re-read the opening sentence again. “For the last couple of decades…” Why you insist on bringing 20th century data up as pertinent to this discussion is beyond me.”

I have no idea where you got 1990-2000 from, the table is clearly labelled <i>since</i> 1990 and the data is plotted to 2020.
 
Check also the graph; Berkeley Earth data support the assertion that the USA is warming rapidly. 

Willis has not shared his code so it impossible to check, nor does he state the uncertainty range. I’d like to see the same method applied to a 30 year period, with proper error bars, but I won’t hold my breath.

The Moyhu site has a trend plotter, I plotted the BEST data for the last 20 years and it produced a trend of 0.22 C / decade but with a huge confidence interval of 0.16 to 0.27. Now that’s global data but I’m pretty sure the CI in the USA data will be of a similar magnitude, without posting the confidence intervals, Willis can say very little about the actual trend.

John Phillips
Reply to  Willis Eschenbach
April 3, 2021 11:39 am

Dial back on the snark, he says, before getting snarky.

I have not said the analysis is wrong, it may well be right. My main criticism is that the interval chosen is too short correctly to conclude anything very meaningful about the actual trend. You could prove me wrong by posting your confidence intervals, but I suspect you will not do so because the range will include the warming you claim to have disproven.

My second point is that Berkeley’s own analysis over a more meaningful thirty years supports those who claim the USA is warming, 0.27 C /decade for the contiguous USA and 0.47C / decade for North America. And yes, they show their confidence range.
 
Thanks for the link to the data, using ordinary least squares I got a slightly positive trend for the annual data and a level trend for the monthly, both with lousy significance, but there are many ways of skinning the trend cat. The 10 year average in the annual figure ending in Mar 2000 was 0.49C, the same average ending Feb 2020 was 0.96C. No warming there, no sir. 

Clyde Spencer
Reply to  John Phillps
April 2, 2021 3:51 pm

“Since 1990, their estimates in degrees C per century are: Contiguous USA: +2.42 (± 0.36), North America: +4.72 (± 0.39)”

From Taylor (1982):

Rule for Stating Answers

The last significant figure in any stated answer should usually be of the same order of magnitude (in the same decimal position) as the uncertainty.

For example, the answer 92.81 with an uncertainty of .3 should be rounded to: 92.8 ±.3

[Taylor, J. R., 1982, an introduction to error analysis; University Science Books, Mill Valley, CA, p.17]

Therefore, BEST should be showing their summary results as: Contiguous USA: 2.4 ± 0.4, North America: 4.7 ± 0.4

It raises questions about the veracity of their answers when they make up their own rules.

An exception to intermediate answers, which will be used for subsequent calculations, is to maintain an additional guard digit, usually shown with parentheses. However, showing additional digits in the answer that go beyond where the uncertainty digit starts is unwarranted and implies more precision than actually is present.

John Phillips
Reply to  Clyde Spencer
April 3, 2021 6:57 am

It raises questions about the veracity of their answers when they make up their own rules.”

You’re just throwing sand in the air. Willis did not post any uncertainties at all, what does that say about the ‘veracity of his answers’? Ah, he gets a free pass.

Tim Gorman
Reply to  John Phillips
April 3, 2021 7:27 am

Willis certainly doesn’t get a free pass but if the data he uses has no uncertainty quoted then it makes it difficult for him to quote an uncertainty value. And that’s the problem I have with so much of the temperature data that is available. It is a conglomeration of a data set formulated by jamming together temperatures and anomalies with no regard for uncertainty or even such gross differences as those between coastal and inland measuring stations let alone those at sea level versus being a mile high! Even such a small geographical difference as being on the north side of the Kansas River valley versus the south side, a difference less than 50 miles can make a vast difference in temperatures depending on prevailing winds. Yet stations on both sides get thrown together when formulating some kind of an “average”. I still don’t know what such an average is supposed to tell you!

Clyde Spencer
Reply to  John Phillips
April 3, 2021 10:01 am

BEST is processing and adjusting historical temperature data-sets. They provide their estimates of uncertainty. Willis, on the other hand, is relying on data published by various agencies, which rarely provide any uncertainty measurements for him to post.

Major problems with climatology are that definitions are often lacking, as with “well mixed,” or perverted as with “more acidic,” and laws of statistics such as the “Law of Large Numbers” are misapplied, and basic principles of metrology such as the appropriate display of uncertainties and handling of significant figures, lead to mischief in the conclusions.

Sometimes approximate numbers, without uncertainties, can be used to answer general questions such as “Is the world burning up?” The assumption is that there is a change in temperatures that are of a gross nature and easily discerned. Other times, for questions such as what is the sign and probable magnitude of the trend, require more precise measurements and considering the impact the quantified uncertainty has in the answers to those questions.

In a rigorous treatment, if the uncertainty has not been, or cannot be estimated, then it should be acknowledged with a statement such as “While the actual uncertainty is unknown, for the purposes of this illustration, it is assumed to be negligible and not having an impact on the conclusion.” WUWT contributions are not held to high academic editorial standards. However, if you think that not acknowledging the uncertainty has a bearing on the argument presented, you are free to bring it to the attention of the author and readers, and state why you think it is a serious omission.

