Page 2 of 9 FirstFirst 1234 ... LastLast
Results 11 to 20 of 83

Thread: 2017's global temperatures

  1. #11
    tWebber TheLurch's Avatar
    Join Date
    Jan 2014
    Location
    Northeast USA
    Faith
    MYOB
    Gender
    Male
    Posts
    1,001
    Amen (Given)
    66
    Amen (Received)
    323
    Quote Originally Posted by Teallaura View Post
    Sparky quoted a section stating they were omitting a part of the data from the analysis - not that they were running a separate statistical analysis (which wouldn't mention an omission of extraneous data).
    Well, the part he quoted was:

    Another way to analyze this is to remove the effect of the El Niño. We have enough historic examples of how El Niño/La Niña strength correlates with global temperatures, and can use these to subtract the effect from the temperature data. That's been done with NASA/NOAA, and the results show that, when this adjustment is made, 2017 is actually the warmest year on record.
    I'm not sure where that implies omission of data.

  2. #12
    tWebber TheLurch's Avatar
    Join Date
    Jan 2014
    Location
    Northeast USA
    Faith
    MYOB
    Gender
    Male
    Posts
    1,001
    Amen (Given)
    66
    Amen (Received)
    323
    Quote Originally Posted by oxmixmudd View Post
    The raw data is adjusted. TOB, heat island effect, wild point editing, and so on. So it is not fair to say the results come solely from 'unaltered data'. From my studies, those adjustments make sense, are valI'd and even necessay to get an accurate result. And further, they don't change the general trend. That is, the averaged raw unadjusted data shows basically the same result. But those that have their doubts will see deception in the way you characterise the results relationship to the raw data.
    Fair enough, poor phrasing on my part; "they didn't alter the data to perform this specific analysis" would have been more appropriate.

    And yes, the underlying data would not produce anything meaningful in its original form, given that it's taken at different times of day, from stations that have moved over the years, etc. etc. NOAA and NASA use similar approaches to compensating for that, while Berkeley Earth specifically set out to use a completely different way of handling it - they ended up producing nearly indistinguishable results (I can get into the technical details if anyone's interested).

    There have also been a few looks at specific issues like the potential for urbanization to have influenced the overall temperatures, and those have ended up validating NOAA and NASA's approach, too.

  3. #13
    tWebber
    Join Date
    Jan 2014
    Location
    Maryland
    Faith
    Christian
    Gender
    Male
    Posts
    1,867
    Amen (Given)
    94
    Amen (Received)
    797
    Quote Originally Posted by TheLurch View Post
    Fair enough, poor phrasing on my part; "they didn't alter the data to perform this specific analysis" would have been more appropriate.

    And yes, the underlying data would not produce anything meaningful in its original form, given that it's taken at different times of day, from stations that have moved over the years, etc. etc. NOAA and NASA use similar approaches to compensating for that, while Berkeley Earth specifically set out to use a completely different way of handling it - they ended up producing nearly indistinguishable results (I can get into the technical details if anyone's interested).

    There have also been a few looks at specific issues like the potential for urbanization to have influenced the overall temperatures, and those have ended up validating NOAA and NASA's approach, too.
    I remember reading one study (I think I still have it on my laptop) where they took the data from the elite sites for the US (USCRN) and compared the unprocessed output from those stations during the years where that data is available with the same homogonized data for the US. The result (in terms of the graph and temperature trend over time) is indistinguishable except perhaps for a small constant delta in the absolute temperature value.

    I've done personal avaluations of the accuracy by averaging sets of stations and discovered through that the (known to climatologists) drop in temperature across the eastern US around the late 1950's due to a shift in certain Atlantic (air/water? can't remember) circulation patterns. The 'signal' for that can be found in the (averaged) data as far west as eastern Texas.

    Jim
    Jorge's trueorigins paper: "...it is known that other volcanic features match what is usually associated with impact craters including ... shatter cones and crystal deformations"

    Planetary Science Institute: "Shatter cones … are found in only two places on Earth, 1) in nuclear test sites and 2) meteorite impact structures. They are formed as a result of the high pressure, high velocity shock wave ...

    maximum pressures from 45 to 200 times greater than found in volcanic events (2->20 Gpa)

  4. #14
    tWebber TheLurch's Avatar
    Join Date
    Jan 2014
    Location
    Northeast USA
    Faith
    MYOB
    Gender
    Male
    Posts
    1,001
    Amen (Given)
    66
    Amen (Received)
    323
    Quote Originally Posted by oxmixmudd View Post
    I remember reading one study (I think I still have it on my laptop) where they took the data from the elite sites for the US (USCRN) and compared the unprocessed output from those stations during the years where that data is available with the same homogonized data for the US. The result (in terms of the graph and temperature trend over time) is indistinguishable except perhaps for a small constant delta in the absolute temperature value.
    This is almost certainly the paper you're thinking of:
    https://www1.ncdc.noaa.gov/pub/data/...e-etal2010.pdf

    My favorite analysis of this sort was a study that looked at the influences of urbanization by figuring out which areas of the globe had urbanized over the last 40 years by comparing the amount of light in night-time satellite images. They then identified all the temperature readings taken in these areas and looked for a trend. Just seemed like a very elegant approach to the problem.

    Quote Originally Posted by oxmixmudd View Post
    I've done personal avaluations of the accuracy by averaging sets of stations and discovered through that the (known to climatologists) drop in temperature across the eastern US around the late 1950's due to a shift in certain Atlantic (air/water? can't remember) circulation patterns. The 'signal' for that can be found in the (averaged) data as far west as eastern Texas.
    Neat. It's great that climatology is one area of science where most of the data is public, along with much of the computer code that analyzes it. There's even a couple of simplified climate models that researchers have adapted to run on personal computers, so they can be used for educational purposes.

  5. Amen shunyadragon amen'd this post.
  6. #15
    tWebber Teallaura's Avatar
    Join Date
    Jan 2014
    Location
    In my house.
    Faith
    Christian
    Gender
    Female
    Posts
    12,194
    Amen (Given)
    5520
    Amen (Received)
    4508
    Quote Originally Posted by TheLurch View Post
    Well, the part he quoted was
    :Another way to analyze this is to remove the effect of the El Niño. We have enough historic examples of how El Niño/La Niña strength correlates with global temperatures, and can use these to subtract the effect from the temperature data. That's been done with NASA/NOAA, and the results show that, when this adjustment is made, 2017 is actually the warmest year on record.

    I'm not sure where that implies omission of data.
    Probably the bit I bolded. 'Subtract' implies omission or removal here to my mind.

    I admit, the need to correct for whatever always makes me skittish about trusting the conclusion drawn - even though I understand enough to know it may be a necessary evil and perfectly correct. Of course, climate isn't as persnickety as people are but weighting is so easy to mess up.

  7. Amen Cerebrum123 amen'd this post.
  8. #16
    tWebber shunyadragon's Avatar
    Join Date
    Jan 2014
    Location
    Hillsborough, NC
    Faith
    Agnostic
    Gender
    Male
    Posts
    12,530
    Amen (Given)
    1215
    Amen (Received)
    830
    Quote Originally Posted by Teallaura View Post
    Probably the bit I bolded. 'Subtract' implies omission or removal here to my mind.
    Reading it fully in context would assure one that 'subtract' does not mean removal of data.
    Glendower: I can call spirits from the vasty deep.
    Hotspur: Why, so can I, or so can any man;
    But will they come when you do call for them? Shakespeare’s Henry IV, Part 1, Act III:

    go with the flow the river knows . . .

    Frank

    I do not know, therefore everything is in pencil.

  9. #17
    tWebber Teallaura's Avatar
    Join Date
    Jan 2014
    Location
    In my house.
    Faith
    Christian
    Gender
    Female
    Posts
    12,194
    Amen (Given)
    5520
    Amen (Received)
    4508
    Quote Originally Posted by shunyadragon View Post
    Reading it fully in context would assure one that 'subtract' does not mean removal of data.
    Being less of a jerk would insure people taking you more seriously - but we can't have everything.

    Actually, omission is better than weighting - you can make a better case for 'we need to leave X out of this for Y reason' than ''okay, we're gonna assume X = Y for <insert goobledy gook here>" - in terms of explaining it to those without a statistical background.

  10. #18
    tWebber
    Join Date
    Jan 2014
    Location
    Maryland
    Faith
    Christian
    Gender
    Male
    Posts
    1,867
    Amen (Given)
    94
    Amen (Received)
    797
    Quote Originally Posted by Teallaura View Post
    Being less of a jerk would insure people taking you more seriously - but we can't have everything.

    Actually, omission is better than weighting - you can make a better case for 'we need to leave X out of this for Y reason' than ''okay, we're gonna assume X = Y for <insert goobledy gook here>" - in terms of explaining it to those without a statistical background.
    The difficulty in explaining why certain adjustments are made is due primarily to three factors:

    1) politically motiviate mistrust
    2) scientific ignorance of the audience
    3) malicious players in the mix who routinely misrepresent or lie about the adjustments being used*.

    #3 includes those hampered by 1 and 2 but who are unaware of that fact.

    I mentioned several types of adjustment in my previous email. But I think more important than that is the study I mentioned in terms of dealing with two. The USCRN is a set of weather stations that are set up far from any manmade influence and in absolutely pristine conditions in terms of the equipment and methods used to gather data from them. That data, unmodified except perhaps for the most basic of wild point editing (e.g. thermometer breaks on day x and gives crazy reading, is replaced and funtional on day x+1. data point from day x must be either removed or normalized), when simply averaged over time, produces an almost identical temperature trend to the data using all those 'questionable stations' and 'questionable modificiations'. The implication then at the most simple level then is those modifications would appear to be either correct of of no consequence whether one understands them or not.

    The next step then in understanding what is going on is to take a look at some of the factors that can give a bad result and require action. One of the less obvious is something called "Time of Observation Bias" (TOB or ToOB). This is a bias that results from taking temperature readings at different times of the day across the various stations. TOB adjustment then (mostly) removes the bias by determining the known, measured bias associated with a station's data reading time(s) against the current established norm for data reading time (s).

    As for whether or not it makes sense to try remove a known bias (can we expect such bias removal efforts to be effective, accurate).

    If possible, listen to a record (vinyl) through equipment not employing the RIAA equalization curve established for record and playback in that media. It will sound pretty bad. Very tinny, and with lots more pops and hisses than you are used to. That is because when the biases are known in a system, equalization or bias removal works very, very well to restore the original signal. In this case, applying the equalizations curve, which is designed to take into account the acoustic energy distrubution of music and muscial instruments, allows your record (vinyl) to produce a relatively quiet and clear audio playback, significantly superior to what the media itself can do unequalized.

    And that is where the trust issue comes into play, and I believe it is the primary factor. Unless one takes a lot of time, and has a non-trivial understanding of the systems used and climate in general, it becomes virtually impossible to explain why the raw data is not used without modifiation**. And with lots of folks out there claiming the modifications are not scientifically justified and designed to favor certain monetary and political interests, the waters become significantly muddied, and a lot of people get led down the primrose path thinking the truth is X when in fact the truth is Y.

    Jim

    *this happens mostly on the political/anti AGW side of the argument, but to jack up the mistrust issue, there have been notable attempts to bias the data on the science/ pro-AGW side as well

    ** consider your <insert gobledygook here> characterization.
    Last edited by oxmixmudd; 01-22-2018 at 01:54 PM.
    Jorge's trueorigins paper: "...it is known that other volcanic features match what is usually associated with impact craters including ... shatter cones and crystal deformations"

    Planetary Science Institute: "Shatter cones … are found in only two places on Earth, 1) in nuclear test sites and 2) meteorite impact structures. They are formed as a result of the high pressure, high velocity shock wave ...

