Re: [asa] Data doesn't support global warming

From: Kenneth Piers <pier@calvin.edu>
Date: Wed Dec 16 2009 - 08:14:11 EST

Friends: Not being a statistician or a data processor I should probably keep out of this. But wouldn't the obvious thing to do with a set of temperature numbers that were collected from a thermometer located near or above an outdoor AC is simply exclude them from the data set? (even though, as Rich points out, if this anomaly had endured during the entire time frame of the data collection, inclusion would likely have little material impact).
Failing exclusion, wouldn't the next obvious thing be to look for data sets collected by nearby thermometers that were not afflicted by location defects and use their results to apply correction factors to the suspect data set?
By the way it does seem fairly inconceivable to me that, if one was interested in collecting valid temperature data, one would locate the thermometer near or above an outdoor AC. I have an old thermometer outside my kitchen window on the east side of my house. In the winter this thermometer always reads about 2-3 deg F above the air temperature because of heat transfer through the nearby window. The actual magnitude of the temperature anomaly depends on the air temperature and the indoor temperature. The bigger the difference between these two, the bigger is the anomaly. I would never recommend using the raw data from this thermometer in a data set for global temperature calculations. The observed temperatures have a systematic bias. If the temperature measurements collected from this thermometer were going to be used in a data set, some "homgenization factors' would need to be applied. Wouldn't the same be true for the AC thermometer?
Ken Piers

Ken Piers

"We are by nature creatures of faith, as perhaps all creatures are; we live by counting on things that cannot be proved. As creatures of faith, we must choose either to be religious or superstitious, to believe in things that cannot be proved or to believe in things that can be disproved."
Wendell Berry
>>> "Glenn Morton" <glennmorton@entouch.net> 12/16/09 7:11 AM >>>
replies for Rich Blinne

 

 

I had asked about correcting a station with an air conditioner. Rich erroneously wrote in reply

 

 

>>>Murray answered that question see Peterson. In addition to that as long as the air conditioner is not new it doesn't materially affect the trend as again shown by Peterson. <<<<

 

 

That was not the question. Clearly you don't understand what the homogeneity correction is doing and what the difference between it and a correction of the daily temperature is. I am asking how one is to correct the DAILY temperature output of a station next to an air conditioner. The signal output--a temperature--consists of two parts. The temperature of the environment PLUS the heating of the local environment by the air conditioner. Tout=Tenviro+Tac. The true temperature is Tenviro and that is what we want to get to. But no records are kept of when the air conditioner is on, what direction the wind is blowing (next to the building might be different than the anemometer on the roof) so we can't always be sure if the heat is carried towards or away from the thermometer.

 

These data are then put into the station record and the homogeneity filter is put on top of all that. I would also suggest that if one is to cool a station, one should identify what is heating the station. If one is to heat it up as seems to be the norm then one should have a physical explanation for what is cooling the station over time. None of this is given in the corrective work. And given Balling and Idso's work of subtracting the raw data from the final data,

it is clear that warming seems to be far more common than cooling a station over time.

http://1.bp.blogspot.com/_Lxqre8hMG3M/Sh7ntlYGwEI/AAAAAAAAAXY/9GQ6qdOqSEQ/s1600-h/weatherRawVsSanitizedUStempDifferences.jpg

 

So please tell me how to correct the daily temperature record, which after all is the original data.

 

 

 

>>>>You need to choose which is superior the raw or corrected data. Between what you have shown and Peterson I would say that the corrected data is superior. <<<

Yes if the correction fits the bias of the climatologist, it is automatically better, but that isn't science. What is the physical cause of cooling in eastern Colorado that requires the changes Peterson makes to his temperature series. He doesn't specify that in the article. He merely waves the magic wand of "badly sited". What is it Rich that is cooling those stations making a heating correction necessary?