John Phillips
Reply to  Willis Eschenbach
April 3, 2021 3:46 pm

In fact, it is NOT statistically significant.

 
Well, I suppose I should be grateful for the concession that the basis of this post, statistically speaking, is worthless. One is left wondering why you bothered?
 
To summarise: a straw man argument, in the sense that no evidence is provided for the claim that anybody is looking at the 20 year trend and claiming a rapidly warming USA. Plus some worthless statistics to assert that the 20 year trend is negative, when the data is actually too noisy to prove any such thing.
 
WUWT. Situation normal.  

Oatley
April 2, 2021 5:20 am

Willis, you never fail to impress. From your innate curiosity, mathematical skill, broad experience and brutal honesty you hold your work up for all to see and to poke at for weaknesses. That makes you TRUSTED in my world.

Tom in Florida
April 2, 2021 5:29 am

“so we’re going for a lovely bike ride for a birthday treat.”
My first quick read thought it said “birthday suit”.
But then, my mind wanders some times.

April 2, 2021 5:56 am

I host the website co2questions.com, the first question of which links to a short video I think must also totally nullify the idea that the globe is warming.

Roger Taguchi
April 2, 2021 8:17 am

Sarah Palin will be angry at you Mr. Eschenbach. Your cherry picked definition of “America” by leaving out Alaska will infuriate her.
.
https://19january2017snapshot.epa.gov/sites/production/files/2016-09/documents/climate-change-ak.pdf
.
But then, you are correct that “America” is not burning if you ignore the part of America that is burning.

Last edited 11 days ago by Roger Taguchi
Roger Taguchi
Reply to  Roger Taguchi
April 2, 2021 9:07 am

Oh yeah, if you want to talk about “burning”…… comment image

Anthony
Reply to  Roger Taguchi
April 2, 2021 9:22 am

I’ve shoes older than 1965, go back 300 years and then report

beng135
Reply to  Roger Taguchi
April 2, 2021 10:33 am

Result of forest mis-management that’s been documented for many decades & continues at the present.

OK S.
Reply to  Roger Taguchi
April 2, 2021 10:36 am

Your chart reminded me of the news a few days ago blaming Alaskan earthquakes on global warming. But I remember 1964, the year before your chart starts: Great Alaskan Earthquake

GreatAlaskanEarthquakeGovernmentHillSchool.jpg
gbaikie
Reply to  Roger Taguchi
April 2, 2021 5:55 pm

Alaska average temperature is about minus 3 C and the lower 48 average temperature is about 11 C.
If Alaska average temperature was 14 degree warmer, would Alaska have less problems with forest fires?

Andrew Wilkins
Reply to  Roger Taguchi
April 2, 2021 12:31 pm

“burning”
ROFLMAO

BobM
April 2, 2021 8:26 am

March UAH temp is out… -0.01 anomaly.

Tom Abbott
Reply to  BobM
April 2, 2021 9:23 am

Down below the line, huh?

Doonman
Reply to  BobM
April 2, 2021 11:09 am

How strange. CO2 concentration keeps going up. It truly is a magical gas as it is now obvious that it heats and cools at the same time.

Tom Abbott
Reply to  Doonman
April 2, 2021 7:38 pm

CO2 up, temperatures down.

That’s not the way things are supposed to go, according to the alarmists. They claim that increases in CO2 equal increases in temperatues, but that appears not to be the case.

Clyde Spencer
Reply to  Doonman
April 3, 2021 9:32 am

It depends on the phase of the moon.

April 2, 2021 9:03 pm

Any article by Willis is well worth the read as we all need a regular dose of logical common sense. Thanks Willis.

The temperature trend since 2000 is of interest but, to me, is far less interesting than the trend over the past century or more when there have been many long records of thermometer data.

Furthermore, I prefer to see plots of max and min temperatures at specific stations rather than an average country-wide figure.

I have recently plotted the long-term Annual, January & July mean Max & Min temperatures at 15 stations in the contiguous USA. There are definite long-term trends and, although these vary in magnitude, they are all surprisingly linear. What do you think?

https://briangunterblog.wordpress.com/2021/01/27/usa-east-temperatures/

https://briangunterblog.wordpress.com/2021/01/27/usa-west-temperatures/

Tim Gorman
Reply to  Brian Gunter
April 3, 2021 5:54 am

I think the graphs show why it is so misleading to cram temperature data from independent sites together to try and come up with some regional or global “average”. Altitude and geography play such an important role in the temperature profile yet get totally ignored by cramming temperatures from independent locations into one data set.

Clyde Spencer
Reply to  Tim Gorman
April 3, 2021 10:11 am

Tim,
On the other hand, if the question is “What is the average global land temperatures?”, then the only appropriate approach is to average all surface land temperatures without regard to confounding factors. That would be in contrast to a more theoretical question, “What would the average temperature of Earth be if all land were at the average topographic height of land?” The latter would require applying the universal gas law or lapse rate to come up with a theoretical value. The approach should be dictated by the intended use of the calculated number.

%d bloggers like this: