Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Donate

Twitter Facebook YouTube Pinterest MeWe

RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

New? Register here
Forgot your password?

Latest Posts


1934 is the hottest year on record

What the science says...

Select a level... Basic Intermediate

Globally the year 1934 was cooler than the 20th century average.

Climate Myth...

1934 - hottest year on record

Steve McIntyre noticed a strange discontinuity in US temperature data, occurring around January 2000. McIntyre notified NASA which acknowledged the problem as an 'oversight' that would be fixed in the next data refresh.  As a result, "The warmest year on US record is now 1934. 1998 (long trumpeted by the media as record-breaking) moves to second place." (Daily Tech).


The year 1934 was a very hot year in the United States, ranking sixth behind 2012, 2016, 2015, 2006, and 1998. However, global warming takes into account temperatures over the entire planet, including the oceans. The land area of the U.S. accounts for only 2% of Earth's total surface area. Despite the U.S. sweltering in 1934, that year was not especially hot over the rest of the planet, as you can see on the 1934 map below. Globally, 1934 temperatures were actually cooler than average for the 20th century.

1934Global T map 1934

2016global T map 2016

Figure 1. Global temperature maps for 1934  (top) and 2016 (bottom). Source NASA.

Climate change skeptics have pointed to 1934 in the U.S. as proof that recent hot years are not unusual. Choosing the year 1934 is an obvious example of "cherry-picking" a single fact that supports a claim, while ignoring the rest of the data. In fact they have to cherry pick both a location (the U.S.) and a year (1934) to find data that is far from the global trend. Globally, the years 2014, 2015 and 2016 are the hottest on record, so far.

Global T anomalies


Figure 2. Global land and ocean temperatures from 1880 to 2015. Source: National Climate Data Center


The fact that there were hot years in some parts of the world in the past is not an argument against global climate change. Regional and year-to-year temperature variations will always occur. The reason we are worried about climate change is that on average, over the entire world, the long term trend shows an undeniable increase in global surface temperatures and global ocean temperatures. This rapid global heating is dramatically altering the planet we live on.



Last updated on 7 August 2017 by Sarah. View Archives

Printable Version  |  Offline PDF Version  |  Link to this page

Argument Feedback

Please use this form to let us know about suggested updates to this rebuttal.

Related Arguments


Prev  1  2  3  Next

Comments 51 to 100 out of 122:

  1. Darkmath,

    Did you really post an average temerature of the USA that simply averages all temperatures without adjusting for differrences in number of sites, location of sites or time of day of measurements?  If you really think that is acceptable data you are welcome to think whatever you want.  Real scientists don't agree and you will convince no-one here that you make sense.

    Your suggestion that the world wide "climate model" has warming built into it is completely false (in addition to being sloganeering).  The adjustments to the early ocean temperatures substaintially raise those temperatures so that the slope of the line is lowered.  see this SkS article.

    adjusted versus raw temperature

    The denier websites you have frequented have mislead you.  If scientists models currently raise temperatures why would they lower the slope with their early adjustments? Your argument is contradicted by the data. 

  2. DarkMath @48, if you don't like NOAA, you can always use the AGW denier funded Berkeley Earth Surface Temperature project results:

    Technically, the BEST data series make no adjustments.  Instead, when there is a known, or reasonably inferred change of equipment, location, or time of observations they treat the data as coming from two distinct stations - a proceedure which Anthony Watts endorsed as having his full confidence (until he saw the results).  It is certainly a proceedure that has the full confidence of Judith Curry (denier enabler), Richard Muller (temperature series skeptic), Zeke Hausfather (luke warmer) and Steven Mosher (Luke Warmer), not to mention three independent scientists selected by at the time, climate skeptic Richard Muller.

    For the record, the highest ranked running 12 month mean temperature in the 1930s according to BEST ranks 23rd.  In contrast, eight of the 12 highest ranked 12 month running mean temperatures are in 2012, with another three in the last three months of 2011.

    But you want to cherry pick just July temperatures.  However, the highest ranked July temperature in the 1930s is 1936 (ranked 3rd) followed by 1934 (ranked 6th).  In contrast, in the 21st century the highest ranked are 2012 (1st), 2006 (2nd), 2011 (4th), and 2002 (5th).  The average July temperature across the 1930s was 0.66 C.  Across the 21st century (to 2012) it was 0.85 C.

    And if you are wondering, BEST uses approximately 8 times as many stations as does the USHCN, with an increasing number in time over the 20th and 21st century.

    In short, your cherry pick of the cherry pick still does not give you the conclusion you desire.

    Your only refuge is to insist that when a station changes its instrument entirely, or its time of day for observations, or is moved to a new location, it should be treated as the same station with no adjustments for differences in recorded temperature between the new and the old; and to take meaningless arithmetic means that do not care that the station density in New York is far higher than that in Nevada, thereby giving more importance to North Eastern state temperatures than to those in the mid-west or west:

    Your bias in favour of rich, Democratic eastern states is noted.

  3. Tom Curtis, you have valid points about why the temperature needs to be adjusted and/or estimated. But the problem is those adjustments and estimates change over time:

    The more the temperature record changes the less confidence I have in it.

    Then there is the discrepency between NASA's land and satellite temperature data. They don't always match up. For example here:

    I don't have any skin in the game here. I have a strong science background but only deal with medical data all day long. I'm an objective observer of climate science. But I got to tell you though is the more I read the more I think the science definitely doesn't appear "settled". And given I've become an expert in observing human scientific endeavor over the past 30 years, :-), I always expect the worst.


    [TD] All the adjustments, both procedures and individual raw and adjusted data, are publicly available--along with the rationales for the repeatedly improved adjustment methods. You are welcome to state on SkS your specific objections to any of those procedures, rationales, or data. But you are not welcome to simply state, without referring to any of that evidence, that you just don't trust the adjustments, because that implies that you don't trust the scientists no matter how publicly and thoroughly they document their work, which implies that you are unwilling to discuss evidence. The SkS comments are for evidence-based discussions. All that I've just told you to address, must appear in the thread that moderator PS pointed you to, not this thread.

    Regarding satellite measurements of temperature, read the post "Which is a more reliable measure of global temperature: thermometers or satellites?" Comment further on that topic on that thread, not this one. I strongly suggest that before commenting there you also read the post "Satellite measurements of warming in the troposphere"--all three tabbed panes, but especially the Advanced one. You should hesitate to assume the satellite temperature indices are superior to surface and balloon indices especially because those satellite indices started to diverge from both surface and balloon indices around the year 2000 when the satellite instruments were switched.

    [RH] Please limit images to 500px.

  4. michael sweet: "without adjusting for differrences in number of sites, location of sites or time of day of measurements"

    There are many different ways to clean up data. For example adjusting for changes in the time of day. Say a weather stations data from 1900 - 1950 was taken at 12:00pm and from 1950 - 2000 it was taken at 1:00pm.

    NASA/NOAA's approach to clean the data is by far the most complicated. It didn't have to be. You could start out in this case without making any adjustments at all. You simply treat them as two separate data sets. Voila. All that matters is that weather stations change in recorded temperature at a specific time.  You reduce it to a rate of change, one set for the 12:00pm and another for 1:00pm. It no longer matters what time the temperature was taken because using 2 datasets instead of one removes time of measurement as a variable. The same would hold true for an elevation change or change in location. 

    NASA/NOAA didn't even attempt this simple option. Instead they went with some hefty calculations to make that one stations 100 years of data appear if it was all taken at the same time, the same location and same elevation. That's great, it's worth doing but shucks that is a lot of work. If anything the much simpler "treat a change in time/elevation/location" as if it were another weather station technique could be used to validate the more complicated approach. If I've learned one thing in my engineering career it's always go with the simplest option first. 


    [PS] This is getting offtopic here. Please post any responses to Darkmath about how the temperature record is adjusted to "Temp record is unreliable".

    Darkmath - put any further responses you wish to make over adjustment there too please.

    [TD] Darkmath, read the Advanced tabbed pane on that thread before commenting there. You also must actually read Tom Curtis's response to you, in which he already explained the BEST team's approach of not making any adjustments. Then you must respond to his comment explicitly (on the thread that moderator PS has pointed you to). I write "must," because SkS comments are for discussion, not sloganeering. "Sloganeering" includes failing to engage with respondents substantively.

    [JH] I flagged DarkMarth for sloganeering upstream on post #48. I also advised him to read the SkS Comments Policy and to adhere to it.

  5. DarkMath:

    In comment #53, you state:

    I don't have any skin in the game here. I have a strong science background but only deal with medical data all day long.

    In comment #54, you state:

    If I've learned one thing in my engineering career it's always go with the simplest option first.

    Does your "engineering career" include dealing with medical data all day long? 

  6. DarkMath @53, first, when comparing temperature anomalies (as shown in the two graphs above), you need to provide them with a common baseline - ie, the interval with a mean temperature of 0.  Failure to do so can create a strong apparent visual discrepancy even between temperature series which are isomorphic.  It is very evident in your first graph that no such common baseline is calculated, with not even a single data point having common values, let alone a 30 year period with a common mean.  If you employ a common mean, the comparison looks like this:

    Second, there are several differences between the 1981 and the 2016 product.  Of these, the most important are:

    1)  An increase in the number of reporting stations from around 1000 (1981) to around 2200 (1987), to around 7,200 (1999-to 2015), to 26,000 (2016 but possibly not yet implimented).  The differences in station numbers reporting at a given time between 1987 and now are shown below.  In 1981, the number of reporting stations is half that of 1987, though no doubt following a similar pattern over time;

    2)  The introduction of adjustments for station moves, instrument changes, etc from the 1990s (detailed in a 1999 publication;

    3)  The introduction of adjustments for the urban heat island effect (1999), and the switch of classification of urban areas from a classification based on population to one based on the intensity of night lights as observed from satellite (2010);

    In addition for the full global Land Ocean Temperature Index, the use of Sea Surface Temperature data started in 1995, and the way temperatures over sea ice was changed in 2006 to better reflect the fact that sea ice insulates the overlying air layer from the SST.  These changes do not effect the above graph, which are based on the Meteorological Station only data.

    IMO, it is thoroughly unrealistic to expect such changes to have no impact on the estimate of global temperature.  Nor is it realistic to expect that because in 1981, no adjustments for station moves etc (because Hansen did not have access to the station metadata to make such adjustments possible, nor under undertaken the research that would provide the theoretical justification for how those adjustments are made), nor to incorporate more data as it becomes available.  And now that you are on record as endorsing the methodology of the BEST temperature series (@54), here is a comparison of the BEST temperature series with that from the NCDC, which uses the same data as does GISTEMP:

    Clearly, if we are to trust the BEST temperature series, we should conclude that the adjustments by NOAA, and ipso facto by GISS, have improved the data.  Further, if we adopt the logic that we should automatically distrust measurements which purportedly improve over time, we should not accept modern determinations of the speed of light, which have inflated by 50% over the original measure in 1675.

    With regard to the satellite data:

    1)  The satellite, TLT data measures a weighted average of atmospheric temperatures from 0 to 12,000 meters, whereas the surface temperture data measures a hybrid of the 2 meter air temperature over land, and the SST over sea.  These are not the same thing, nor are they expected to change in lock step:

    2)  The satelite TLT data is far more greatly effected by ENSO and volcanic temperature fluctuations, making it much noisier.  As a result of this, the strong El Nino in 1997/98 along with the strong La Ninas in 2008 and 2011/2012 have a much larger effect on the short term trend post 1998.

    3)  The satelite TLT data has four or five major versions from different teams, all using precisely the same data but with much larger differences in trend etc than the different land surface series (most using different data, and all using different methodologies).  Prima facie, that indicates the TLT temperature series is less well known than is the surface temperature series.

    4)  The particular satellite temperature series you use (RSS) has just had a major revision increasing the post 1998 trend in its TMT dataset.  That revision will have a similar impact on the as yet unrevised TLT dataset once the revision is made, so we know the data show in not currently accurate.

    The use of satellite data to construct a temperature series requires far more adjustments than is required for the surface temperature series; and there is no consensus among those working in that field as to the correct way to make those adjustments.  Further, as noted above, the different way of making those adjustments has a significant impact on the final product (unlike the case with the surface products, where different methods come up with essentially the same result).  Given that, in a case where surface and satellite data disagree, there is no question that the surface data should be considered a more reliable indication of the surface trend.


    [TD] Thank you, Tom, for carefully reading DarkMath's comment, for responding specifically to his/her points, for responding in detail and thoroughly, and for responding with referenced evidence rather than handwaving, personal incredulity, and implications or even accusations of conspiracies. DarkMath, please follow Tom's example in your commenting style.

    (Tom, we are trying to move this discussion to the appropriate threads, so in future please respond to DarkMath on those other threads.)

  7. DarkMath,

    I responded to you here.

    Most experienced readers follow the comments page here where all your posts, and the responses, will show up.

  8. The science of carbon is not up for debate. it is well known. And it is well known what carbon does in the atmosphere, it warms it. And we know where the carbon is coming from, human activity. A small child of three could look at the photos of the antarctic and tell you whays happening, the ice is melting. When it all melts, 200 feet of sea level rise, billions displaced a world we can hardly recognize today. All for burning petroleum, what is a unique and precious chemical goldmine, that  cannot be duplicated in the laboratory. And we burn it, to enrich a few people. There are more jobs, more money and a cleaner future in renewables but greed and ignorance have prevented it. We now talk about protecting cola miners, the most dangerous job in the world, instead of retraining coal miners to work on renewable energy programs like wind and solar. Why do we want to keep them and their children down in the mines, greed and ignorance. When you hear "drill baby drill" or "energy voter" you can count on a deep and brutal ignorance of the facts.

  9. The chart of global temperature on this page in 1934 appears to be exceptionally misleading. As I understand it we have nothing like so clear a picture of global temperatures in 1934, with significantly less than 50% global coverage and many areas having only a handful of readings. Such charts do not appear to be justifiable.


    [DB] Not counting 2018 (which is almost ready for inclusion), 1934 is the 7th-warmest year in the US.  You can look this up yourself.

    US Temperatures

    Globally 1934 is nowhere near the warmest year, coming in at the 86th-warmest.



  10. LTO:

    Please provide a reference to support your wild claim that Global temperature in 1934 was inadaquate.  BEST (financed by deniers) starts global coverage in 1850 and GISS (more conservative) starts at 1880.  Both  are way before 1934.

  11. Michael: See here:

    I find your response disingenuous. What % of the globe do think was being sampled at least once a day in 1934, or indeed 1880? Common sense would tell you it's relatively low, with the southern hemisphere exceptionslly low. What were the 'deniers' in 1850 denying, pray tell?

    The 1934 chart pretends to have accuracy to a few degrees Fahrenheit. Independent of whether you believe in AGW, this is fanciful thinking.

  12. Sigh, if you want to rely on John McLean, then you will never want for moonshine. See here. A pretty simple check is construct a temperature series from the GHCN stations that have been around since 1934 and see if you can spot the difference. See here for time series with just 60 stations for comparison and also a proper discussion of coverage bias.

    The chart does not pretend any such accuracy - go to the appropriate papers for each of the temperature records to see what the error bars are.  

  13. LTO @61 ,

    your link is to the work of Dr John McLean.

    To add to Scaddenp's comment: The short story is : McLean has made a fool of himself.  And not for the first time.

    Please, LTO, try to be logical and scientific in assessing important issues, such as AGW.   Everywhere you look on science-denier websites, you find deluded crackpots who continue to tie themselves in knots . . . cherrypicking and/or doctoring data . . . doing all sorts of crazy stuff in trying to deny the "bleeding obvious".   LTO, you owe it to yourself to dig deeper and really look into the rubbishy propaganda (which you seem so attracted to).

    Check out Andthentheresphysics on Dr McLean's ideas.  Plenty of other respectable sources critiquing his nonsense.  ( In particular, the McLean paper is an exercise in triviality. )

  14. LTO:

    Your citation is to an obscure PhD thesis.  Here is a discussion of the thesis from And then There's Physics.  The thesis states "The audit covers a broad range of issues but leaves the quantifying of the impact of such errors to others".  That means the writer has not checked to determine if the supposed "errors" affect the result.  All this data was reviewed and argued about in the 1970's.  Scientists agreed that the data was properly collected and analyzed.  You are 50 years too late.  An unreviewed PhD thesis cannot be compared to papers published in Nature and AScience.

    Please provide a peer reviewed citation to support your wild claims.

    NASA GISS averages their data over 1200 km.  They have good coverage over the globe since 1880.  You can check their errror bars at their web site here.  Scientist have determined that the data since 1880 are sufficient.  It is well known in the scientific community that the HADCRU4 record does not have very good coverage of the globe.  That is why their estimate of warming is too low.  Other records like GISS and BEST have better coverage.  

    Common sense tells me that the data is sufficient since the IPCC report, accepted by every nation on the globe, accepts the data.

    In 1850 there were no deniers.  Everyone agreed that CO2 would casue an increase in global temperatures.  By 1896 Arhennius had estimated the increase from doubling CO2 and got a number that is still in the range of sensitivities.  Here is his peer reviewed paper.

    For someone who is just starting to learn about AGW you are very well informed about obscure denier papers.  You are aware that most of the deniers have given up arguing because the evidence of warming is so obvious that it is not necessary to even measure the temperature any more.  Rising seas, disappearing ice, fire storms and unprecedented hurricanes all tell a story.

  15. Hi everyone

    Sounds like there's some history with this McLean fellow, but let's set it aside for now, as whether or not he's said silly things about other topics is neither here nor there. A phd thesis is absolutely peer reviewed, and thoroughly challenged. Mine certainly was, admittedly at a far more renowned university, but snobbery on such matters is uncalled for. Your comments on peer review and Science/Nature are a bit naive generally, but particularly so in the wake of this debacle

    Appeal to (lack of) authority is not science. Nor are we bound by what some now-dead scientists thought 50 years ago (notably when they thought a new ice age was upon us). I've learnt a lot from this site, but the 'ignore that person because he's an idiot' line of argument is not persuasive. Play the ball, not the man.

    Michael: Arhennius was hardly the last word, as you presumably know. I'm not really aware of much of the past GW politics (or interested in it), having previously taken it at face value. I recently became interested when someone I respect - Scott Adams - started looking at it. Do follow Scott's discussion on twitter / periscope, I'm sure he'd find your contributions useful.

    Back to the topic at hand. Nobody has yet answered my questions, so I'll formalise it. So NASA GISS averages out temperatures over 1200 km? That's almost the length of the UK, which in itself raises an eyebrow from someone who lives in London and is familiar with the weather in scotland. You probably mean 1200 sq km(?), but this still covers many degrees C of gradation in the UK and probably most places in the world. Nevertheless, let's go with that for now, which equates to ablut 42,000 grids globally, 21k in each hemisphere. Please correct if wrong.

    1. What % of all the ~42k (or however many therr are) grids had daily temperature readings from at least 10 different locations, split out by north/southern hemispheres percentage, in 1880, 1920, 1934, 1960 and 2000 respectively?

    If the answer for 1936 is >80% I'll withdraw my criticism of the chart.

    2. As above, but the % that had at least one daily max/min temperature reading within each grid for those years.


  16. Calm down please, LTO.   We were discussing Dr McLean's work.

    And I am sorry you are not cynical enough to realize that there are PhD's . . . . and there are PhD's.    To put it politely  ;-)

    Dr McLean is criticized because he puts forward idiotic ideas ~ and more than one idiotic idea and on more than one occasion.   He is a repeat offender (and therefore deserves no presumption of innocence).   The likely explanation is that his emotional bias provides Motivated Reasoning for his intellect to deny the "bleeding obvious".   This is very typical of denialists (of all levels of intelligence).

    Even you yourself, LTO, should try some introspection to identify the underlying causes of your apparent determination to oppose the scientific evidence by means of rhetoric & sophistry.   Look at the overall picture please.   Melting ice, rising seas, alteration of weather patterns, migration of plant & animal species in response to global warming [global warming at a time in this interglacial when the world had been on a natural multi-millennial cooling trend].   All "bleeding obvious" ~ and irrelevant as to whether you classify Year 1936 as a this or a that.

    LTO, if you are a true skeptic, then you will present some reasonable evidence to support your "viewpoint".   But so far, you have only made handwave rhetorical comments.   There is a reason why (over recent decades) the number of climate scientists disagreeing with the mainstream consensus . . . has steadily dwindled to a minuscule minority.   Quite simply: they have no valid evidence to support their (often mutually contradictory) assertions.

    LTO, please get your act together, and present something substantive.   And good luck with that!   Indeed, I suspect you will need Divine intervention more than good luck  ;-)

  17. LTO,

    You demonstrate again a deep knowledge of denier literature, contrary to your claimed recent introduction to AGW.  You have chosen a particularly obscure issue to hang your hat on.  I cannot find a reference with 30 minutes of GOOGLE time.  This demonstrates that  the issue is not important even to deniers.  Please link the denier site (and the post about global coverage) you are getting your information from.

    Hansen 2006 discusses the problems with the HADCRU data set.  That is the one referred to in your PhD thesis.  As you can see, Hansen beat McLean in finding this issue.  Hansen discusses how GISS resolved the issue so that they are not affected.  BEST is also not affected.  I note that the HADCRU issue results in HADCRU underestimating global warming because they do not include the Arctic and Antarctic.

    Cowtan and Way web site discuss the issue in more detail and show how they correct the HADCRU issue.

    Your attitude has changed from someone who claimed actual questions to someone demanding answers to obscure denier garbage.  I am not your GOOGLE boy.  Unless you make particularly wild claims I will no longer respond.

  18. LTO,

    According to this video (, the BEST record covers 80% of the Earth's land area from about 1900 to 1950.  Only the Antarctic continent is not covered.  From about 1950 over 95% of Earth is covered.

  19. Hmmm. Intriguing change in tone. I don't yet have a viewpoint; what I'm trying to do is evaluate the evidence being presented. 

    First, thank you for putting in the time to try and find an answer - I really appreciate it, even if it's made you grumpy in the process. Having done some fuether research myself, I have some answers. 

    First, I see that the 1200 km figure is actually a 'smoothing radius', which assumes that a climate measuring station within 1200 km 'influences regional temperature'. Again, that is the length of Britain, and only a radius, so the diameter is twice this. A bit odd on its face, but depends how the smoothing is done I suppose. Note: it appears to come from a 1987 paper discussed below. Dodgy, but not necessary to go into now.

    I also had success on the gridding, and it looks like GISS breaks the globe down into 16,200 grids, (each presumably ~31,500 sq km - ie size of belgium) which are used to build the charts above. I base this on the data you can export from their site. So my question can be reformulated as:

    1. What % of all the 16,200 grids used to create the charts above had daily temperature readings from at least 10 different locations, split out by north/southern hemispheres percentage, in 1880, 1920, 1934, 1960 and 2000 respectively?

    Michael, an aside: That you can't find the information could be a sign that the question isn't important. However, given that the question is in essence one of how good the coverage of actual measurement data is and therefore what inferences can be drawn from it, the question seems to me to be of primary importance. You may well take the view that if this Hansen fellow says something then it must be true, but as I said earlier appeals to authority are not science. It's not very reassuring if you can't answer basic questions about the quality of the data set you're relying on and using to draw trend lines,

    The discussion in the Hansen paper you cite is trivial and adds nothing. It does however link to a 1987 paper that was perhaps the foundational work for this data set. Link is here:

    To my pleasant surprise Fig 1 goes some way to answering the question, which you can see here:
    Each circle has a diameter of 2,400 km (two Britains!) and within it a single meteorological station. Figure 2 shows the globe divided into just 80 grids(6 million sq ft each!), and you can see that for many grids continuous coverage didnt even start until well after 1934, and further the number of stations in many is tiny (far fewer than 10) despite covering enormous areas that will have variances in temperature of many degrees C.

    The paper is an absolute must read, if you can do so skeptically. Hansen's done a good job with a very limited data set. The problem is thst that data now appears to be being massively overinterpreted.

    For fun, I overlaid the 1930 station coverage from Hansen's 1987 paper against the 1936 chart on this page here: You can see the chart is just making up data showing a dramatic 4F decrease in temperature across much of the globe despite there not being a meteorological station within many thousands of kilometers. Remember each circle is two Britains wide, and contains just one meteorological station.

    So can I answer my own question? Unfortunately not, but I can answer a similar questions using Hansen's 1987 paper:

    Q: if the globe was divided into just 80 grids of roughly 6 million sq km each, how many contained at least 10 meteorological stations in 1987? For reference Australia is just 7.6 million sq km.

    A: Roughly 65%

    Given the explosion of stations in 1960 comapred to 1930, the answer for 1934, even at such a low resolution, would have been much smaller.

    My conclusion from all of this is unchanged: that to try and pretend that you can show a chart of global temperatures in 1934 with certainty of within a few degrees F is totally misleading. It doesn't pass the sniff test.

  20. Michael: That video isn't what it purports to be. The percentages appear to be of mathematically sampled land area, not land area that actually had a weather station on it. Further, the analysis of past data has a pretty major assumption:

    "Our calculation assumes that the regional fluctuations in the Earth’s climate system during the entire study interval have been similar in scale to those observed in the reference period 1960 to 2010'

    Ummm... How can that be justified, if ihe period from 1960-2010 is apparently one of unprecedented climate change?

  21. LTO, it seems to me that you are focused on absolute temps, whereas the global avg. temp. reconstructions are given in anomalies. The distinction is important, and this series of posts explains it all very well: Of Averages and Anomalies, especially Parts 1B and 2A, for your other hang-up on "coverage".

    scaddenp, up thread, pointed you to this post at AndThenTheresPhysics which is the most recent look at the amazing fact that you don't need thermometers covering every sq. meter of the globe to get a good sense of how the temps are increasing. One of the first to do this analysis is Nick Stokes here: Just 60 stations.

    You can do so yourself using Kevin Cowtan's "temp tool".

  22. LTO:

    Regarding your question about how BEST justifies averaging anomalies over large areas I will point out that the BEST study was financed by the Koch brothers (fossil fuel deniers) for the specific purpose of finding errors in the surface temperature record.  No errors were found.  I presume that their data analysis would withstand rigorous examination since it was designed by deniers.

    I am interested to find out that you are so expert at temperature records that you can dismiss the work of multiple scientific groups for the past 50 years without even reading their papers.  Arguing that you do not believe scientists can average anomalies over 1200 km is simply an argument from ignorance.  Arguments from ignorance do not carry any weight on this web site, you must provide evidence to support your wild claims.

    Your tone changed so I changed my tone.

  23. David Kirtley @71 , 

    thank you for that reference to Nick Stokes's "Just 60 stations".

    I recalled him saying that he could get a good approximation of global temperature change from a fairly small number of observation stations [ less than 100 ] . . . but I did not recall the exact number he had used in his test case.   ( Also, slightly amusing to see the paucity of USA continental stations used in the analysis! )

    All of which, is leaving LTO's argumentation looking even more hollow.

  24. The 1200km correlation in temperature anomalies comes from the data, and while initial work done in 1987, it has been reproduced by numerous workers. And the reason is no great surprise either - 1200 km is about the size of a weather system.

    Please note the anomaly definition, it is critical. It is saying that if have a station that is measuring say 2 degrees above the local  average for that thermometer, then you expect thermometers with 1200 km to also be measuring 2 degree above their  local  average, especially if you consider monthly average which takes the time factor of the weather system out of it.

    Absolute temoperatures vary wildly over very short distances - that is why anomaly methods are used.

  25. Hi all

    Thanks for your responses, but either you've misunderstood my criticism or you aren't being serious. Let me restate it.

    The 1934 chart in this topic purports to show the 'global temperature' for 1934 was cooler than "average for the 20th century", with massive swathes of the globe and southern hemisphere oceans and Antarctica in particular apparently being multiple degrees cooler. It is an absolute nonsense to suggest we know anything about the temperature across the globe to this level of resolution. There is a huge fraction of the southern hemisphere that was much more than 1,200 km from the nearest measuring station in 1934. ( See ) There wasn't even a single weather station on Antarctica then (1.5x size of continental US), but you'd like me to believe it was 1-4 degrees colder than 'average' in 1934.

    Further, I don't believe that the 1,200 km radius is valid for temperature anomalies over the ocean to any certainty, where currents must play a far greater role than on land. In other words, the error across the chart is far greater in most places than the purported effect. It is totally misleading and asserts something we don't know.

    As to the people talking about needing only a handful of measuring stations. Unfortunately you appear to have not read the fine print. This is only if they are very strategically placed around the globe. They were not strategically placed in 1934, with massive gaps. One of the big elephants in the room with historical temp records are ocean temperatures.

    As always, thoughts on what I may have got wrong very welcome. Please do acknowledge and quantify uncertanties. Thanks!


    [PS] This is verging on sloganeering and strongly suggests you have either not read or understood the resources offered to you on the anomaly method. Arguments from Personal Incredulity have no place here. Either present data supporting your claims or show us the faults in the published analyses of SST data. Scientists do the hard sweat over data.

    If you are determined not accept the science, then this site is not for you. I would encourage you however to engage in some critical thinking and decide what data would change your mind. Then we might be able to help.

  26. LTO @75,

    The graphics in the OP show the distribution of "global temperature" for 1934 & 2016, but not the actual "global temperature". You appear to be disputing the method by which that distribution can be extended to places that have no measured temperatures.

    Assuming you are happier with 250km smoothing (as opposed to the more normal 1,200km), here are two graphics based on the same GISS data with such 250km smoothing.GISS 2016 temperature mapThe ERSST data has less gaps, even in 1934, probably because the weather stations sail around the oceans and are thus always filling in gaps. Or it may be that the methodology used in ERSST isn't impacted by the 250km smooting button at GISTEMP.

    You will see in the top right corner of these graphics the resulting global temperature anomaly. Note that the 1934 global anomaly (1900-2000 base) is still negative. It would be a very brave man who suggested these graphics do not entirely support the message of the OP.

  27. Oops!!
    I've lost a graphic @76.

    Gistemp 1934 temperature map


    [DB] Try this:

    GISS Temeprature Map

    Vs this:


    GISS Map Graphic output is in the form of a temporary file and need to be hosted somewhere.

  28. LTO,

    Since this is a scientific blog you must support your claims with citaations to the relevant literature.

    Please support your wild claim " It is an absolute nonsense to suggest we know anything about the temperature across the globe to this level of resolution".  This data was hashed over in the 1070's and 1980's.  In the 2010's the BEST group, specifically designed by deniers to find errors in the conclusions of GISS and HADCRU, found that the evalations of GISS and HADCRU were accurate.  On what basis do you now claim errors??  You initially claimed that you had no relevant experience with this data.  Now you claim that on your own  authority all scientists are wrong.  That is a very weak argument.

    Your second wild claim " I don't believe that the 1,200 km radius is valid for temperature anomalies" was also resolved in the 1970's.  GISS went with 1200 km and HADCRU with 250 km.  Data accrued since then has shown that HADCRU is less accurate than GISS.  On what basis do you challenge this scientific conclusion?  It appears that you are again citing your experience without reviewing the data.

    I look forward to your citation of peer reviewed data to suppoprt your wild claims.  In the absence of peer reviewed citations you must at least read the relevant literature before you make the claim that scientists have incorrectly concluded 1200 km is accurate.  Since the deniers who funded BEST were unable to find any errors I expect you will be hard pressed to find any.

  29. Hi Michael

    I’m afraid you’re confusing different things. Please try to stay on topic. Merely posting irrelevant guff in an offensive manner trying to shift the burden of proof isn’t conducive to a productive conversation. If you feel a need to caricature someone’s arguments in order to address them, it just comes across as looking like you don’t know what you’re talking about and are trying to mask this. I’m certain that isn’t the case, and if you feel like addressing my actual comments using science rather than bravado and appeals to authority  I’d be really interested to hear your thoughts.

  30. Hi MA

    Could you use or similar to post links to your graphs? I can only see a 2016 one.

    I’ve had a quick look at the ERSST data set available here  A sentence jumps out:

    ”Note that the data are more reliable after the 1940s”

    Again, this goes to my point: the chart suggests we have spatial resolution to map the temperature anomaly in 1934 against an average that runs from 1900 to 1999 in areas of the world, including the antarctic, where we had virtually if not literally zero measurements within 1200 km until well after 1934. This is a nonsense.

    Based on the rather rude ideological Responses I’ve had from some people, I suppose it’s necessary to state explicitly that my criticism here is specifically about that chart giving a totally misleading impression of the information available in 1934, particularly in the Southern Hemisphere. Not a general global warming criticism. If your model pops out an anomaly for a region of -2 against a heavily derived baseline for a region in which there have been zero reliable measurements, it simply fails the red face test (unless you acknowledge the error is probably something like 10, rendering those parts of the chart worthless).


    [DB] Please revisit MA's comment here to see both graphics in question.

    Off-topic snipped.

  31. LTO,

    I have already provided you  a peer reviewed reference to support the 1200 km measurement when I cited Hansen 2006 in post 67.  I provided you data from BEST showing their numbers cover 80% of the globe in 1934.  It is not my problem if you do not read the citations you are given.  You need to cite data to support your wild claims.

    Your argument is an argument from incredulity.  If you wish to stand on that basis you are welcome.

  32. Michael, it appears you missed my responses to the references you cited at #69 and #70. They did not support your argument. If you think I misunderstood something please clarify. You're the one making wild claims, in my view. The issue may be that you're looking at things from a big picture perspective. I'm trying to actually understand the detail, which is where the devil always is.  The temperature data is so incredibly derived, and yet never seems o be presented with error bars. This is an enormous issue when precision has massively increased in recent decades

  33. MA: Thanks for thr graphs at 77. They seem really weird on two counts, but hopefully you can explain what I'm missing:

    1. How is it possible that there are more blanked out areas in 1934 than 2016? The baseline is apparently being calculated as a 1900-1999 average, and so this implies there is baseline data to compare 2016 anomaly data against for which there was no actual data in 1934. But if there was no data in 1934, how could a baseline for the period 1900-1934 be calculated in the first place?

    2. In the 1934 chart, I've drawn a couple of circles of radius roughly 1000km. Could you tell me how many temperature measurements there were within them in your dataset in 1934?

    If I'm reading the 1934 chart correctly, it's asserting that 1934 was 0.1 C cooler than the 1900-2000 baseline, but with no data for roughly 15% of the globe. What is the error on this 0.1C figure?



  34. Link to chart for Q2 above here:

  35. LTO @83,

    (1) The usual advice on choice of anomaly period (often ignored) is to use the latest 3 complete decades as the anomaly base. This is because these data will be more accurate than that of earlier periods.

    The choice of the 1900-2000 period as an anomaly base was because I set the GISS mapping-engine to that base, there having been talk up-thread of 1934 being below the 29th century average. (This indeed appears to, be tha case.)

    To provide an anomaly when any data is missing requires some form of in-filling. With a 1900-2000 base period there will be a lot of in-filling as there is a lot of blank spaces for the early years of this period.

    (2) Concerning the number of temperature measurements within your circles:-LTO marked-up map

    The answer would be "lots." The data is from ICOADS who reckon to something like 25% ocean coverage in 1934. So there are plenty of gaps. (But do note it depends on wht you call a gap. The [almost] million 1934 measurements may provide 25% cover but they are but point-measurements in oceans which stretch 361 million sq km.) Yet the areas you circle would have been well covered having had pleanty of whaling ships traversing them in 1934.

    You also ask about confidence intervals for global temperature records in years-gone-by. Perhaps the quickest way to find such data is Berkeley Earth. They put the 1934 global anomaly in the range -0.120ºC to -0.246ºC (95%CI)  and 2016 +0.905ºC to +0.993ºC, this with a 1951-80 anomaly base. Subtract 0.066ºC for a 1900-2000 base.

  36. LTO - you seem to be trying to assert that globally,1934 might have hotter than now and doing that by inferring that coverage bias is higher than calculations from comparing spatially separated stations indicate.

    Would you still doubt that it is hotter now than 1934 if you looked at a temperature series that was constructed only from stations operating in 1934? (ie global temperature is being measured on the same basis).

    Infilling. How could you gain confidence that infilling was doing a good job in sparse stations? How about estimating the value of a station only from those around it and comparing with actual value? This is how the error estimates are built. Does it seem a reasonable assumption that the observed variability in SSTs from modern satellite era can be used as basis for estimating the underlying variability in SST in the past as well?

  37. Hi MA, thanks for the response.

    1. Could you please elaborate on what this "in-filling" is?

    2. With respect, 'lots' isn't a great answer. 25% coverage means no coverage for 75% of the ocean, and yet the 1934 chart shows anomaly readings for roughly 90% of the ocean. Are you saying that the vast majority of the ocean data in the chart is a result of this 'in-filling'. This isn't even getting into issues of reliability of data and calculation of anomalies when a given ship instrument probably never measured precisely the same place at the same time twice.

    3. Yes the Berkeley earth data seem to be one of the few sources that show their errors loud and proud, and as I understand it they use a wide selection of proxy data and acrual measurements to arrive at their estimates. I haven't disputed whether or not 1934 was a cold year globally, but I'm rather saying the chart at the start of this article is massively misleadingp and something of a work of fiction, pretending we know far more about thr global spatial temperature anomalies and with more certainty than we actually do. It raises alarm bells.

  38. You work on basis that "if you dont have thermometer there, you dont know anything", but as the methodology papers show, that is not the case. If this was so then couldnt reconstruct the temperature from just 60 stations, (massively less than 25% coverage). but as I pointed out earlier, you can.

    I wasnt aware the BEST used proxy data. Can you elaborate? I dont see it in a quick scan of methodology report.

    I found error estimates for other series in a few seconds of search. Which do you want to know?

    The chart is only "a work of fiction" if you discount all the work on estimates of error in global temperature series. The producers of data series publish papers on the methodology which discuss the issues and errors. For you to ignore these and declare the series as "fiction" means the burden of proof does move to you. They have peer-reviewed papers justifying their series. If want to cast doubt, then you need to show us what is wrong with their analysis.

  39. Hi Scadden

    In #86 you assert that I’m trying to assert something that I never have. It’s a weird thing to do. Please just focus on my actual criticisms rather than invent new ones on my behalf.

    In #87 you assert the basis I apparently work on, which is again incorrect.  You either don’t understand the “60 stations” analysis (it’s not actually 60, but that’s not really the point. I address this in #75), or you don’t understand what my criticism is. Read my actual criticisms rather than leaping to conclusions about what you imagine them to be.

    Re Berkeley earth, yes you’re right re proxy data. I meant to make the point that it is land only, and the discussion here is largely about the ocean anomaly ‘data’ shown in the 1934 chart at the top of This page.

  40. ITO,

    Reviewing your posts at 69 and 70 I see no links to any peer reviewed data (and no data analysis at all).  I see only a rambling account of your uninformed opinion.  You provide no statistical evidence (or mention of statistical calculations) to support your wild claims that the temperature record is not supported by enough data.

    We need to review the rules.  This is a scientific web site.  You must produce data to support your wild claims.  I have produced peer reviewed articles to support my claims.  You have provided only your opinion of those articles.  The burden is on you to provide data to support your claims.  If you do not like the articles I cite it is your responsibility to find articles that detail their mistakes or provide detailed calculations that they are incorrect.  Simply stating your unsupported opinion as a random guy on the internet who claimed at the start to not beingvery informed does not carry any weight. 

    Do your homework and read the literature.  If you want to know all the details they are published on the internet.  Find them yourself.  It is  not our responsibility to find all the obscure details that you are interested in.

    You have not demonstrated that you know what a temperature anomaly is.  That knowledge is crucial to our discusion.  As for your unsupported claim that GISS cannot estimate the anomaly over a radius of 1200 km BEST says:"the correlation structure [of the temperature field] is
    substantial out to a distance of ~1000 km, and non-trivial to ~2000 km from each site".  BEST uses a different averaging technique than GISS uses.  When two different groups using different mathmatical techniques agree that the data can be averaged over long distances you need to provide more than "I don't believe it".  Knowing what an anomaly is would help you understand.

    I suggest you restart your discussion and try to learn the basic teminology before you try to reanalyze the data.  You might benifit from reading some of the old posts at Open Mind although you will have to look way back in his blog to find basic discussion of the temperature record.  Realclimate also has good technical discussions of the temperature record in their library.

    You have demonstrated an extraordinary knowledge of denier references to the temperature record.  Please link the web site that you are using to inform yourself so that we can see the original argument.   It probably has already been debunked and we can simply link to that debunk.

  41. My apologies. I was really guessing and trying to understand what your beef was. Can perhaps we clarify that?

    I am assuming your beef is with Fig 1. I understand that you are claiming it to be work of fiction because you believe ocean temperature anomaly, particularly in southern hemisphere was unknowable in 1934. (I believe NASA would assert that this is best estimate available from available data and reanalysis methods but they would also acknowledge large error bars). If I have misunderstood you, then please be clearer.

    My point with the 60 station analysis, is that is demonstrates how well methods that perforce assume very large spatial correlation of anomalies work in practise.

    My primary source for how SST derived is Huang et al 2016 but you have chase down the older papers (particular Smith and Reynolds contributions) for errors and validation studies. I do not believe the error estimates for SST in 1900-1941, southern regions (as high as 0.9C) invalidate the figure 1 but maybe I have again misunderstood you.

  42. Hi Scadden

    Accepted! :) My beef is that the 1934 chart gives the impression we know things about the global distribution of temp anomalies that we simply don’t. I consider it to be extremely misleading, even if the point of this article (1934 was only hot in localised areas) is true. Generally, I‘m finding there is a real lack of acknowledgement and proper treatment of uncertainties. Unfounded certainty raises red flags, and is unscientific. 

    If you’ve got a solid background in experimental science you’ll know that the vast majority of scientists (particularly those in softer fields) don’t really understand statistics and treatment of errors. It’s not a trivial point, nor is validation of data. As we get more And more measurement data, it becomes less and less reasonable to rely on old analyses of validation and associated errors. As I understand it Berkeley Earth was set up precisely because of myriad errors in climate data analysis (even if the eventual conclusions on land temperature changes were in line with prior analysis).

    To illustrate what I mean, see the annotated image here. There are ‘features’ in areas where there was probably no data collected in 1934, meaning that what we’re looking at in the areas I’ve highlighted is a largely imaginary baseline (I.e. using infilled data modelling for much of the 20th century) overlaid with imaginary 1934 data. If you look at the areas I’ve highlighted in the 2016 data, you can see there are real features that you’d only identify from having actual measurements. No such measurements exist in 1934, and so the model just assumes everything is cold, and makes up some features to give the impression of being realistic. Do you see my point?

    Re the SST data, I address the Huang et al data set in #80 above - ERSST explicitly acknowledge on their website unreliability of data from the 1940s and earlier. I must say I’m a bit confused now about whether the chart is supposed to be sea surface temperatures or air temperatures. Obviously the two are linked, but they’re not the same thing.

  43. Hi Michael

    I don’t think you’re a serious part of the discussion here. My post at #69 specifically discusses data from Hansen’s 1987 paper, and overlays those figures against the 1934 chart at the top of this article to illustrate my point, which you haven’t addressed. Your asserting that that isn’t there makes you look like a fool. I’m relatively sure that isn’t actually the case, although the uncertainty in that belief is increasing rapidly. Your only purpose here appears to be to insult me, as opposed to explaining concepts or arguing points of view. I’m learning a huge amount about this topic and the myriad complexities from the discussions I’m having on this site, but not from you. Please either change your tone or stop harassing me.


    [JH] Antagonistic, ad-hominem snipped. In addition, you are now skating on the thin ice of excessive repetition which is also prohibited by the SkS Comments Policy

    Please note that posting comments here at SkS is a privilege, not a right.  This privilege can be rescinded if the posting individual treats adherence to the Comments Policy as optional, rather than the mandatory condition of participating in this online forum.

    Please take the time to review the policy and ensure future comments are in full compliance with it.  Thanks for your understanding and compliance in this matter.

  44. LTO @92/93,

    Perhaps the difficulty you have with your "beef" is that it has a great similarity to that more commonly persented by folk who are in denial about AGW and who argue robustly against the very evidence itself. Given that similarity, you do need to be on the very top of your game to differentiate yourself from such contrarians. So not grasping that GISS LOTI is a combination of surface air temperature (GHCN) surface sea temperature (ERSST) is a rather profound omission. I would also add that the lack of patience with those comments challenging you is also something of a contrarian characteristic.

    Picking up on your assertion that the ERSSTv5 info web page says that "explicitly acknowledge(s) on their website unreliability of data from the 1940s and earlier"; this is incorrect. The statement made is "Note that the data are more reliable after the 1940s" and this refers to both land and ocean data. What is 'reliable' depends on your purpose. Note the description following that statement. So if that is to compare the ERSST 2° × 2° grided data cell-by-cell (2° of latitude is 210km), then 'reliability' pre-1950 may be more of an issue for you.

    If you have specific concerns with 'uncertainty' regarding these data, it would be good if you set them out. And by that, I mean describing them properly rather than presenting accusative statements such as "the vast majority of scientists ... don’t really understand statistics and treatment of errors. It’s not a trivial point..." or suggesting there is "imaginary 1934 data ... No such measurements exist in 1934" being presented as real. If your "beef" were to rattle out levels of accuracy in temperature records, we wouldn't expect such language.

  45. LTO , my comment is a general one :-

    Granted that our knowledge of localized climatic conditions becomes increasingly imperfect as we look backward through the past century (and indeed, past millennium) . . . is there any substantive conclusion (from that truism) which you wish to present to the readers of this thread?

    You make the valid point that there should be better understanding of "statistics and treatment of errors".   But (so far) you have not drawn any substantive conclusion (from that point) regarding the science of modern climate.   Are you trying to show that the mainstream/consensus assessment of AGW is wrong in a major way?   And if so . . . how is it wrong?

    You yourself have indicated ( in #87 , though in a passive way ) that the contrarian scientists who undertook the BEST project did end up in confirming the mainstream science.   Which is also confirmed by the simple physical evidence of melting ice / rising sea level / oceanic acidification / species migration / etcetera.

    And as for "statistics and treament of errors" . . . none of us should forget that the methodology of statistics has the primary purpose of illuminating/assessing scientific "fact" (or more grandly, "truth").   I am sure you would like to agree that statistics ought not to be used to obfuscate, or to persuade the reader toward falsehoods & pseudo-science.   Of course.

  46. Hi MA

    You may notice that ‘beef’ was first used in #91, and not by me. 

    I’ve set out my concerns now a number of times. I asked you some questions in #87 you haven’t answered. If you don’t know the answers, it‘s fine to admit it. My purpose here isn’t to trick anyone, just to increase my understanding of the relative uncertainties. There are indeed many profound omissions in my knowledge on this topic, which is precisely why I’m here. I’ve never pretended otherwise. I’d add with certainty that there are many profound omissions in the knowledge of people who have made up their minds that climate change will destroy the world in 12 years, as well. 

    There’s an irony here that a site called sceptical science treats questions about extremely complicated data manipulation with dismissive rudeness. You haven’t done this - your answers have been exceptionally informative - but the same can’t be said of some of your colleagues. It isn’t my job to ‘distinguish’ myself from different people asking different questions, although frankly I’d have thought that the ‘asking different questions’ element would do this.

    Your claim that the comment about ERSST data before 1950 being unreliable (which I quoted in full in #80) is referring to both land and sea data isn’t apparent at all, and the page appears to be talking solely about The Extended Reconstructed Sea Surface Temperature (ERSST) dataset. It may be another profound omission of mine to think that this data set is just, you know, sea surface temperatures, but I suspect I’m on more solid footing here. I’m not sure what the relevance of the point is, either way, as it doesn’t move anything forward from my comments at #92

    As to the point about statistical error analysis, I suspect you’re being disingenuous. If not, read this and prepare to be amazed:


    [DB] Inflammatory and off-topic snipped.

  47. LTO @96 , 

    further to my points @95 , have you come to any substantive useful conclusions about AGW and mainstream climate science?

    Thus far, your posts have been remarkably discursive and deficient in critical thinking.

  48. Validation and errors have been re-examined a number of times, as the explicitly stated in the Huang 2016 method paper and easily followed in the bibliography. As a side note BEST was set up because of perceived errors in the temperature record. They tried a new methodology (always good) but ended up discovering that that existing methodology worked after all. Perception is not reality, but BEST credit, this is science and skepticism working as it should.

    What is peeving for me and other commentators is that your insistence "we know things about the global distribution of temp anomalies that we simply don’t" Uncertainty and large error bars are not the same as "dont know". Huang comments on pre-1940 data are more deeply discussed in the methods and earlier papers. 1940 is when change in measurement methodology occurred as well as issues with coverage. These all combine to uncertainties of up to 0.9C in lower southern hemisphere oceans. Double that if you like and the difference between 1934 and now is still pretty much as depicted by the figure.

    Sparcity of measurement does not translate into unknowable. You can use modern satellite data to show anomalies are spatially correlated over large distance and estimate errors in unsampled areas. You dont appear to have read how the anomaly baselines are actually calculated. Try Smith and Reynolds 2003 and note especially the discussion about high and low frequency distributions. And, yes, weather models are useful. Humidity measurements made down wind constrain what temperatures upwind can be etc etc. And, no you dont have to take it on faith. Reanalysis models do extensive validation otherwise noone would trust them.

  49. Remarkable that someone considered something in my post at #96 to be inflammatory and yet lets eclectic spew insult after insult cf #97.

    Scadden: you're persisting in refusing to engage on the points I'm actually making about the 1934 chart. You say that uncertainty and large error bars aren't thr same as don't know, but this depends on what the question is, and what you're trying to do with the data. The 1934 chart doesn't present any error bars, uncertainties or indications of statistical significance at all nor acknowledge the massive spread in uncertainties across different parts of the globe. The error is not homogeneous. This, as I've been repeatedly saying, is incredibly misleading as it leads people to think we know far more than we actually do.


    [JH] Moderation complaint snipped.

    Please note that posting comments here at SkS is a privilege, not a right.  This privilege can and will be rescinded if the posting individual continues to treat adherence to the Comments Policy as optional, rather than the mandatory condition of participating in this online forum.

    Moderating this site is a tiresome chore, particularly when commentators repeatedly submit offensive or off-topic posts. We really appreciate people's cooperation in abiding by the Comments Policy, which is largely responsible for the quality of this site. 
    Finally, please understand that moderation policies are not open for discussion.  If you find yourself incapable of abiding by these common set of rules that everyone else observes, then a change of venues is in the offing.

    Please take the time to review the policy and ensure future comments are in full compliance with it.  Thanks for your understanding and compliance in this matter.



  50. LTO @96.

    Firstly, my apologies for the "... and this refers to both land and ocean data" within my comment @94. This was part of a rather-too-long passage that was cut, and should have been cut in its entirety. (This cut passage concerned the "reliability" comment on the ERSST info web-page being part of the caption for a missing graphic.)

    Secondly, looking at your many comments down this thread, you "beef" is that you consider it inappropriate to in-fill temperatures on an anomaly map. You term this in-fill for the 1934 map in the OP (actually a 5-year period centred on 1934) "exceptionally misleading," "massively misleading and something of a work of fiction," "fanciful" and suggest it is unjustifiable. You appear to expect coverage to be at least ten stations per ~200km square. And despite the long-long interchange since #59 up-thread, I don't see any movement from that position.

    You appear to want folk here to furnish you with station-number/measurement-numbers for some remote part of the world at some time in the past; this information to demonstrate that you are justified in branding the anomay map in the OP "exceptionally misleading." But such a conclusion ignores the ability of folk to produce such anomaly maps that remain remarkably accurate with far fewer data. I think you are less after understanding (which you have said is your purpose) and more are looking for a research assistant to furnish you with ammunition for your erroneous "beef."

    Thirdly, I am aware that you asked a question @87. Demanding an answer isn't a good way of getting an answer. Besides, I am not entirely sure an answer would be helpful to your understanding.

    So perhaps there is need to restructure your 'enquiry' here. So far, it is producing a lot of words and with very little coherence.

Prev  1  2  3  Next

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

Link to this page

The Consensus Project Website


(free to republish)

© Copyright 2021 John Cook
Home | Links | Translations | About Us | Privacy | Contact Us