    maximum pressures from 45 to 200 times greater than found in volcanic events (2->20 Gpa)

  11. #19
    tWebber TheLurch's Avatar
    Join Date
    Jan 2014
    Location
    Northeast USA
    Faith
    MYOB
    Gender
    Male
    Posts
    1,001
    Amen (Given)
    66
    Amen (Received)
    323
    I think the other thing here is that, on most complicated topics, there is no one-size-fits-all explanation that will work for everyone. I phrased things in a way that made sense for me; Teal, because of her background, would find different explanation more compelling, etc. It's important to have a diversity of explanations, even if they all distill down to the same thing, because you never know what'll click with someone.

  12. Amen Cerebrum123, Teallaura amen'd this post.
  13. #20
    tWebber Teallaura's Avatar
    Join Date
    Jan 2014
    Location
    In my house.
    Faith
    Christian
    Gender
    Female
    Posts
    12,194
    Amen (Given)
    5520
    Amen (Received)
    4508
    Quote Originally Posted by oxmixmudd View Post
    The difficulty in explaining why certain adjustments are made is due primarily to three factors:

    1) politically motiviate mistrust
    2) scientific ignorance of the audience
    3) malicious players in the mix who routinely misrepresent or lie about the adjustments being used*.

    #3 includes those hampered by 1 and 2 but who are unaware of that fact.

    I mentioned several types of adjustment in my previous email. But I think more important than that is the study I mentioned in terms of dealing with two. The USCRN is a set of weather stations that are set up far from any manmade influence and in absolutely pristine conditions in terms of the equipment and methods used to gather data from them. That data, unmodified except perhaps for the most basic of wild point editing (e.g. thermometer breaks on day x and gives crazy reading, is replaced and funtional on day x+1. data point from day x must be either removed or normalized), when simply averaged over time, produces an almost identical temperature trend to the data using all those 'questionable stations' and 'questionable modificiations'. The implication then at the most simple level then is those modifications would appear to be either correct of of no consequence whether one understands them or not.

    The next step then in understanding what is going on is to take a look at some of the factors that can give a bad result and require action. One of the less obvious is something called "Time of Observation Bias" (TOB or ToOB). This is a bias that results from taking temperature readings at different times of the day across the various stations. TOB adjustment then (mostly) removes the bias by determining the known, measured bias associated with a station's data reading time(s) against the current established norm for data reading time (s).

    As for whether or not it makes sense to try remove a known bias (can we expect such bias removal efforts to be effective, accurate).

    If possible, listen to a record (vinyl) through equipment not employing the RIAA equalization curve established for record and playback in that media. It will sound pretty bad. Very tinny, and with lots more pops and hisses than you are used to. That is because when the biases are known in a system, equalization or bias removal works very, very well to restore the original signal. In this case, applying the equalizations curve, which is designed to take into account the acoustic energy distrubution of music and muscial instruments, allows your record (vinyl) to produce a relatively quiet and clear audio playback, significantly superior to what the media itself can do unequalized.

    And that is where the trust issue comes into play, and I believe it is the primary factor. Unless one takes a lot of time, and has a non-trivial understanding of the systems used and climate in general, it becomes virtually impossible to explain why the raw data is not used without modifiation**. And with lots of folks out there claiming the modifications are not scientifically justified and designed to favor certain monetary and political interests, the waters become significantly muddied, and a lot of people get led down the primrose path thinking the truth is X when in fact the truth is Y.

    Jim

    *this happens mostly on the political/anti AGW side of the argument, but to jack up the mistrust issue, there have been notable attempts to bias the data on the science/ pro-AGW side as well

    ** consider your <insert gobledygook here> characterization.
    You do realize I have no issue with the idea of having to adjust to account for a bias (or skew in my terminology). But in this case, and in the political arena, I think you're (general) better off assuming your (personal - post above) Number 2 rather than Number 3. You answer the why portion yourself in your footnotes.

  14. Amen Cerebrum123 amen'd this post.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •