1321: "Cold"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Sat Jun 21, 2014 5:16 pm UTC

SFX wrote:Indeed, in fact the "couple pof days that dipped below zero" were in reality some of the coldest weather the city of St Louis has ever experienced, certainly the coldest most people who read xkcd ever experienced.
This I will grant you, simply because xkcd readership is disproportionately young and if you're talking about daytime temperatures, and consider that young children are often kept from "experiencing" especially cold weather, 1989 is probably before most of them can remember cold weather.

Hell, it was cold by Alaskan standards.
The people of Fairbanks, with an average of 16 days a year that never get above -15F, would have to disagree. (It's not really fair to only consider cities at the southern extreme of Alaska, which have always had much more mild winters even than much of the upper Midwest.)
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Sat Jun 21, 2014 6:18 pm UTC

gmalivuk wrote:
SFX wrote:Indeed, in fact the "couple of days that dipped below zero" were in reality some of the coldest weather the city of St Louis has ever experienced, certainly the coldest most people who read xkcd ever experienced.
This I will grant you, simply because xkcd readership is disproportionately young and if you're talking about daytime temperatures, and consider that young children are often kept from "experiencing" especially cold weather, 1989 is probably before most of them can remember cold weather.
It's not "me" that is the issue, it's the facts of the weather, and climate, that are so hotly disputed. I'm just pointing out the reality, which is far different than the glib explanation and handwaving away of the record cold, we see in the comic.
gmalivuk wrote:
Hell, it was cold by Alaskan standards.
The people of Fairbanks, with an average of 16 days a year that never get above -15F, would have to disagree.
It's the wind chill that really matters. The arctic and Alaska rarely have wind after it snows, so even -25 in Fairbanks doesn't feel as cold as -7 with the wind blowing at 30 mph a 8:51 AM in St Louis. Which feels like -32.7 °F, except it's actually far more hideous weather than still air at -25. Yeah, I am actually that guy mentioned in the alt text.

gmalivuk wrote:Which data do you consider to be the St. Louis climate data?
Now there is a really good question. What data actually shows the climate of St Louis, stretching back in time, to that first office at 210 Olive Street? Where are the concurrent observations that started in July 1929, when a weather station was opened at Lambert Field, 15 miles northwest of the downtown office? Where are the changes noted in the St Louis records, when in July 1958, the airport became the only official weather station in St Louis? Were there any obvious changes when in June 1996 the airport, now greatly expanded, began using the new automated only weather data?

This one.
Image

It's in the middle of a huge, very busy jetport.

Image

A hundred yards from a busy runway. Which is de-iced, and constantly cleared of snow.

Image

That's the official climate station for St Louis. Which shows, even from the weather yesterday I linked to. While the green areas were 85, the official high was 92 for St Louis. The average temperature (because they use the airport) was four degrees higher for the entire area, Even when it wasn't anywhere close to that, except at the airport, or other industrial areas. They all show 5 to 1o degrees higher maximums than the green areas do. What a surprise.

The actual carefully selected climate stations show a different story of course. But that set only starts in 2005.

Image

gmalivuk wrote:If you're not being intentionally misleading, maybe link to your sources ever?
Speaking of, the source for every image is actually linked to each image. Every image I've used has the source url attached to it. I was going to mention this, since it was obvious you didn't know that. But after I read your request, I saw this nonsense.
Izawwlgood wrote:Again, this is incredibly dishonest and demonstrative of cherry picking.
gmalivuk wrote:Wow, yeah, that is pretty damn dishonest of SFX.
But even so I was going to tell you, because it would make my life easier, but then you did it again.
gmalivuk wrote:Your repeated inability to answer this question makes me suspicious of whatever you will come up with, now that you've had ample time to cherry pick whatever sourse most strongly supports your position that winters aren't getting warmer.
Now take that suspicion, and realize that's how I feel about climatecentral, who did not list the source of their graphs, much less show the entire picture. They cherry pick the time series. They are known for being deceptive as hell about it.
gmalivuk wrote:Could that possibly be because the image supports your argument less if you go back farther?
Oh gosh, that's exactly what I've been pointing out about climatecentral, and their deceptive graphics! It's almost like I was making a point about time periods! But nah, that couldn't be it.
gmalivuk wrote:Yes, when you disregard the cooler 1970s and 80s and start in the abnormally warm 1990s, you do indeed see the trends you want. I've pointed this out to you repeatedly, and yet you still just keep on cherry picking every single graph you post.
Now here's where it gets so interesting. I say, and the data shows it clearly, that in the last few decades, the St Louis winters are trending colder. I say using the seventies as a start point is deceptive. If the winters are getting colder in the last decade, or decades, then using 44 years for the trend is bullshit. especially when that is starting at the coldest period ever. That is a prime example of cherry picking, and being deceptive. Show the entire period, and show the recent trend. People can think as they will, but at least they have the data, not some bullcrap climatecentral slanted view.
Izawwlgood wrote:This is an honest question.
No, it was not. You made that obvious.
Izawwlgood wrote:Your data source does not actually show this.
And there is why arguing on the internet is a joke. Try to laugh bout it.
speising wrote:what those graphs show is that a linear trendline is simply not a useful indicator, when slight variations in period produce wildly different results.
Yes, but that doesn't mean trends don't matter. It's exactly why you have to look at the big picture, if you want to actually know something about it.

Not some tiny little view climatecentral wants to to see. It's why they didn't just provide an actual source (like I did) of where they got the data from.

User avatar
Crissa
Posts: 294
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1321: "Cold"

Postby Crissa » Sat Jun 21, 2014 6:29 pm UTC

As if no one had thought of that!

Oh, wait, they did.

Heat island effect on climate measurements.

Another short answers to stupidly long questions.

-Crissa
Last edited by Crissa on Sun Jun 22, 2014 7:28 am UTC, edited 1 time in total.

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Sat Jun 21, 2014 6:44 pm UTC

schapel wrote: It's a fool's errand to argue about it. Even if you show a clear warming trend, someone will just post one photo of a weather station next to the heat output of an air conditioner and claim the data is just showing the urban heat island effect.
I tend to agree, and it goes both ways.

I show that the Columbia data shows a more realistic picture, because the weather station isn't in the middle of a huge city, next to a runway at a major airport. Using the NOW data it's very obvious that the St Louis data is not good enough for climate data.

And it will somehow be brushed aside, like it means nothing.

You show the weather station right in the middle of a modern airport, and it just doesn't mean anything to somebody who has already decided it must be warming there.

Then you show, even with all that, the winters (and summers) show no warming for decades. Cooling actually, and that gets brushed aside, ignored, or worse.

What's the point? facts and data aren't going to sway anyone. If that were the case, there would be no arguments about the weather/climate. We would all be one happy bunch of people holding hands and singing songs.

I mean, once again the lush green Rolling Hills are showing 84.2, but the industrialized stations are showing 92.3

Right at this moment in Missouri. But the airport in St Louis will be the final word. The climate record.

Haha you have to laugh.

Image

(source is listed on the picture)

Of course they tell you the data is adjusted to avoid bias, so as to not make it appear things are hotter, just because of growing cities, and stations moving locations, and new instruments. But when you look into it, that isn't the case at all most of the time.

You can spend a year watching the actual data, each day, and see for yourself. It's one reason I am a skeptic. I like to see for myself.

And speaking of, since I find climatcentral not providing a link to the source reprehensible and rude, here's how to find the source for any image I use

Just remove the .png from the image url, you will go to a page that includes the source, and sometimes more
information about the image.

For example

If I use an image, say to show what the St Louis weather station looked like in 1875

Image
http://i.imgur.com/IfbtqqL.png

Just remove the .png to get
http://i.imgur.com/IfbtqqL

which will show you the source of each picture I have used. (unless I screwed up, in which case let me know and I will fix it)

Yeah, the source was always there. Using an image with out a source is deceptive and cruel.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Sat Jun 21, 2014 7:32 pm UTC

SFX wrote:The actual carefully selected climate stations show a different story of course. But that set only starts in 2005.
So the different "story" they tell is not yet long enough to be considered a story. I'm not sure why you bothered even posting the image.

gmalivuk wrote:If you're not being intentionally misleading, maybe link to your sources ever?
Speaking of, the source for every image is actually linked to each image. Every image I've used has the source url attached to it. I was going to mention this, since it was obvious you didn't know that. But after I read your request, I saw this nonsense.
Izawwlgood wrote:Again, this is incredibly dishonest and demonstrative of cherry picking.
gmalivuk wrote:Wow, yeah, that is pretty damn dishonest of SFX.
But even so I was going to tell you, because it would make my life easier, but then you did it again.
gmalivuk wrote:Your repeated inability to answer this question makes me suspicious of whatever you will come up with, now that you've had ample time to cherry pick whatever sourse most strongly supports your position that winters aren't getting warmer.
Now take that suspicion, and realize that's how I feel about climatecentral, who did not list the source of their graphs, much less show the entire picture. They cherry pick the time series. They are known for being deceptive as hell about it.
So... what? You declined to inform us that you'd already sourced your images out of spite? You didn't like our tone so you decided to make yourself look more dishonest than you were already being accused of? How does that make sense?

gmalivuk wrote:Could that possibly be because the image supports your argument less if you go back farther?
Oh gosh, that's exactly what I've been pointing out about climatecentral, and their deceptive graphics! It's almost like I was making a point about time periods! But nah, that couldn't be it.
Look, you can't blame us for not understanding what badly communicated point you were trying to make. Complaining that climatecentral didn't use a sufficiently long timescale by using even shorter timescales yourself doesn't make you look like an intellectually honest participant in this discussion.

gmalivuk wrote:Yes, when you disregard the cooler 1970s and 80s and start in the abnormally warm 1990s, you do indeed see the trends you want. I've pointed this out to you repeatedly, and yet you still just keep on cherry picking every single graph you post.
Now here's where it gets so interesting. I say, and the data shows it clearly, that in the last few decades, the St Louis winters are trending colder. I say using the seventies as a start point is deceptive. If the winters are getting colder in the last decade, or decades, then using 44 years for the trend is bullshit. especially when that is starting at the coldest period ever. That is a prime example of cherry picking, and being deceptive. Show the entire period, and show the recent trend. People can think as they will, but at least they have the data, not some bullcrap climatecentral slanted view.
But you haven't been doing that either. You're just criticizing their data with equally shitty pictures of your own.

SFX wrote:Now here's where it gets so interesting. I say, and the data shows it clearly, that in the last few decades, the St Louis winters are trending colder. I say using the seventies as a start point is deceptive. If the winters are getting colder in the last decade, or decades, then using 44 years for the trend is bullshit. especially when that is starting at the coldest period ever. That is a prime example of cherry picking, and being deceptive. Show the entire period, and show the recent trend. People can think as they will, but at least they have the data, not some bullcrap climatecentral slanted view.
SFX wrote:I show that the Columbia data shows a more realistic picture, because the weather station isn't in the middle of a huge city, next to a runway at a major airport. Using the NOW data it's very obvious that the St Louis data is not good enough for climate data.

And it will somehow be brushed aside, like it means nothing.
Okay, fine. Let's consider Columbia, MO, and let's consider the whole record instead of just starting with the cold 1970s.

For every single winter before 1998, the linear trend in the number of days with lows strictly below 0F, from that winter to the one that ended this year, is negative. (As before, slopes below are per-century.)

Code: Select all

1890   04   -01.45
1891   01   -01.48
1892   04   -01.64
1893   07   -01.67
1894   03   -01.59
1895   11   -01.67
1896   01   -01.41
1897   04   -01.57
1898   00   -01.60
1899   13   -01.82
1900   05   -01.45
1901   00   -01.44
1902   14   -01.66
1903   05   -01.22
1904   06   -01.19
1905   17   -01.11
1906   01   -00.46
1907   02   -00.61
1908   00   -00.72
1909   04   -00.94
1910   08   -00.95
1911   02   -00.73
1912   09   -00.85
1913   00   -00.56
1914   01   -00.80
1915   05   -00.98
1916   03   -00.93
1917   03   -01.00
1918   15   -01.07
1919   03   -00.36
1920   03   -00.41
1921   01   -00.47
1922   01   -00.66
1923   00   -00.87
1924   07   -01.16
1925   07   -00.95
1926   01   -00.73
1927   02   -00.96
1928   04   -01.12
1929   02   -01.13
1930   09   -01.31
1931   00   -00.90
1932   00   -01.25
1933   05   -01.62
1934   02   -01.56
1935   02   -01.78
1936   19   -02.01
1937   01   -00.56
1938   00   -00.84
1939   00   -01.25
1940   11   -01.69
1941   00   -00.94
1942   05   -01.38
1943   03   -01.28
1944   04   -01.40
1945   00   -01.41
1946   02   -01.93
1947   02   -02.23
1948   06   -02.55
1949   04   -02.34
1950   00   -02.40
1951   07   -03.05
1952   02   -02.69
1953   00   -03.08
1954   01   -03.84
1955   04   -04.50
1956   00   -04.69
1957   01   -05.62
1958   02   -06.45
1959   03   -07.17
1960   03   -07.75
1961   03   -08.38
1962   07   -09.07
1963   12   -08.92
1964   09   -07.55
1965   04   -06.72
1966   03   -07.05
1967   00   -07.67
1968   04   -09.18
1969   02   -09.71
1970   07   -10.90
1971   01   -10.63
1972   03   -12.28
1973   02   -13.43
1974   06   -15.09
1975   02   -15.40
1976   02   -17.33
1977   16   -19.50
1978   07   -15.65
1979   12   -15.33
1980   03   -12.35
1981   04   -13.57
1982   08   -14.37
1983   00   -12.83
1984   15   -16.05
1985   09   -09.50
1986   02   -05.86
1987   03   -06.90
1988   04   -07.26
1989   05   -06.74
1990   07   -05.00
1991   01   -00.57
1992   00   -01.98
1993   02   -05.08
1994   03   -06.23
1995   01   -06.09
1996   07   -09.30
1997   07   -01.75
1998   01   08.82
1999   03   08.68
2000   00   13.93
2001   04   11.43
2002   01   22.53
2003   02   25.52
2004   01   34.55
2005   00   41.21
2006   00   41.67
2007   02   38.10
2008   00   53.57
2009   02   40.00
2010   02   50.00
2011   05   60.00
2012   00   350.00
2013   00   700.00
2014   07   

So unless you're basing your "clear cooling trend" argument on only a decade and a half of data, it remains the case that winters on average are getting warmer in the Saint Louis area. If climatecentral had also used the Columbia data and started in 1980, the trend would have looked even stronger.

(And if we dropped last winter and count only to the one that ended in 2013, then only the years 2002-2006 show what looks like a cooling trend based on this data. Every other year in the entire record shows warming.)

Then you show, even with all that, the winters (and summers) show no warming for decades. Cooling actually, and that gets brushed aside, ignored, or worse.
Your arguments get brushed aside because you have failed to show any such cooling as you claim. As I just showed, you might be able to make such an argument about one decade, or maybe two depending on what data you're looking at, but if you extend farther back than that the trends you think you're showing disappear. At least right up until you decide to start your own trends just before the heat spikes in the 1930s, doing exactly the same kind of cherry picking you complain about from climatecentral. (Not that data back to the 1930s even makes any sense when your own claim is a cooling trend in recent decades.)

Yeah, the source was always there. Using an image with out a source is deceptive and cruel.
If you had said that at the beginning, instead of intentionally keeping quiet out of spite or whatever, people might have been less inclined to dismiss you as just another boring climate change denier.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Sat Jun 21, 2014 8:19 pm UTC

gmalivuk wrote:So unless you're basing your "clear cooling trend" argument on only a decade and a half of data, it remains the case that winters on average are getting warmer in the Saint Louis area.
I think the twenty year trend line shows up clearly in the graphs. So does a 25 year trend line.

Image

Yep, seems clear enough. Since the warm nineties (which strangely enough had some really cold nights), Missouri winters are trending colder. Not in the long term, just the regular climate cycles that show up clearly looking at the longterm data. Which I showed you.

Image

Yep, there it is.

Image

Fits with the long term graph, which clearly shows a natural cycle of warm periods, and cold periods, so it's not even a surprise to see them turning cold again.

gmalivuk wrote:If climatecentral had also used the Columbia data and started in 1980, the trend would have looked even stronger.
You keep calling their cherry picked method "a trend", which it is not. You could pick -5, or 10 degrees to create a graph, and see something different. It's a fraudulent method of avoiding what the data actually shows. We do not use that for climate data, for a reason.

gmalivuk wrote:If you had said that at the beginning, instead of intentionally keeping quiet out of spite or whatever, people might have been less inclined to dismiss you as just another boring climate change denier.
It was more like I couldn't believe nobody here knows how imgur works. Don't confuse my lack of caring with spite. To be spiteful you have to actually care. I find it humorous in the extreme when people get all personal and angry over the weather.

I mean, how absurd can you get? It's talking about the weather.

gmalivuk wrote:
SFX wrote:The actual carefully selected climate stations show a different story of course. But that set only starts in 2005.
So the different "story" they tell is not yet long enough to be considered a story. I'm not sure why you bothered even posting the image.
You misunderstand. The official "we really checked this time and were careful" climate data is only from 2005.

Certainly many studies have compared long term rural stations and long term temperature records against the NCDC and found the problem obvious.

Image

And of course, like anything even remotely connected to climate, that will be fought over tooth and nail. I mean, it's so important some people will actually want to kill you over the weather. or the weather records.

Fascinating really. How did that happen?

User avatar
Izawwlgood
WINNING
Posts: 18686
Joined: Mon Nov 19, 2007 3:55 pm UTC
Location: There may be lovelier lovelies...

Re: 1321: "Cold"

Postby Izawwlgood » Sat Jun 21, 2014 8:36 pm UTC

SFX wrote:And of course, like anything even remotely connected to climate, that will be fought over tooth and nail. I mean, it's so important some people will actually want to kill you over the weather. or the weather records.
You're definitely right about that point.

SFX wrote:I think the twenty year trend line shows up clearly in the graphs. So does a 25 year trend line.
Again, why do you insist on selecting spans of time like that? Why do you continue to ignore the fact that going back further, i.e., a more inclusive average, reveals the opposite of what you keep claiming? And everytime you actually link long term data, you conveniently neglect to add a trend line that covers it.

THAT, was not an honest question. No one really wonders.
... with gigantic melancholies and gigantic mirth, to tread the jeweled thrones of the Earth under his sandalled feet.

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Sat Jun 21, 2014 9:24 pm UTC

Izawwlgood wrote:
SFX wrote:And of course, like anything even remotely connected to climate, that will be fought over tooth and nail. I mean, it's so important some people will actually want to kill you over the weather. or the weather records.
You're definitely right about that point.
That fraud was debunked a while ago.

Izawwlgood wrote:
SFX wrote:I think the twenty year trend line shows up clearly in the graphs. So does a 25 year trend line.
Again, why do you insist on selecting spans of time like that? Why do you continue to ignore the fact that going back further, i.e., a more inclusive average, reveals the opposite of what you keep claiming? And everytime you actually link long term data, you conveniently neglect to add a trend line that covers it.
It's sad the level of ignorance that we see these days. Maybe a story would help.

"Over the last twenty years, even 25 years, the data shows a trend of colder winters. not warming, as climatecentral tries to say"
Why do you ignore the 44 year trend?
"Did you not read what I just said? In the last few decades. I clearly said what climatecentral did, starting at the coldest period is fraud"

User avatar
Izawwlgood
WINNING
Posts: 18686
Joined: Mon Nov 19, 2007 3:55 pm UTC
Location: There may be lovelier lovelies...

Re: 1321: "Cold"

Postby Izawwlgood » Sat Jun 21, 2014 10:32 pm UTC

SFX wrote:That fraud was debunked a while ago.
Unfortunately, it's not the only story.

SFX wrote:"Over the last twenty years, even 25 years, the data shows a trend of colder winters. not warming, as climatecentral tries to say"
Why do you ignore the 44 year trend?
"Did you not read what I just said? In the last few decades. I clearly said what climatecentral did, starting at the coldest period is fraud"
Yes, your ability to cherry pick what you respond to has been repeatedly observed.
... with gigantic melancholies and gigantic mirth, to tread the jeweled thrones of the Earth under his sandalled feet.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Sat Jun 21, 2014 11:10 pm UTC

SFX wrote:
gmalivuk wrote:If climatecentral had also used the Columbia data and started in 1980, the trend would have looked even stronger.
You keep calling their cherry picked method "a trend", which it is not. You could pick -5, or 10 degrees to create a graph, and see something different. It's a fraudulent method of avoiding what the data actually shows. We do not use that for climate data, for a reason.
No, I do not call their method a trend. I call the linear trend of the data they showed a trend. Because that's what it is. It might not be a very accurate trend, and it might rely on cherry-picked data, but what both of us mean when we've talked about "trends" in this thread is the linear regression model fit with some set of time series data.

Also, picking other temperatures does not show something different. I've mostly stuck with minimums < 0, because that's what the comic and climatecentral are both about, but I've posted <32 data as well. And lots of other temperature cutoffs and the like show a trend that looks like slight recent cooling just so long as you include the admittedly unusually cold 2013-2014 winter. But you're talking about a trend, so presumably it should still show up at least to some extent if you drop the last winter, right? I mean, if winters have been getting colder for 25 years, then surely they were already getting colder for the first 24, right? But as I pointed out above with the Columbia <0 data, dropping that most recent outlier just makes the warming trend show up even more strongly than it already did. Here's Columbia <=10 slopes up through the 2012-2013 winter:

Code: Select all

1890   13   -03.51
1891   11   -03.70
1892   18   -03.97
1893   35   -03.97
1894   13   -03.25
1895   35   -03.44
1896   07   -02.67
1897   15   -03.11
1898   10   -03.20
1899   33   -03.53
1900   31   -02.80
1901   07   -02.13
1902   26   -02.57
1903   14   -02.11
1904   26   -02.22
1905   33   -01.72
1906   13   -00.83
1907   08   -00.95
1908   09   -01.35
1909   12   -01.71
1910   26   -01.92
1911   09   -01.34
1912   25   -01.72
1913   15   -01.18
1914   09   -01.20
1915   20   -01.59
1916   12   -01.32
1917   24   -01.54
1918   41   -00.98
1919   09   00.75
1920   14   00.40
1921   05   00.38
1922   09   -00.29
1923   10   -00.71
1924   21   -01.09
1925   16   -00.65
1926   08   -00.56
1927   07   -01.12
1928   15   -01.79
1929   24   -01.84
1930   24   -01.11
1931   01   -00.33
1932   06   -01.57
1933   15   -02.44
1934   08   -02.52
1935   08   -03.29
1936   33   -04.11
1937   09   -02.43
1938   12   -03.14
1939   06   -03.58
1940   23   -04.72
1941   06   -04.00
1942   11   -05.22
1943   17   -05.94
1944   13   -05.96
1945   16   -06.49
1946   10   -06.66
1947   16   -07.66
1948   21   -07.90
1949   12   -07.43
1950   09   -08.24
1951   21   -09.58
1952   16   -09.13
1953   02   -09.44
1954   08   -12.15
1955   11   -14.06
1956   12   -15.61
1957   08   -17.11
1958   23   -19.52
1959   22   -19.14
1960   20   -18.88
1961   17   -18.99
1962   20   -19.74
1963   31   -19.84
1964   19   -17.20
1965   18   -17.24
1966   15   -17.52
1967   09   -18.61
1968   14   -21.53
1969   14   -23.28
1970   17   -25.21
1971   13   -26.34
1972   12   -28.94
1973   16   -32.20
1974   19   -34.31
1975   13   -35.40
1976   13   -39.10
1977   34   -43.24
1978   36   -37.92
1979   31   -30.42
1980   14   -24.03
1981   19   -25.94
1982   24   -25.04
1983   02   -20.56
1984   20   -30.17
1985   26   -28.37
1986   21   -21.18
1987   06   -16.36
1988   17   -24.03
1989   16   -22.46
1990   15   -21.30
1991   17   -20.75
1992   05   -17.17
1993   13   -29.35
1994   18   -32.48
1995   11   -27.37
1996   20   -33.54
1997   18   -21.57
1998   04   -08.97
1999   14   -30.36
2000   05   -28.35
2001   23   -58.79
2002   09   -22.73
2003   12   -38.18
2004   11   -42.42
2005   08   -55.00
2006   07   -105.95
2007   18   -214.29
2008   12   -211.43
2009   12   -320.00
2010   16   -540.00
2011   17   -700.00
2012   02   100.00

As you can see, nearly every single starting point, even the ones that looked like they showed cooling in the <0 data, shows warming now, except for the very last one because there were 2 days that got below 10F in 2012 and three in 2013. (I'll note for your reference, though, that a starting point in 1919, 1920, or 1921 also seems to show slight winter cooling, if you want to cherry-pick a 90-year trend in the future.)

Let's also look at your other lovely "trends" with last winter dropped:
St. Louis, 1990-2013: -5.6F/cen - If your trend gets cut in half from the removal of one data point, it's not very robust.
1994-2013: +0.5F/cen - The "twenty-year trend" that you say "shows up clearly in the graphs" is entirely nonexistent without last winter.

Columbia's twenty-year trend is warming, +3.5F/cen
Its 1990+ trend is cut down to -1.9F/cen

The Missouri average 20-year trend is up at +0.3F/cen

So who's cherry picking now? All the trends you claim exist either greatly shrink or disappear entirely with the removal of just one data point: the abnormally cold 2013-2014 winter.

And since the comic was implying that winters had been getting warmer prior to the cold snap last January, I'd say that's very strong evidence in favor of Randall's position and against your own.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
addams
Posts: 10274
Joined: Sun Sep 12, 2010 4:44 am UTC
Location: Oregon Coast: 97444

Re: 1321: "Cold"

Postby addams » Sun Jun 22, 2014 4:49 pm UTC

PinkShinyRose wrote:
addams wrote:
Pfhorrest wrote:
schapel wrote:I think days below zero should be banned.

Vote Lord Kelvin for Hegemon of Earth and it shall be so!

Even at High Altitude?? :shock:
May we have an exception for High Altitudes?

We may be able to meet the goal at Sea Level.
But; At Atlitude? That's asking a lot.

We can, just, choose to ignore those cold days at 15,000 feet.
Brisk. We can call them Brisk, not cold.

It will Work!
We are well on our Way.

It's great to have Global Mission Doable!
All bow to the Hegeman. (whoever That is)

Below 0K days at altitude at 15000 feet (that's about 5000m right?)? I don't get it.

I think it is closer to 3000m.
It's cold up that high.

Even when close to the Equator.

Is it still too cold?
Yes.

I have been looking for evidence I can measure that shows global warming.
Like the comic, it is hard to tell. Too cold is still, too cold.
Life is, just, an exchange of electrons; It is up to us to give it meaning.

We are all in The Gutter.
Some of us see The Gutter.
Some of us see The Stars.
by mr. Oscar Wilde.

Those that want to Know; Know.
Those that do not Know; Don't tell them.
They do terrible things to people that Tell Them.

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Sun Jun 22, 2014 10:32 pm UTC

gmalivuk wrote: I mean, if winters have been getting colder for 25 years, then surely they were already getting colder for the first 24, right? But as I pointed out above with the Columbia <0 data, dropping that most recent outlier just makes the warming trend show up even more strongly than it already did.
It's hard to see what you are trying to show, but this may have a technical side. Part of the problem might be linking to trends with the NCDC data doesn't work, even if you see it on your end, I've checked, and for anyone else it always shows the trend starting in 1895, which is why I take snapshots of the graphs, rather than linking. There is currently no way to make a link work with the trend as you want it.
gmalivuk wrote: Let's also look at your other lovely "trends" with last winter dropped: St. Louis, 1990-2013: -5.6F/cen - If your trend gets cut in half from the removal of one data point, it's not very robust.
Here's what I think you are trying to link to. The St Louis min temp winter trend leaving off 2014.

Image

And you say when you use 2014 the trend doubles

Image

Which leads you to say it's "not robust". Probably because you discovered, while trying to find any trend from 1990 that would show warming, it was hard to find. But you seem dedicated, so I suspect you have.
I think you are trying to show this.

Image

And you did it! You found a 21 year period, starting with two cold years, and ending with two warm ones, that supports your belief! ! You cherry picked, and it's wonderful. If the winters continue like that for a hundred years, we will see winters on average .3F warmer!
gmalivuk wrote:So who's cherry picking now?
You are!

But why not use 1990-2014? Oh right. Cooling. Why not just back the start date of the trend backwards? Go back 30 years?

And why did you switch from minimum to average temps when you used 1993-2013?

Image

Oh, the minimum temperature shows cooling. No matter, it's best to just ignore anything that doesn't support your view.

Image

gmalivuk wrote:All the trends you claim exist either greatly shrink or disappear entirely with the removal of just one data point: the abnormally cold 2013-2014 winter.
No, and I hope you can see why that's ridiculous now. You have to also pick the right start year, and avoid using minimum temperatures. Then you can make it vanish. Otherwise, no matter how hard you try, you can't make it look like winters are warming in the last two decades, Or even 25 years.

Image

Isn't that interesting.

But wait, there's more!

gmalivuk wrote:And since the comic was implying that winters had been getting warmer prior to the cold snap last January, I'd say that's very strong evidence in favor of Randall's position and against your own.


I don't know what Randall's position is, but mine is to look at the evidence, all of it, and use rational thinking, and a skeptical approach, to see what is actually happening.

And the Missouri winters show a very similar pattern to many other areas, when it comes to winter.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Sun Jun 22, 2014 10:56 pm UTC

Pay some attention to what I'm replying to. I chose locations and statistics (minimum vs average temps) based entirely on what you yourself had just posted before me. I picked 1990 and 1993 or 1994 because you were talking about 20-25 year trends. I was just posting your own data but without 2014, basically.

You yourself used St. Louis minimum, Columbia minimum, and Missouri statewide average.
In your post.
Which is what I was responding to.

And in any case, the fact that you are able to do some more cherry picking to find other things that support your belief isn't the point. The point is that if you can add or subtract a single outlier to significantly change (possibly to the point of reversing or eliminating) the alleged trend, it was never a very strong trend in the first place.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
Crissa
Posts: 294
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1321: "Cold"

Postby Crissa » Mon Jun 23, 2014 4:12 am UTC

Did you recently read this article, gma?
http://www.serpentine.com/blog/2014/06/ ... jackknife/

-Crissa

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Mon Jun 23, 2014 4:33 am UTC

No, I didn't, but that is interesting.

I just noticed how much of an outlier 2014 was among recent winters, in pretty much every metric. So I decided to take a look at the (alleged) trends up to 2013 and saw that most of them went away.

Even long-term the variation from one year to the next tends to be much bigger than the difference between the start and end of the period we're looking at, so an outlier or two can have a pretty strong effect on whatever we're trying to compute. But it makes it especially difficult when talking about particularly short trends, like SFX's repeated claim that winters have been cooling over the past decade or two.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

schapel
Posts: 244
Joined: Fri Jun 13, 2014 1:33 am UTC

Re: 1321: "Cold"

Postby schapel » Mon Jun 23, 2014 1:28 pm UTC

SFX wrote:I don't know what Randall's position is, but mine is to look at the evidence, all of it, and use rational thinking, and a skeptical approach, to see what is actually happening.

You may want to use a Theil-Sen estimator and compute a 95% confidence interval for the trend. This will help you see whether the warming or cooling trend is real or is just showing the temperature escalator, which is essentially finding a "trend" in noise.

Update: I performed these calculations on the July-August data presented earlier in the thread, for the periods from 1960, 1970, 1980, and 1990 to the present. The median slope always showed a warming trend, but even a 50% confidence interval (!) includes the possibility of a cooling trend.

Code: Select all

1990 to 2013
Median slope: 0.100 degrees per year
50% confidence interval: -0.11 to 0.29

1980 to 2013
Median slope: 0.027 degrees per year
50% confidence interval: -0.14 to 0.19

1970 to 2013
Median slope: 0.029 degrees per year
50% confidence interval: -0.10 to 0.15

1960 to 2013
Median slope: 0.035 degrees per year
50% confidence interval: -0.07 to 0.12

So the data shows a warming trend, but the data is so noisy it's hard to be really sure that we're not just seeing a "trend" in the noise.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Mon Jun 23, 2014 5:22 pm UTC

schapel wrote:even a 50% confidence interval (!) includes the possibility of a cooling trend.
Doesn't that just mean that a lot of years are followed by other years which are colder? Naively computing a confidence interval with periodic data like sea level includes trends with negative slope, at least for the 95%CI, simply because adjacent data points are weighted the same as distant ones.

The Siegel variation, where you compute the median for each point and then the median of those, seems more robust to finding underlying trends even in very noisy data.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Mon Jun 23, 2014 6:02 pm UTC

schapel wrote:You may want to use a Theil-Sen estimator and compute a 95% confidence interval for the trend.

Tamino thinks both methods will work, but mentions no scientific papers on the matter.
http://tamino.wordpress.com/2013/03/02/theil-sen/

Which strange since I easily found hundreds.

Tyndmyr
Posts: 11443
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 1321: "Cold"

Postby Tyndmyr » Mon Jun 23, 2014 6:36 pm UTC

SFX wrote:
schapel wrote: It's a fool's errand to argue about it. Even if you show a clear warming trend, someone will just post one photo of a weather station next to the heat output of an air conditioner and claim the data is just showing the urban heat island effect.
I tend to agree, and it goes both ways.

I show that the Columbia data shows a more realistic picture, because the weather station isn't in the middle of a huge city, next to a runway at a major airport. Using the NOW data it's very obvious that the St Louis data is not good enough for climate data.

And it will somehow be brushed aside, like it means nothing.

You show the weather station right in the middle of a modern airport, and it just doesn't mean anything to somebody who has already decided it must be warming there.

Then you show, even with all that, the winters (and summers) show no warming for decades. Cooling actually, and that gets brushed aside, ignored, or worse.

What's the point? facts and data aren't going to sway anyone. If that were the case, there would be no arguments about the weather/climate. We would all be one happy bunch of people holding hands and singing songs.

I mean, once again the lush green Rolling Hills are showing 84.2, but the industrialized stations are showing 92.3

Right at this moment in Missouri. But the airport in St Louis will be the final word. The climate record.


Heat islands are an issue, but they're not entirely avoidable...we simply are industrializing previously unindustrialized areas, so it's going to be an ongoing thing. However, climate is long term, so I think it is unlikely that I can replace weather stations with personal observation. I could use personal observation to, perhaps, validate my closest weather station's accuracy, but realistically, that requires more work that I want to put in for relatively modest amounts of data.

So, how do we go about collecting data? We can use places not subject to heat island effects, or attempt to correct for known effects, or both, comparing the two against each other for accuracy. This latter method seems most reasonable, and I believe that is, in fact, what is commonly done currently.

Yes, if you go quite a while back, our climate data isn't as good, because less stations, less accuracy, and so forth, but it's still SOME data. It's not entirely without value.

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Mon Jun 23, 2014 6:46 pm UTC

gmalivuk wrote:[The Siegel variation, where you compute the median for each point and then the median of those, seems more robust to finding underlying trends even in very noisy data.
You should try it on a hundred year period, using the Missouri data. Even with the past being adjusted down, to make it seem colder, you can still see the most amazing things. Using the actual data, not the adjusted data, it's really amazing.

Missouri was actually a great example to use for the comic, since it sort of sits right in the middle of a large continental climate zone, has a lot of data, and is subject to both the arctic air in winter, and the great plains and warm ocean climate the rest of the year.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Mon Jun 23, 2014 7:06 pm UTC

SFX wrote:
schapel wrote:You may want to use a Theil-Sen estimator and compute a 95% confidence interval for the trend.

Tamino thinks both methods will work, but mentions no scientific papers on the matter.
http://tamino.wordpress.com/2013/03/02/theil-sen/
That may be the case for worldwide data, which already does a pretty good job smoothing itself out.

Also, it may be the case that the median T-S estimator is almost always close to the least-squares slope, but that doesn't necessarily hold for the entire confidence interval.

Take the 1975-2012 data from the tabular version of the second graph here
Image

Whatever we may say about the methods and significance behind this graph, we can all agree that there's a rising trend from the 70s to today, right?
The least-squares slope is 2.20degC/cen
The T-S median slope is 2.16degC/cen
But the naive 80% CI for the T-S slope is -0.60degC/cen to +5.17degC/cen, both of which are absurd given the data. For a confidence interval to have meaning with this method, I think you'd at the very least need to give more weight to more distant data points.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

schapel
Posts: 244
Joined: Fri Jun 13, 2014 1:33 am UTC

Re: 1321: "Cold"

Postby schapel » Mon Jun 23, 2014 7:14 pm UTC

SFX wrote:
gmalivuk wrote:[The Siegel variation, where you compute the median for each point and then the median of those, seems more robust to finding underlying trends even in very noisy data.

You should try it on a hundred year period, using the Missouri data.

Why a 100-year period? The warming started around 1970. Using the Seigel variation shows clearer evidence of warming, but it's still not entirely conclusive. This is why I tend to look at the sea level rise and the melting of ice, which clearly show warming, no doubt about it!

Code: Select all

1960 to 2013
Median of medians: 0.032 degrees per year
50% confidence interval: 0.009 to 0.048 degrees per year

1970 to 2013
Median of medians: 0.029 degrees per year
50% confidence interval: -0.006 to 0.067 degrees per year

1980 to 2013
Median of medians: 0.040 degrees per year
50% confidence interval: -0.071 to 0.100 degrees per year

1990 to 2013
Median of medians: 0.114 degrees per year
50% confidence interval: -0.025 to 0.194 degrees per year

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Mon Jun 23, 2014 9:03 pm UTC

schapel wrote:
SFX wrote:You should try it on a hundred year period, using the Missouri data.

Why a 100-year period? The warming started around 1970.
There's exactly why you should actually see for yourself. It's a fascinating thing to look at long term data, and poke around for a few hours, being all skeptical and stuff. You learn things that somebody just telling you things will never learn ya.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Mon Jun 23, 2014 9:48 pm UTC

Yeah, it is interesting to look at long-term data and do various analyses yourself (as I'd been doing with the lists of regression slopes by start year in a few earlier posts).

It's also interesting that after tons of posts about recent cooling, once we start discussing how more robust statistical methods show continued warming even when restricted to the latest couple of decades, you now advocate going back to the beginning of the time series with the method you already know gives a 0 as the result.

But the thing about this particular robust estimator is that, while it is less resistant to being pulled one way or another by a few outliers, it is actually more susceptible than least-squares to giving nonsensical results when the trend isn't linear. (If you have a straight horizontal line for most of the time period, followed by literally anything at all for the remainder, including a sharp and significant rise, the median of medians method will still give you zero slope.)
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

schapel
Posts: 244
Joined: Fri Jun 13, 2014 1:33 am UTC

Re: 1321: "Cold"

Postby schapel » Tue Jun 24, 2014 12:03 am UTC

gmalivuk wrote:more robust statistical methods show continued warming even when restricted to the latest couple of decades

Ah, just the urban heat island effect!
Image

speising
Posts: 2354
Joined: Mon Sep 03, 2012 4:54 pm UTC
Location: wien

Re: 1321: "Cold"

Postby speising » Tue Jun 24, 2014 12:06 am UTC

schapel wrote:
gmalivuk wrote:more robust statistical methods show continued warming even when restricted to the latest couple of decades

Ah, just the urban heat island effect!
Image


that's a cooling unit, isn't it? will that have an impact on winter measurements?

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Tue Jun 24, 2014 12:29 am UTC

speising wrote: will that have an impact on winter measurements?
If it's a scientific question, the matter could be resolved. Using science. How cool would that be? Of course, if you start making temperature measurements and station location and records all scientific, there won't be much to argue about.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Tue Jun 24, 2014 12:32 am UTC

Science like, "this cooling unit won't be running during the winter, and therefore has no impact on winter measurements"? I'm not sure what more you're suggesting in that case.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Tue Jun 24, 2014 12:48 am UTC

schapel wrote:
gmalivuk wrote:more robust statistical methods show continued warming even when restricted to the latest couple of decades

Ah, just the urban heat island effect!
Image

In 1900, this thermometer was measuring the temperature of miles and miles of pasture. Today, it is measuring the temperature of acres of asphalt in the middle of a growing city.

http://www.coyoteblog.com/coyote_blog/2 ... ionin.html

I think the point that they add an AC unit, 10 feet away, in 1985, and the annual temperatures happen to increase after this event, is pretty damn interesting.
Image (img src http://www.coyoteblog.com/coyote_blog/2 ... ionin.html)

But the true warmer believes that somehow they changed the daily readings to account for the changes. Science would have us compare the data with all the other records from the area, to see if there really was a huge annual increase in temperatures, starting in 1984.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Tue Jun 24, 2014 1:05 am UTC

Okay, but how about winter temperatures?
Last edited by gmalivuk on Tue Jun 24, 2014 2:44 am UTC, edited 2 times in total.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
Crissa
Posts: 294
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1321: "Cold"

Postby Crissa » Tue Jun 24, 2014 2:34 am UTC

SFX wrote:But the true warmer believes that somehow they changed the daily readings to account for the changes. Science would have us compare the data with all the other records from the area, to see if there really was a huge annual increase in temperatures, starting in 1984.

Hmm. Maybe someone already answered when you said this...

But hey, we're having a polite conversation, aren't we?

gmalivuk wrote:Okay, but how about winter temperatures?

Depending on the local features, winter temperatures can either be warmer or colder, depending. The thermal mass of large objects which block the wind raise the temperature, while pavement and walls lower the temperature. Airport temps will be lower while suburb temps will be a wash while city temps will remain higher. In the summer suburbs have high evaporation so change temperature more gradually while airports spike (no thermal mass) and the city core goes up faster and comes down slower.

So it depends.

-Crissa

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Tue Jun 24, 2014 2:45 am UTC

gmalivuk wrote:Okay, but how about winter temperatures?

djf-forestgrove.png

Looks like nothing much changed before and after 1985...

(I know UHI can have different effects on winter readings. I was more specifically addressing the AC thing.)
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Tue Jun 24, 2014 4:11 am UTC

Crissa wrote:
SFX wrote:But the true warmer believes that somehow they changed the daily readings to account for the changes. Science would have us compare the data with all the other records from the area, to see if there really was a huge annual increase in temperatures, starting in 1984.
Hmm. Maybe someone already answered when you said this...
I would venture you didn't actually read your "source", but just copied and pasted from an app or something. Because the very first thing that idiot blog does is use Hansen 2001 as a source. That's some humor right there you know.

You have to laugh at the logic and reason behind the handwaving away the very real problem of station location and modern civilization and the effect on local weather and climate. I mean, I brought up the issue of the St Louis station used (in the comic), and the response is to a page that actually says
They found in most cases, urban warming was small and fell within uncertainty ranges. Surprisingly, 42% of city trends are cooler relative to their country surroundings as weather stations are often sited in cool islands (eg - a park within the city). The point is they're aware of UHI and rigorously adjust for it when analysing temperature records.
And no, they don't. I showed you how to see this for yourself, just observe the various stations daily around St Louis, and compare to the official NWS records. The problem is extremely obvious, and anyone can prove this. But wait, there's more.
This confirms a peer review study by the NCDC (Peterson 2003) that did statistical analysis of urban and rural temperature anomalies and concluded "Contrary to generally accepted wisdom, no statistically significant impact of urbanization could be found in annual temperatures... Industrial sections of towns may well be significantly warmer than rural sites, but urban meteorological observations are more likely to be made within park cool islands than industrial regions."
Once again, the example was the St Louis NWS situated now in the middle of an airport. I even showed you pictures of it. Didn't matter, won't matter, can't matter. You are sure everything is OK with that "climate" data, even when it is nowhere near what the other stations report each day.

And FYI using skepticalscience as a substitute for a discussion means you lose. Completely and with no qualms, you are done. Over. It's like a Jesus witness pointing to the Bible in a debate about the existence of God. You lose. But you will never know it.

User avatar
Izawwlgood
WINNING
Posts: 18686
Joined: Mon Nov 19, 2007 3:55 pm UTC
Location: There may be lovelier lovelies...

Re: 1321: "Cold"

Postby Izawwlgood » Tue Jun 24, 2014 4:12 am UTC

SFX wrote:And FYI using skepticalscience as a substitute for a discussion means you lose. Completely and with no qualms, you are done. Over. It's like a Jesus witness pointing to the Bible in a debate about the existence of God. You lose. But you will never know it.
Oh, this was good for a laugh. Oh yes.
... with gigantic melancholies and gigantic mirth, to tread the jeweled thrones of the Earth under his sandalled feet.

User avatar
Crissa
Posts: 294
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1321: "Cold"

Postby Crissa » Tue Jun 24, 2014 5:16 am UTC

Izawwlgood wrote:
SFX wrote:And FYI using skepticalscience as a substitute for a discussion means you lose. Completely and with no qualms, you are done. Over. It's like a Jesus witness pointing to the Bible in a debate about the existence of God. You lose. But you will never know it.
Oh, this was good for a laugh. Oh yes.

Yeah, how honestly dishonest can you get?

SFX wrote:Because the very first thing that idiot blog does is use Hansen 2001 as a source.

Yes, because you don't link to any refutation, heck your link to the study (as if that would self-refute?) doesn't work. Do you believe NASA is a conspiracy or something? I fear to ask.

SFX wrote:I mean, I brought up the issue of the St Louis station used (in the comic), and the response is to a page that actually says
They found in most cases, urban warming was small and fell within uncertainty ranges. Surprisingly, 42% of city trends are cooler relative to their country surroundings as weather stations are often sited in cool islands (eg - a park within the city). The point is they're aware of UHI and rigorously adjust for it when analysing temperature records.
And no, they don't.

Okay, so did you know that in most states, despite the urbanizing, there are more trees today than a hundred years ago? I guess not. It's a normal anti-environmentalist talking point. Out here in the west, we forget that, since urbanizing is actually destroying forest faster than we can grow it back, and logging in the 80s was cutting more board feet of timber than at any time in US history, including the redwoods. But the forest I live in was flattened a hundred year ago to rebuild San Francisco, and few of the trees are older than 60 years.

Anyhow, did you compare the site you've been pecking about to nearby ones to make sure the overlapping charts didn't match, like the skeptical science linked to Hansen doing? Did every site install an air conditioner the same year?

-Crissa

User avatar
Izawwlgood
WINNING
Posts: 18686
Joined: Mon Nov 19, 2007 3:55 pm UTC
Location: There may be lovelier lovelies...

Re: 1321: "Cold"

Postby Izawwlgood » Tue Jun 24, 2014 10:48 am UTC

It's a loosing battle; you find data that runs contrary to what SFX wants to believe, and they'll say the data is bad because the collection site is next to a Starbucks. You point out errors or biases in their analysis, and they shift to different time spans until there's a trend that fits what they want. You refute a claim, they ignore it and instead post a battery of sat photos of airports. We're looking at St Louis just kidding Missouri just kidding Columbia just kidding 95-2014 just kidding 80-2014 just kidding 100 years.

I'm dizzy with the frequency at which they've shifted goal posts. But remember kids, SFX is a true scientist and seeker of truth, a skeptic who can only be swayed with rational thinking!
... with gigantic melancholies and gigantic mirth, to tread the jeweled thrones of the Earth under his sandalled feet.

Tyndmyr
Posts: 11443
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 1321: "Cold"

Postby Tyndmyr » Tue Jun 24, 2014 4:53 pm UTC

Izawwlgood wrote:I'm dizzy with the frequency at which they've shifted goal posts. But remember kids, SFX is a true scientist and seeker of truth, a skeptic who can only be swayed with rational thinking!


Yeah, this is what bugs me. I'm very open to finding a different set of assumptions/data, and fully exploring that to see what that reveals. Might be enlightening. But when the instant you get your teeth into it, they ditch, and change the ground rules again, and again...it becomes clear that it is not an honest exploration, but merely avoiding anything that looks like it might go somewhere they don't wish to think about.

User avatar
SFX
Posts: 194
Joined: Sun Jun 21, 2009 1:33 am UTC

Re: 1321: "Cold"

Postby SFX » Wed Jun 25, 2014 2:18 pm UTC

Crissa wrote:
SFX wrote:Because the very first thing that idiot blog does is use Hansen 2001 as a source.

Yes, because you don't link to any refutation, heck your link to the study (as if that would self-refute?) doesn't work.
The joke is pretty simple. That link is what the idiot SS uses to "source" their argument on the page you inked to. It's a dead link.

Priceless.

Here's a working link, and if you ever bother to actually read the paper, you will laugh so hard, looking back on the recent discussion here. And here's another good paper to read, for good measure.

But back to the truly hilarious part. SS (and idiots who blindly believe it) wants to use Hansen to justify no UHI effect, but they are so actually stupid, they don't realize what the papers actually say. For example
We find evidence of local human effects ("urban warming") even in suburban and small-town surface air temperature records, but the effect is modest in magnitude and conceivably could be an artifact of inhomogeneities in the station records.
Hansen et al 2001

That is an actual example of whatt SS considers evidence. But further down in the abstract we see the money shot. (in regards to our current discussion about St Louis)
Nevertheless, it is clear that the post-1930s cooling was much larger in the United States than in the global mean. The U.S. mean temperature has now reached a level comparable to that of the 1930s, while the global temperature is now far above the levels earlier in the century. The successive periods of global warming (1900-1940), cooling (1940-1965), and warming (1965-2000) in the 20th century show distinctive patterns of temperature change suggestive of roles for both climate forcings and dynamical variability. The U.S. was warm in 2000 but cooler than the warmest years in the 1930s and 1990s.
Yes, which is what I showed with the graphs and links and scientific official web sites. But if you were responding to Hansen I can hear you telling him, "2000 is not cooler if you start in 1970, or of you cherry pick just the right period, so you are wrong".

It's too funny.

User avatar
Crissa
Posts: 294
Joined: Sun Mar 20, 2011 8:06 pm UTC

Re: 1321: "Cold"

Postby Crissa » Wed Jun 25, 2014 4:25 pm UTC

You think it's funny that you don't know the difference between regional and global?

It's hilarious.

No, wait, it's not. It's tragic.

-Crissa

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26767
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1321: "Cold"

Postby gmalivuk » Wed Jun 25, 2014 4:42 pm UTC

SFX wrote:That is an actual example of whatt SS considers evidence. But further down in the abstract we see the money shot. (in regards to our current discussion about St Louis)
Nevertheless, it is clear that the post-1930s cooling was much larger in the United States than in the global mean. The U.S. mean temperature has now reached a level comparable to that of the 1930s, while the global temperature is now far above the levels earlier in the century. The successive periods of global warming (1900-1940), cooling (1940-1965), and warming (1965-2000) in the 20th century show distinctive patterns of temperature change suggestive of roles for both climate forcings and dynamical variability. The U.S. was warm in 2000 but cooler than the warmest years in the 1930s and 1990s.
Yes, which is what I showed with the graphs and links and scientific official web sites. But if you were responding to Hansen I can hear you telling him, "2000 is not cooler if you start in 1970, or of you cherry pick just the right period, so you are wrong".

It's too funny.
The US was not warmer in 2000 than the highest peaks of the 1930s or 1990s, but it did beat the 1930s peaks and almost match the '98 peak in 2006, and then beat all previous peaks in 2012. And the 5-year mean centered around 2000 was higher than any previous time, and the 2000-2010 average is the highest of any decade.
Image

But also what Crissa said about how tragic it is that you apparently don't understand the difference between what is happening in the continental US and what is happening globally.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: No registered users and 43 guests