Today, at the big 100 year anniversary shindig of the American Meteorological Society (AMS) there was a press release session that featured NOAA and NASA GISS talking about how their climate data says that the world in 2019 was the second warmest ever.
Here is their slideshow presentation, released today: https://www.ncdc.noaa.gov/sotc/briefings/20200115.pdf
In my opinion, the NOAA/NASA press release (and slideshow) is inconsistently presented. For example, they can’t even agree on a common base period for comparisons. Some graphs use 1951-1980 while others compare to 1981-2010 averages to create anomaly plots. NOAA and NASA owe it to the public to present climate data with a consistent climate period for comparison, otherwise it’s just sloppy science. NASA GISS has consistently resisted updating the 1951-1980 NASA GISS baseline period to the one NOAA and other datasets use, which is 1981-2010. GISS stubbornly refuses to change even though they have been repeatedly excoriated for keeping it.
That 1951-1980 period just so happens to be the coolest period in the 20th century, so by using that as a baseline, the peak amount of warming anomaly is magnified in NASA GISS plots. Most laymen will never spot this. A simple comparison of the two maps show the difference in the peak values:
The difference between the two analyses is NOAA @ 0.95°C/1.71 ° F and NASA GISS at 0.98 ° C/1.8 ° F
There isn’t much separation between some years at all:
One wonder if NOAA/NASA can accurately measure the Earth’s temperature with that sort of precision down to 1/100th of a degree. My experience suggests no.
Now lest you think I’m splitting hairs, bear in mind that during the last decade changeover, NASA declared this about 2009 and the 2000-2009 decade:
NASA Research Finds Last Decade was Warmest on Record, 2009 One of Warmest Years
…
The past year was a small fraction of a degree cooler than 2005, the warmest on record, putting 2009 in a virtual tie with a cluster of other years –1998, 2002, 2003, 2006, and 2007 — for the second warmest on record.
…
In the past three decades, the GISS surface temperature record shows an upward trend of about 0.36 degrees F (0.2 degrees C) per decade. In total, average global temperatures have increased by about 1.5 degrees F (0.8 degrees C) since 1880.
“That’s the important number to keep in mind,” said GISS climatologist Gavin Schmidt. “The difference between the second and sixth warmest years is trivial because the known uncertainty in the temperature measurement is larger than some of the differences between the warmest years.”
Source: https://www.nasa.gov/home/hqnews/2010/jan/HQ_10-017_Warmest_temps.html
So, small fractions of a degree matter when it comes to making climate claims.
But this year, Dr. Schmidt says:
“We crossed over into more than 2 degrees Fahrenheit warming territory in 2015 and we are unlikely to go back. This shows that what’s happening is persistent, not a fluke due to some weather phenomenon: we know that the long-term trends are being driven by the increasing levels of greenhouse gases in the atmosphere,” Schmidt said.
Source: https://www.nasa.gov/press-release/nasa-noaa-analyses-reveal-2019-second-warmest-year-on-record
So much for the concern about “uncertainty” and “trivial” differences in temperature. Interestingly, most people don’t know this but NASA GISS is not a truly independent analysis. They use GHCN data prepared by NOAA NCEI, the same data gets the NASS GISS “special sauce” and magicically come out a little bit warmer, allowing them to make the kinds of claims they make of “warmest ever” year after year.
While we are on that subject of datasets and small temperature differences, here is a bit of inconvenient data that never gets mentioned.
NOAA’s U.S. Climate Reference Network (USCRN) has the best quality climate data on the planet, yet it never gets mentioned in the NOAA/NASA press releases. Commissioned in 2005, it has the most accurate, unbiased, and un-adjusted data of any climate dataset.
The USCRN has no biases, and no need for adjustments, and in my opinion represents a ground truth for climate change.
In this graph of the contiguous United States updated for 2019 comes out about 0.75°F cooler than the start of the dataset in 2005:
The data show above was not plotted by me, nor adjusted in any way. It is entirely NOAA’s, and the fact that 2019 in the USA48 is cooler than 2005 is indisputable. The two large peaks are related to natural El Nino events which warmed not just the USA, but the world.
While the U.S. isn’t the world, and the dataset is shorter than the requisite 30 year period for climate data, the lack of warming in the contiguous United States since 2005 shown in the graph above suggests that the data NOAA and NASA use from the antiquated Global Historical Climate Network (GHCN) reflects warmer biases due to urbanization and adjustments to the data. I’ve demonstrated this with my own work here: https://wattsupwiththat.com/2015/12/17/press-release-agu15-the-quality-of-temperature-station-siting-matters-for-temperature-trends/
Imagine if the entire world had a high quality state of the art temperature monitoring network like the USA does. Arguments over data quality, adjustments, UHI, time of observation, and many other niggles which affect and bias the data would disappear.
But here’s also something interesting. All of the temperature plots used to represent climate change are highly magnified. This is so variations of one degree or less are highly visible. Unfortunately, these huge variation often scare the public since they perceive them as “massive” temperature increases.
Fortunately, the NOAA online plotter allows adjustment of the vertical axis, and when the vertical axis of the climate data is adjusted to fit the scale of human temperature experience, they look less alarming.
Here are some before and after examples, with the “after” plots set to the scale of human experiences with temperature, which typically range from around 0°F to 100°F. The zero mark on these anomaly plots assumes an average global temperature of about 58.4°F.
US Climate Reference Network:
Contiguous USA Climate Divisions (GHCN/COOP) Network (old style weather stations):
NASA GISS for the world (Land and ocean temperature):
“Climate change” certainly looks a lot less scary when the temperature change is presented in the scale of human experience.
UPDATE:
For those who care about replication, here is the USCRN data for 2005-2019, independently compiled into a CONUS presentation, but using a different baseline. This was done by Dr. Mark Albright from the University of Washington and sent to me via email.
Dr. Albright writes:
I use the base period 2002* – 2018 for determining “normal” or climatological values at each CRN site. I make no attempt to adjust these to 1981 – 2010. I believe NCEI does attempt this sort of adjustment to the CRN data. I keep adjustments to a minimum which pretty much means no adjustments.
* The beginning year actually varies from 2002 to 2008 depending on when the site was installed.
By the way, on 3 January 2020 I also plotted California for 2009 – 2019. You can see the big fall-off in temperature since 2014. https://atmos.uw.edu/marka/crn/ca/ca.crn.2009-2019.pngAnd the Pacific Northwest shows a large decline since the peak temperature in 2015: https://atmos.uw.edu/marka/crn/pnw/pnw.crn.2008-2019.png
What is most interesting to note is that due to the different baseline period and method Albright uses, the peaks on his graph are lower than the peaks on the NOAA USCRN graph.
For example: 2012
NOAA peak is 2.53°F Albright’s peak is about 1.7°F
Choice of baselines and methods use to create them matter, especially when it comes to any single year being the “hottest ever”. When the group presenting anomaly data gets to choose their own baseline period, and how it was created, they can essentially make the peak amplitude higher or lower based on those choices.
They don’t have good data for Africa, but just colour a giant red circle around a few sparse, god knows what happend to them, weather stations. Not sure if this was a good sentence, but it sums up what I think about weather stations and using it for climate.
And probably the weather stations in Africa are in urban environments which are subject to UHI.
Most of them I guess will be in Airports.With the massive increase in the number of flights i think you’d be hard pressed to find worse sites for detecting changes in climate.
Locations without stations became ‘fill-ins”. Including parts of Africa and the southern hemisphere.
Make that most of Africa, as well as most of Asia and all of the oceans.
“Imagine if the entire world had a high quality state of the art temperature monitoring network like the USA does.”
You don’t need to imagine it as there is a global CRN since it is an idea of the WMO. I remember a report saying that the US was actually behind schedule implementing it.
Is this the same high quality monitoring that the surface stations project showed was anything but by any chance?
Just showing how few places record temperature globally on the pretty world maps might well shock a lot of people to the extent that they may wonder about ‘global temperature’.
The entirety of Africa had 14 thermometer stations continuously measuring temps from 1880-1950, with 11 of those being in Mediterranean countries (1 – Morocco, 2 – Egypt, 8 – Algeria). The vast interior had nothing.
https://data.giss.nasa.gov/gistemp/station_data_v3/
How dare you suggest such a thought!
The world is much much more than just the United States.
Simple Simon says.
So does Wonderful Werme.
LOL
And the actual historical temperature data for that much much bigger area are much much less, which means those data are much much less reliable for anything other than local determinations. For instance, there were fewer than 50 thermometers in the entire southern hemisphere continuously measuring temperature during the first half of the temperature record (1880-1950), and most of those were concentrated in SE Australia.
There was only one thermometer in the entirety of sub-equatorial Africa continuously recording temperature from 1880-1950. That was located in Cape Town on the coast, which is hardly representative of the rest of sub-equatorial Africa. The unadjusted data from that location shows no discernible trend.
https://data.giss.nasa.gov/cgi-bin/gistemp/stdata_show_v3.cgi?id=141688160000&dt=1&ds=7
The entirety of sub-equatorial South America had 6 thermometers with 4 of those being on the coasts. Brazil only had one on the coast.
Antarctica had zero thermometers for that period.
The temperature data for all locations with no real data is simply made up.
https://data.giss.nasa.gov/gistemp/station_data_v3/
Sub-equatorial doesn’t appear to be the right term. What’s the correct word for everything south of the equator?
It’s two words: southern hemisphere
Southern hemisphere?
I was looking for an adjective that describes that part of a continent that lies in the southern hemisphere.
So where are the temperature sites where all the dark red is ?
Alaska, Siberia, and Canada are warming up fast. They may even become livable. This must be stopped!
Minnesotans for global warming!
I told a warmist I’d be delighted for him to be right as it would be a nice warm retirement for me.
As long as the continent spanning glaciers don’t return to the northern hemisphere, I’m good with minor temperatures fluctuations we have. Long Live the Holocene Interglacial!
1 – America is one of the few places for which there is good data.
2 – We’ve been warming out of the Little Ice Age for much more than a century. Of course it’s getting warmer. Global warming is all about mischaracterising natural phenomena.
Warming since the LIA – Yes – but as the LIA is recognised as being 2C and more COLDER than the long term ‘average’ then, at just ~1.5 C above 1850, Earth hasn’t fully recovered yet.
It begs the question why the IPCC want to lock earth into the temperature range of the LIA.
Politics, politics and ever more socialist-marxist politics coupled with $ multi-billions is the answer.
“The most terrifying words are “I am from the government. I am here to help you.”
Ronald Reagan
The world is much much more than just the United States
And the world doesn’t have the coverage that the good ol’ USA has.
Simon, that was fully explained in the article , but you ignored it to make your stupid point.
You expect a troll to actually read an article?
“The world is much much more than just the United States …”.
Which is one of the reasons the surface global temperature record from 1880 (or 1850) is not fit for purpose.
“The world is much much more than just the United States”
The United States temperature record is a good proxy for the temperature profile of the whole world since all regional temperature charts show the same temperature profile as the U.S. temperature profile, i.e., that the 1930’s were just as warm as today.
This fact destroys the Human-Caused Climate Change narrative. This fact destroys the bogus, bastardized, fraudulent, “hotter and hotter” Hockey Stick global temperature profile. No unmodified regional temperature chart shows a “hotter and hotter” “Hockey Stick” profile. It’s only after NOAA and NASA Climate put the unmodified numbers in their computers that they get a Hockey Stick. A computer-generated fantasy and science fiction world. A very expensive computer-generated fantasy and science fiction world.
Until the satellite era, most of the rest of the world wasn’t well enough monitored to know it’s temperature down to a tenth of a degree.
But if it is GLOBAL warming then isn’t the US part of the globe?
If it is regional warming then why isn’t it called such?
You should have used degrees Kelvin and then the temperature increase would have looked much smaller.
Fahrenheit isn’t an SI unit.
use kelvin but scale it for min/max i.e.
0 Kelvin to 1.417×10^32 kelvin (plank temperature).
This would prove that the earths temperature has not changed since the big bang! Just what you need!
When dealing with small fractions of a degree, claims of warmest years in the range of those small fractions are meaningless hype.
It took me five tenths of a second longer to type the previous sentence than it would have taken me last year. Oh God, my ability to type is degenerating!
My heart rate is one beat per minute faster, after eating a meal, than on an empty stomach. Oh God, eating is going to shoot my blood pressure too high!
I usually drink about three quarts of water per day, but today I drank an ounce less. Oh God, I’m going to get kidney stones!
I’m going to make my y axis three inches per gradation, while making my x axis a fourth of an inch per gradation, because drawing scary pictures is fun.
Very few people, except engineers and some scientists, and Anthony Watts of course, understand the phenomenon known as noise. Amplify noise, and you can show all kinds of meaningless “trends” if you select the right time period.
If you think very few people don’t understand what noise is, spare a thought for those who don’t even know what a trend is.
“trends”
Most trolls, such as yourself, believe that a trend is when you pick two points and draw a straight line between them.
You still alive Loydo? I would have thought the climate or the white man would have got you by now.
Like you know anything about science or math, Loydo. You still have never demonstrated that you even know the difference between accuracy and precision. Nor whether you can reduce either or both using mathematical procedures (like averaging multiple measurements). Until you can answer those basic questions correctly, nobody should take your comments on the science seriously.
Speaking of noise…
Noise?
All the surface “measurements” are “noise”.
No one with sense would believe the infilled, repeatedly adjusted, homogenized surface data are useful scientific data.
Near global temperature data began in 1979, using satellites measuring in the troposphere, where the relevant warming from greenhouse gases would occur.
Before 1979, we have non-global garbage temperature “data”.
Before 1920, we have minimal temperature data, except from the US, Europe and Australia.
Even worse:
All real time measurements were made DURING a warming trend that started in the 1690s — so record highs SHOULD BE EXPECTED until that warming trend ends.
And we’re only missing 4.5 billion years of real time global average temperature compilations !
Real news would be a lack of record highs for a decade or more — that could mean the 300+ year warming trend is ending, and a cooling trend is starting.
Now for the most significant question of all:
What is a “normal” global average temperature ?
Since no one knows the answer, I wonder why leftists have chosen 1750 as “normal”, when it was during the Little Ice Age — people at the time thought their climate WAS FREQUENTLY TOO COLD !
The current average temperature of our climate is the best it has been for humans in at least 800 to 1,000 years, since before the Little Ice Age centuries.
No one with sense would complain about the current climate, or about the warming since the 1690s, unless they had ulterior motives.
Our planet supports the most life when it is warmer than today, and has much more CO2 in the air than today.
A warm, high CO2 level, planet is best for all living things, plant and animal … so why do leftists insist those conditions are bad news for humans?
That doesn’t make sense.
The past 300+ years of global warming was 100% good news, but any more global warming will be 100% bad news ?
I’ll believe that when people in cold Michigan USA stop taking winter vacations in hot Florida — they get +40 to +50 degrees F. warmer after a two to three hour flight — and Michiganders are happy with the heat when they get off the plane !
Of course if Michigan gets a few tenths of a degree warmer in the next ten years, from global warming, that would be a catastrophe — we’d all have to move from Michigan to Alaska, to get away from the heat !
NOAA’s made up numbers are a pack of lies.
The super El Niño years of 1998 and 2019 were both warmer than this year’s ordinary El Niño. Earth has been in a cooling trend for almost four years, and from 1998 to 2016, global temperature, to the extent it is measurable, was practically flat. This despite supposedly steadily rising CO2.
If the people who pay GISS to lie knew how mendacious are NOAA and NASA, heads would roll.
Both base periods are as good as useless in the context of the entire Holocene. The data represents about one per cent of the interstadial, and it’s right at the end of the coolest quartile. So, the data’s statistical significance is highly questionable, and cherry-picking baselines to create temperature anomaly graphs as well as using exaggerated Y-axis is just too tricky by far.
Bureaucracies are blobs, they never go away but only grow. NASA is one big giant blob. It’s the blob’s blob if you’re studying blobs.
I’m a Brit, not an American, but as a boy I was in awe of America & NASA & followed the Apollo missions avidly with great enthusiasm. I fear that if NASA were tasked with returning a man (or woman, I don’t want the Feminists on my back) to the moon, they would have years of research to do to find out exactly just how they managed it, otherwise lives would be at far greater danger than would ordinarily be the case for such special mission!
“NOAA’s made up numbers are a pack of lies.
The super El Niño years of 1998 and 2019 were both warmer than this year’s ordinary El Niño.”
What numbers are you working off?
Loydo. 1998 has been adjusted down because it spoiled the warming trend. Fact.
Yes, the “Ministry of Truth” has been hard at work erasing “inconvenient” history…
Tanya Harding fair-play.
Re baselines for T anomaly calculation.
We in Australia often read about the many adjustments made to T data. These must alter the baseline values as well. Any adjustment that changes baseline values, but not the rest, will need a new anomaly map. And vice versa.
So, my question is, why are there not dozens or hndreds of different anomaly maps, one for each time an adjustment to the baseline values or part there of, is different to the adjustment to any of the rest of the T values? Geoff S
“Any adjustment that changes baseline values, but not the rest, will need a new anomaly map”
It doesn’t. Changing baseline just adds a number, the same one, to all the readings. It doesn’t change relative values. The most that would be needed is a relabelling of the color scale.
However, adjustments make very little difference to the global average. They make even less to the baseline, which is the average of at least 30 years.
Nick
“It doesn’t. Changing baseline just adds a number, the same one, to all the readings.”
I think you misunderstood Geoff Sherrington.
He is imho talking about single values within the reference period aka baseline. These modifications of course modify the entire data set.
*
Never and never would anybody keep all these anomaly maps alive. This makes no sense.
After all, the same is valid even if you would publish absolute values only. When you add a microwave sounding instrument, or surface stations, the averaged absolute values change!
Rgds,
J.-P. D.
Bindidon’s correction is correct Nick,
Now, do you have an accurate response to my careful question?
Do we in reality generate multitudes of anomaly maps because values in the reference period can be adjusted differently with respect to the rest of the values? Cheers Geoff S
Geoff Sherrington
“Do we in reality generate multitudes of anomaly maps because values in the reference period can be adjusted differently with respect to the rest of the values?”
Sorry: this question makes few sense to me.
1. Again: the problem has nothing to do with anomalies: you have to change any average of absolute values whenever you add, change or remove one or more single values.
2. Either a modified absolute values interferes with the reference period, or it doesn’t. If it does, then of course the entire time series has to be reconstructed. This is the reason why all anomaly-based time series change so often.
The UAH time series was modified numerous times, e.g. to perform recomputations of absolute values due to satellite orbit decays, when values within 1981-2010 (or earlier 1979-1998) were affected.
What exactly is your point?
Geoff,
“Bindidon’s correction is correct Nick,”
No, my statement was correct. Changes within the baseline modify the whole anomaly set, but only by adding the same offset to every data point. No change in relative values.
It’s just as if you had a contour map and then sea level changed. The map wouldn’t change, except for the writing on the contours, to which a uniform value would be added.
This must be the only branch of “science” where “scientists” go to all the trouble to set up a pristine new set of measuring stations a a “reference network”, then begin the anomaly readings at +0.8 instead of zero.
Is it a reference network or isn’t it? Why not start at zero anomaly in 2005?
Ktm
“Why not start at zero anomaly in 2005?”
This makes no sense. Nobody starts at any ‘zero anomaly’.
Anomalies are constructed by
– building the mean of the absolute values over a period;
– subtracting that mean from all absolute values.
{ To remove seasonal dependencies in daily or monthly series, this mean is computed and subtracted separately for every month. }
Thus an anomaly at a given time is not zero because you have set it to zero, but because the difference between the absolute value and the mean is zero.
Month one, you have one reading. You could subtract that reading from itself and get zero. Instead they subtract some other average from a non-reference network.
Now they have 14+ years of measurements from the reference network, yet they still subtract all values from an average derived from the non-reference network.
So now, commentators are forced to say a value today is “below 2005” instead of just saying it’s below average for the dataset or simply quoting the anomaly.
Ktm
“Month one, you have one reading. You could subtract that reading from itself and get zero.”
I know, I use that here:
https://drive.google.com/file/d/1y1zmzMt_1gD5jxCOH13UVYvbocYulbNz/view
to show how small the post-Nino differences are between the two most recent major Ninos
*
“Now they have 14+ years of measurements from the reference network, yet they still subtract all values from an average derived from the non-reference network.”
NO.
Anomalies are constructed not only series by series, but also station by station; anything else would make no sense.
I’ll come back with an example comparing the absolute data of two stations and the anomalies constructed out of them – when I have some idle time for 🙂
Eh, in the update above, Dr Albright has already compiled the data with what I think is a more useful baseline using only USCRN stations.
2005 happens to be near zero, instead of +0.8, and 2019 happens to be below zero. The overall trend since the beginning of the series appears to be negative, as before, but I think the scale makes much more sense for a “reference” network.
Ktm
1. “Eh, in the update above, Dr Albright has already compiled the data with what I think is a more useful baseline using only USCRN stations.”
But Ktm! NOAA computes CRN anomalies on the base of CRN data as well. Why should they use anything from outside?
2. “The overall trend since the beginning of the series appears to be negative…”
It appears indeed, but isn’t.
A trend in a time series is not obtained by drawing a line from the first point to the last one.
It is obtained by using a statistical method named ‘Ordinary Least Squares’, developed 200 years ago by the German mathematician Carl Gauss.
This method is used, amoung all other software, by e.g. every spreadsheet calculator, for trend computation and graphical representation. And so it looks like:
https://drive.google.com/file/d/1t9eR2lsnKANAbGQqT9ZISpsuA3-RGhG2/view
This trend, though represented, is not very statistically significant, due to the extreme deviations from the mean in the series: it is 0.32 ± 0.29 °C , i.e. somewhere between 0.03 and 0.61 °C per decade.
The +0.8C anomaly used on the default USCRN graph is NOT derived from the USCRN network, since that network was implemented between 2002 and 2008. According to USCRN, “the 30 years from 1981 through 2010 provide the basis for the normal period for each month and network.”
Albright used the mean of the entire USCRN dataset, and I agree with his approach to set the anomaly basis. Using a non-reference (non-calibrated) instrument to adjust your gold standard “reference” instrument is scientific heresy.
Nick,
Please concentrate.
If you adjust one temperature value (or more) that has a date inside the reference period of 30 years, you will alter every anomaly value outside that period and so need a new map.
Similar outcome if you alter some values inside and some outside the reference period, unless every value is altered by the exact same amount. If the reference period is unadjusted but any values outside it are adjusted, you need a new map anyhow.
Why do we not see a plethora of anomaly maps over the years?
Geoff S
so lets employ some logic.
A) CRN is the best.
B) Anthony hypothesizes that the other stations in the US are shit.
C) Therefore, if we compare the gold standard with the shit over the same time
period we should see differences
https://www.ncdc.noaa.gov/temp-and-precip/national-temperature-index/time-series?datasets%5B%5D=uscrn&datasets%5B%5D=climdiv&datasets%5B%5D=cmbushcn¶meter=anom-tavg&time_scale=12mo&begyear=2005&endyear=2019&month=2
CRN is slightly WARMER than the “bad stations”. Opps
Don’t put words in my mouth:
“Anthony hypothesizes that the other stations in the US are shit.”
Never said that, you jerk.
And if you understood what is really going on, and how heat sinks work, you’d know why CRN is warmer during the northern hemi winter.
Another Mosher drive by fail.
Never said that, you jerk.
…
Name calling?
…
http://archive.is/e5S64
Satyendra
Mosher works hard to earn that characterization. It is quite fitting, even if most people would consider it “name calling” if it were directed at themselves.
What would you call putting words in someone’s mouth?
Calling him a jerk is quite mild compared to what steve deserves.
Well, thirty years of being called genocidal Nazis can bring one’s dander up, can’t it?
Mosher is one of these inherently dishonest types, who hides it by arguing minutia, while not doing anything to correct the utter Stalinistic alarmism.
God’s in the process, the Devil’s in the details – and that’s where these guys operate.
LOL, you just showed that Anthony is totally correct about American temperatures.
How many toes do you have left ??
You keep shooting them at every opportunity. 🙂
Anthony Watts
The graph shown by Steven Mosher tells us that there is few difference between the CRN stations and all ClimDiv stations.
The same is valid for example when comparing the 71 USHCN stations selected by surfacestations.org as being ‘well sited’, and all GHCN daily stations.
It has nothing to do with heat sinks or whatever else. It has to do with the fact that many stations may measure absolute temperatures differing by a lot, but will show very similar departures from their respective mean over a common reference period.
A typical example is the comparison of Anchorage Intl AP with a CRN station nearby (Kenai 29 ENE), distant from Anchorage by no more than 54 km:
https://drive.google.com/file/d/14aS2UEkD0_Uw2rC05ywbQNdIURBVW-GW/view
Looking at their absolute values would give a well different picture.
Rgds
J.-P. Dehottay
Mosher, USCRN is keeping them honest.
You take the thumb off the scale and suddenly the hockey stick flatlines…
Elsewhere around the world, there is no pristine network keeping them honest, which is why Warmists love polluting out the good USCRN data with global garbage data.
Steven,
There is actually a significant difference, the USCRN trend is 0.12 C larger than that of ClimDiv
Hence, there is a cool bias in the large, mixed quality, urban/rural, adjusted network.
(The figure will be updated for 2019 when ClimDiv data arrive..)
The unit for the differential trend is 0.12 C/ decade, the graph equations show per year..
I see our professional strawman builder, Mosher, is back. And it’s “oops”. “Opps” rhymes with “tops”. One would think an English major would know that. Makes me doubt everything you say when you make so many mistakes in the area you are supposed to have some expertise in.
If they were to use common baselines, that would make it too easy to check their results.
It is very easy to convert to common baselines. I post daily plots with baseline 1981-2010. You can choose alternative baselines and the page will do the conversion for you.
Agreed it is easy.
Then explain why NOAA and NASA don’t use a common baseline when presenting importnat things to the public.
My Answer: Pigheaded holier than thou (I’ve met him in person) Gavin Schmidt doesn’t want to lose his alarmist edge that older baseline gives.
“Then explain why NOAA and NASA don’t use a common baseline”
It’s pretty much like the US doesn’t want to change from F to C. NASA doesn’t want to change because they have a large amount of information published, online and in print, using the old baseline. There is a lot to be said for self-consistency; otherwise you have to keep explaining which baseline you are using. And mistakes will be made.
NOAA is less consistent, and that causes trouble. There are good technical reasons for using a 30 year baseline, and one where you have the most stations with data. They publish data with 1981-2010, and also 1901-2000 as shown above. They argue for the century because it seems a natural period to readers, but I think it just causes trouble.
WMO recommends always using the latest 3 complete decades, so update every ten years. There are technical reasons for that too, but it means everything has to be revised with that frequency.
Stokes
You claimed the conversion is easy. Technically, yes. However, you then said, “NASA doesn’t want to change because they have a large amount of information published, online and in print, using the old baseline.” That demonstrates why what is technically easy becomes a difficult and expensive task when there are a lot of changes to be made.
By resisting making changes that are inevitable, NASA makes it difficult to make direct comparisons with other agencies and independent researchers. Graphical obfuscation can be self-serving for those who are more interested in supporting a position than they are in communicating facts.
Right on, Anthony. And I am better informed (not smarter) after looking up “niggles”, which I think you might have had some commentators in mind when you used it: “niggle” a small but persistent annoyance!
On slide 7 they show all indices relative to base 1951-80.
Look in the document posted by S. Mosher:
https://www.ncdc.noaa.gov/sotc/briefings/20200115.pdf
and go to page 7.
around 1880….Berkeley was a lot colder than NASA
Re page 7.
How many stations were there in the world other than USA in the 1880’s/
How many stations were continuous for the whole period?
And how much of the world is fill-ins that chart.
The outputs are actually problematic and its notable what they don’t show
Place a graph of world CO2 emissions along side figure 7.
https://ourworldindata.org/grapher/annual-co-emissions-by-region
1950 we hit 5Billion Tonnes .. model output 0 degree from a -0.4 start point.
Now fast forward to 2017 36Billion Tonnes … model output let call it 1.0
1950 …. 5Billion Tonnes increase = 0.4 degree change in model
2017 …. 31Billion Tonnes increase = 1.0 degree change in model
6 fold increase only doubles temperature change????
That presentation throws up more questions than it answers.
It looks like a sharp rise because time is the x axis.
I would guess we will never again have a 60 year period where emissions increase 6 fold.
So plot the x axis as CO2 emissions, y axis as model temp output .. very different answer 🙂
You could put an aweful lot of CO2 emissions up for the next 1 degree of rise if the models are working.
Why? There is nothing real there. I can make sh!t up too.
Question for extra credit: Can you find the UHI adjustment? LOL
It would but, to be fair, a glacial maximum would look a lot less scary presented on the same scale.
Well they weren’t using Temps from western WI. Last Jan and Feb was bitter cold
March was nice
April was sh!t cold
May was average
June was hot
July – September was cool
October was warm
November was cold
December was average
This is 4th year in a row that it was too chilly to swim in the lake in Mid August. It was a very wet year.
Where is the warmth?
Derg
“Where is the warmth?”
Oh, for example, in Northern Germany!
We had
– two super warm summers in a row,
– each preceeded by a supermild winter, and
– this one right now is the third in a row.
Snow? Ice? What’s that?
The last really cold winter here was… 2009/10.
Wunderbar!
And the “warming” isn’t causing it. Are you seriously going to argue that 1.8 degree Fahrenheit difference from 1950 to 1980 caused “super warm” and “super mild” winters in a row?
If they were 1.8°F warmer than previous warm ones…then yes.
Loydo,
Nonsense. That small a change in average temperature isn’t in the slight bit noticeable without careful measurement and record keeping. People will differences in precipitation long before average temperature.
Seriously, what level of paranoia rules your life to be afraid of that? And what level of SHEER arrogance to demand that the entire world roll back its standard of living, to suit you?
There was at least a 2 degree temperature change on my twenty-minute drive to work this morning.
Again, I can make sh!t up too. And I can make my own invented fantasies especially damning, if I am allowed to ignore history and facts.
When the day comes that lefties acknowledge all facts and the entirety of history, they will no longer be lefties. Oh happy day…
Luke
“Are you seriously going to argue that 1.8 degree Fahrenheit difference from 1950 to 1980 caused “super warm” and “super mild” winters in a row?”
No. I simply see it’s unusually warming around us since years. That’s all.
On this blog, an incredible amount of commenters claim about the Globe cooling just because their CONUS is!
And I know that there were peculiarly mild winters and hot summers BEFORE purported 1.8° Fahrenheit warming.
Bindidon
Something to consider is how sparse temperature measurements are in Antarctica, Africa, and Siberia, and the poor quality of many of them. They all go into the stew pot to make the global average. If one of the best, if not THE best instrumented regions on Earth is not showing the warming that is touted for the globe, then one should consider the hypothesis that CONUS is a better sample than all the measurements. Mosher has assured us that we can get a representative sample that is much smaller than what is used.
“No. I simply see it’s unusually warming around us since years. That’s all.”
What makes you think it’s unusual? Let’s not think short term here.
As if one country with slightly less than 2% of the world’s surface area should have its temperature anomaly being in lock step with that of the rest of the world.
Donald L. Klipstein
Exactly! I never was able to understand this strange americanocentrism.
But… what is good for Uncle Sam is good for the rest of the world, isn’t it?
“Exactly! I never was able to understand this strange americanocentrism.”
Let’s see, the American surface temperature profile shows the 1930’s as being just as warm as it is today.
All unmodified regional surface temperature charts from around the world show the same temperature profile as the American temperatue profile, i.e., that the 1930’s were just as warm as today.
So, the unmodified American surface temperature profile IS representative of the temperature profile of the globe and saying so is not being americanocentric, since the whole world is represented.
If all the unmodified surface temperature charts from all over the world, regardless of the coverage, show the same temperature profile, then that profile is the temperature profile of the globe to the best of our knowledge. How could it be otherwise?
Here’s a slightly modified U.S. surface temperature chart that shows the 1930’s as being just as warm (actually warmer) that subsequent years (Hansen 1999):
I can show you numerous regional charts from around the world and in both hemispheres that show this same temperature profile.
There are no unmodified surface temperature charts which show the “hotter and hotter” temperature profile of the bogus, bastardized, fraudulent Hockey Stick chart. The Hockey Stick chart is NOT the global temperature profile. It’s a lie meant to promote the CAGW (Catastrophic Anthropogenic Global Warming) fraud.
Bogus, Bastardized Hockey Stick
You can see where NOAA and NASA Climate get their “hottest years evah!” meme, from this bogus temperature chart. Of course, it is all a big lie. It’s computer manipulation as a means of human manipulation. No unmodified regional surface temperature chart from anywhere in the world looks like the bogus Hockey Stick chart. The Hockey Stick chart is all alone. It’s the only chart that shows this “hotter and hotter” profile. It’s a dispicable lie.
I believe Donald gave away something personal, when he mentioned lock step.
I really am amused that there are still people who actually believe that Earth’s temperatures have a “normal”. I would ask what planet, but they keep telling us there is no “planet B”.
best slide in the deck, slide 11
https://www.ncdc.noaa.gov/sotc/briefings/20200115.pdf
NASA GISS compare to IR satellite view of the earth.
Graph I would love to see does Berkley have it
x axis CO2 emissions in tonnes, y axis temperature increase from model output?
Wonderfully rational analysis.
“American Meteorological Society (AMS) there was a press release session that featured NOAA and NASA GISS talking about how their climate data says that the world in 2019 was the second warmest ever”
Except that AGW is a theory about long term temperature trends and NOT about temperature events.
Mr. Layman here. Maybe OT.
Some familiar with my comments may have noticed that I’ve paid attention to my little spot on the globes record high and low temperatures and how they’ve been changed over the years, not just new records but old records changed.
I’ve only used TheWayBackMachine to recover the 2002 list. Since 2007 I’ve copy/pasted some of them “live”.
One thing I’ve noticed is that the web address for the records for my little spot on the globe has changed. (IE, “.com” has been replaced by “.gov” an a few other changes.)
Perhaps a list could be built of the “old” addresses for various other “little spots on the globe” so that people could enter them into the TheWayBackMachine’s search bar and see for themselves how “The temps, they are a changin'”?
PS Love that last graph. Almost as scary as “Sea Level Rise”! 😎
PPS A new record high for my little spot on the globe was set (not just tied) this past Saturday.
It broke the record set in 1890.
What was happening in 1890’s?
“What was happening in 1890’s?”
Responsible people were fastidiously recording temps in well-placed stations.
Who knew they were wasting their best efforts?
Pity there isn’t a plot of the US 48 states from GHCN to compare to the reference network.
A C Osborn
I did a similar job using GHCN daily, by comparing all GHCN daily stations in CONUS (over 8000) with all CRN stations active since 2004 (about 200):
https://drive.google.com/file/d/1zg9M-GZwNoIBln404Ay0voAL8V4PmSdK/view
It is probably not quite what you expect. I guess you would prefer to compare GHCN with the CRN data processing. But… who lives in Europe has no interest in data useful for the USA only.
Rgds
J.-P. D.
Bindidon, thanks for that.
No overall warming over that period then.
“second warmest year ever”
Eric, where do they say that? “On record” is not “ever”.
” when the vertical axis of the climate data is adjusted to fit the scale of human temperature experience, they look less alarming.”
This is very similar to TSI graphs that are used. The vertical axis is just a few W/m2 so very small changes are amplified into looking like very large changes. If plotted on a 0-1600 W/m2 vertical axis you may not even be able to see the changes.
Gunga Din, you made reference to the historic high temperatures in Australia dating back to the 1800’s. It has been suggested that the reason the historic temperatures are not taken into account is that it was possible that the thermometers of that time may have been out by as much as two degrees. Lame I know, given that some of the modern readings are from airports and car parks. I hope that there are historic records being kept somewhere.
These days when the MSM talks about record breaking high temperatures, they seem to cover a fairly short and recent time frame. My own real life experience goes back at least far enough to know that they are telling porkies.
Is anyone able to tell me that when I check my weather app on a daily basis (looking for rain) why each month has at least a week of missing daily highs and lows? And does that mean that the monthly averages are based only on the data that is shown. And I if that is the case then the monthly averages are total BS.
Why should we believe the world is heating up when they refuse to give us correct information.
Present a large amount of daily data is missing, upwards to half, so you are right there should be no trust given to the so called yearly temperature for that reason alone, add in about the dozens of other variables not account for all we have here in creative math which in the means very little. Put out by and for useful idiots. If you were to put this kind of work product in most jobs you would be fired.
Megs,
I said nothing about Australia. ( https://wattsupwiththat.com/2020/01/15/while-noaa-nasa-claims-2019-as-the-second-warmest-year-ever-other-data-shows-2019-cooler-than-2005-for-usa/#comment-2894276 )
My “little spot on the globe” is in central Ohio.
That’s where I’ve seen the records changed.
But I would suggest that you find the list of records for you little spot on globe from it’s official weather site. Make a copy. (I just highlighted and copied the whole thing into Excel. Messy result but “cleanable”.) Then copy/paste into wherever you want to save it. Go to “TheWayBackMachine” (archive.org/web/web.php) and paste it into the Search. Maybe someone had achieved a past list.
If so, compare.
(PS They do appreciate donations to keep it going.)
Megs
“I hope that there are historic records being kept somewhere.”
Of course they are. Even if BoM has decided to restrict the measurement era by starting in 1910, the data before is still available.
You see this when processing the Australian stations present in NOAA’s worldwide GHCN daily data set.
Please don’t compare a weather app with daily professional work.
Bindidon, no comparison was ever implied, but short of studying to become a meteorologist, the weather app is the only option readily available to me to have an idea of how we are tracking in regard to the hysteria about temperature rises every other day. There has been averaging, and distortion of temperature data without proper explanation.
The weather app I have been follwing for years is obviously unreliable at the very least though I suspect it is intentionally misleading.
Our Prime Minister has totally misread the public and is hearing only the activists. That is partly because the only information publicly available to anyone is from a leftist point of view. Our country is likely to go bankrupt because he chooses to further the rolling out of wind and solar renewables at the expense of fossil fuels, and all for nothing benefit to anyone, when we have no other reliable power source as backup.
Yes I am looking to find out the truth, other than this site I have no voice. The fact that I can’t even get accurate data from an app that used to put out all of the data and now doesn’t is extremely frustrating. The whole CAGW premise revolves around the supposed rising temperatures. So it’s not just my weather app it’s about trying to get to the truth about something that could change the direction of mankind, to our detriment. All, potentially based on a lie.
Now it may not have been intentional but you came accross as very condescending, if you feel that my level of reasoning is beneath you then don’t respond to me.
My apologies if I offended you or anyone else for that matter, that was not my intention.
Megs
“… but you came accross as very condescending…”
Ooops?! Please show me where I was, in the few lines I wrote!
It was by no means my intention to be “condescending”. And no: you did not offend me at all.
*
“The whole CAGW premise revolves around the supposed rising temperatures.”
I understand your point. But by accident, due to the bush fires in your country, I tried, from my far Germany, to understand a bit of the recent temperature history in NSW / Canberra / Victoria, and generated out of all available GHCN daily data for that region, a yearly time series:
https://drive.google.com/file/d/1oC13owo5AeumsX8YHaIbGhsJuko2bp-g/view
Sorry, Megs, but while 1880-1970 indeed shows a falling trend, it doesn’t quite look like cooling since then.
But please look at this:
https://drive.google.com/file/d/1O1jKHHMvT3VKyN58uzTFZjW3vtAkoWwC/view
The difference in the daily comparison between 2019, 2009 and 1939 is too small to be the only explanation for this year’s bush fire intensity.
That is, I suppose, the reason why your country installed the FDDI index based on much more than temperature.
Regards
J.-P. D.
Meg, I don’t know if you’ve seen this Kenskingdom blog article, it may be helpful to understand the historic data … and ongoing BoM shenanigans.
https://kenskingdom.wordpress.com/2020/01/16/australias-wacky-weather-stations-final-summary/
Recovery from Little Ice Age. Start the record in 1000 AD and see what happens to the silly claims.
Thanks Chas, I have realised on reflection that I am an interloper on a higher level conversation than I have a right to comment on. I’ll just step back for now and let the scientists do their job. I’ll be back, I need this site.
CRN is probably the very best temperature dataset we have anywhere in the world despite the limited history. And there is very little controversy about it; way less than even the satellite data.
Funny how it’s never mentioned in the mainstream media coverage of “climate change”.
It really deserves much more attention.
Mike Smith
“CRN is probably the very best temperature dataset we have anywhere in the world despite the limited history.”
Correct. But it seems to me that the non-CRN stations could be worse:
https://drive.google.com/file/d/1zg9M-GZwNoIBln404Ay0voAL8V4PmSdK/view
The graph above compares all 200 CRN stations (they all are in the GHCN daily data set), with all GHCN daily stations available in CONUS (over 8000).
This is a real, apple-to-apple comparison because both sets are processed using the same software.
Mike, Not apples to apples, you need to exclude the 200 CRN stations in the comparison. else you are influencing the outcome of the bad stations by including the data of the good stations in the All stations. And when you are trying to discern 10ths of a degree difference, it is important.
davetherealist
“Not apples to apples, you need to exclude the 200 CRN stations in the comparison.”
It seems you replied in fact to ‘Bindidon’.
In theory, you might be right, when we were comparing 200 CRN stations with say 500 GHCN daily stations.
But… there are for the period 2004-2019, about 8000 active GHCN daily stations in the CONUS area (over 18000 for the period 1880-2019).
Look at
ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/daily/ghcnd-stations.txt
There you will find all GHCN daily stations worldwide (over 100000, but only 40000 of them report temperature, what is specified in the inventory file).
I could of course extract the 200, but the resulting values would not differ by more than 0.001 °C.
Recording and reporting average temperatures to hundredths of a degree when the thermometers for the vast majority of the data are accurate to plus-minus 0.5 degrees is simply bad data practice, meaningless to any real Scientist or engineer.
Lincoln said it best, “You cannot fool all of the people all of the time.” This, too, shall pass…
““That 1951-1980 period just so happens to be the coolest period in the 20th century…”
I’m not sure how that can be correct, every year up to the late 1930s was below the 1951-1980 average.
Re the Update, and Mark Albright:
The difference between the choice of base goes directly into the arithmetic, in an obvious way. Albright’s method effectively sets 2005 value to zero (by coincidence, since it is equal to the mean, 2005-2018 whatever the base). But NOAA shows that it is about 0.85 above the 1981-2010 mean. That is why the 2012 peaks differ by the same amount. All values do.
But there is a subtlety. Mark says:
“The beginning year actually varies from 2002 to 2008 depending on when the site was installed.”
That matters, and it is why it is essential to use a fixed anomaly period for all stations. It doesn’t matter which period, but it does matter that they are all the same. Otherwise part of the trend is taken up by a drift in the anomalies. I show that effect using GHCN data here. It is real and big.
Another point is that you should always calculate anomalies based on differences from monthly climatology averages, even to compare annual data.
An analogy here: suppose you were drawing a contour map of Illinois, and you draw contours relative to sea level. Someone objects that that makes the Illinois look too mountainous, so it should be redrawn with contours relative to Lake Michigan. Well, you could do that, and the numbers on the contours would be smaller, but the map and its meaning should be exactly the same. Its the same with changing anomaly base.
”The deviation is due solely to the pattern of missing values.“
This does not apply to USCRN, they aren’t a patchwork of zeroes interspersed with data.
I don’t think it’s too much to ask that the scale of the “reference” network be set using only data from the “reference” network. Alright attempted to do this, and I give him credit. If you don’t fully agree with his methodology, please explain how YOU would set the anomaly for USCRN using only USCRN data.
Ktm
“If you don’t fully agree with his methodology, please explain how YOU would set the anomaly for USCRN using only USCRN data.”
1. But… this is exactly what I have shown you above (I didn’t use the CRN data because it is homogenised, but rather the raw, completely unadjusted data provided by the GHCN daily data set, within which the USCRN stations all are incorporated).
Here is, to have it similar to NOAA’s and Albright’s plots, a yearly plot of the raw CRN data:
https://drive.google.com/file/d/1t9eR2lsnKANAbGQqT9ZISpsuA3-RGhG2/view
The plot differs from theirs because it is not CRN’s homenised data.
*
2. You still don’t understand that there is no need to set the first anomaly at zero.
Once more: anomalies wrt any reference period you choose are generated out of absolute values by
– computing the mean of the absolute values (for each year / month / day, depending on how fine your time series is)
– subtracting this means from all absolute values.
Did you never had a look at the UAH time series managed by Roy Spencer?
The reference period (baseline) he uses has nothing to do with the aim of starting at zero: it is a period proposed last year by the WMO, replacing the previous one which was 1971-2000. Spencer anticipated this, and moved around 2013 from his previous period (1979-1998) to 1981-2010.
Why should Spencer choose an anomaly reference period useful only for his local needs?
*
3. What matters when you compare anomalies from different providers is to use the same reference period for all time series.
Otherwise you get nonsense like this:
http://www.woodfortrees.org/graph/hadcrut4gl/from:1979/mean:36/plot/gistemp/from:1979/mean:36/plot/best/from:1979/mean:36/plot/rss/mean:36/plot/uah6/mean:36
No wonder: you were comparing anomaly data wrt the means of
– 1981-2010: UAH
– 1979-1998: RSS
– 1961-1990: HadCRUT
– 1951-1980: GISS, BEST
To get it right, you must displace the time series to have them all on par:
http://www.woodfortrees.org/graph/hadcrut4gl/from:1979/offset:-0.29/mean:36/plot/gistemp/from:1979/offset:-0.42/mean:36/plot/best/from:1979/offset:-0.35/mean:36/plot/rss/offset:-0.13/mean:36/plot/uah6/mean:36
*
All these time series use only their own data for anomaly construction. Why should this be different for USCRN?
“they aren’t a patchwork of zeroes interspersed with data”
Many of them have missing data at the start, which is the worst case from the point of view of bias.
The proper way to do it is to take a fixed base interval for all stations, even if it is just the last ten years.
All the slides here are good. The best may be the one which compares USHCN Adjustment vs Atmospheric CO2. with an R2 of 0.973. That looks like a high mark. Perhaps Gavin Schmidt can explain it.
The attachment was missed:
All the slides here are good. The best may be the one which compares USHCN Adjustment vs Atmospheric CO2. with an R2 of 0.973. That looks like a high mark. Perhaps Gavin Schmidt can explain it.
https://realclimatescience.com/61-fake-data/