 

 

Of Siberia

Rich Replied:

 

>>>You have a bigger question since it's abundantly clear there is warmth from the buildings falling into the permafrost and the permafrost itself gaining two degrees and the coastlines eroding why doesn't your data show it? The effect in the Arctic is so patent now you don't even need a thermometer to show it.<<<

 

No, I don't you have not yet answered MY question, another masterful dodge on your part. You posted a sad story about Tiksi Russia and claimed that warming is causing subsidence. I showed that the entire area is subsiding. Please prove that your subsidence is the cause or what part it accounts for. You seem never to actually answer questions, Rich.

 

As to your claim that the permafrost is gaining temperature, I note that that is mere assertion on your part. Please provide a reference with DATA. Anyone can assert anything but a scientist gives his sources and lets other people examine the data. Mere assertion doesn't hack it in science Rich.

>>>why
the satellite data has a step function in the middle of it,

 

It's called the El Nino of the Century and is why when anybody picks 1997/1998/1999 as a start point of any trend it should be viewed with extreme suspicion. In fact, the longer the trend the better. The trends I quoted were 31 years or greater.<<<

 

Finally an answer to the question. So, now, if it is El Nino, the warming from the satellite isn't caused by CO2. Thank you.

 

Rich, 31 years is the length of the satellite data which started in 1978. My picture IS the data.

 

>>>and why are
modern thermometers made to read warmer than those of 1900?

 

Again it doesn't matter since the very homogenization process you seem to hate deals with different thermometers. Where you need to deal with this is when you change the thermometer. As long as you have an existing thermometer whether it reads warm or cold for whatever reason it's still the same thermometer and thus you can then glean a decent trend from even a biased thermometer, particularly if you homogenize it with the more accurate nearby rural stations. This is not speculation on my part and it was proven, see Peterson. http://ams.allenpress.com/archive/1520-0477/87/8/pdf/i1520-0477-87-8-1073.pdf Peterson said, "Analysis of a small subset of U.S. Historical Climatology Network data does not find a time-dependent bias caused by current poor station siting."<<<

 

Anyone who believes this is quite capable of buying swamp land. Would you put your thermostat for your AC next to the stove, so that everytime you cook a meal the AC goes on even if it is in the middle of winter???

 

CAN ANYONE ACTUALLY BELIEVE THAT AC EXHAUST DOESN'T AFFECT THE TEMPERATURE WHICH IS READ?

 

Sorry Rich, that is physically impossible.

 

here is what they are doing. They take rural stations and compare them to urban stations. .They expect that the rural stations are free of it, but I have posted links to rural stations which are next to AC units as well. So their methodology of making that conclusion is flawed.

 

 

Here are rural stations:

http://2.bp.blogspot.com/_Lxqre8hMG3M/Sq1-UFgmF2I/AAAAAAAAAlA/JFDN1f0QzlM/s1600-h/weatherOHUrbanaThermometer.JPG

 

http://3.bp.blogspot.com/_Lxqre8hMG3M/Sq1-1FMr53I/AAAAAAAAAlI/GcvI9h3iZ9M/s1600-h/St_Ignatius_AC_discharge.jpg

 

http://4.bp.blogspot.com/_Lxqre8hMG3M/Sq1_VmjCpiI/AAAAAAAAAlQ/zHd5inydsss/s1600-h/weatherWYRiverton_closeup_95-97_airconditionera.jpg

 

 

http://1.bp.blogspot.com/_Lxqre8hMG3M/Sq1_thRP4cI/AAAAAAAAAlY/S1Gk3KOW2n0/s1600-h/weatherWYNewCastleEastAirconditioner.jpg

 

So, comparing rural stations with air conditioners to city stations with air conditioners and concluding that there is no problem is stupid, Rich. The devil is always in the details.

  ----- Original Message -----
  From: Rich Blinne
  To: Glenn Morton ; Murray Hogg ; Randy Isaac
  Cc: asa
  Sent: Tuesday, December 15, 2009 11:23 PM
  Subject: Re: [asa] Data doesn't support global warming

  On Dec 15, 2009, at 9:26 PM, Glenn Morton wrote:

    For Rich.

    Would you care to explain specifically how one corrects a thermometer next to an air conditoiner so that the data is useful for determining global climate?

  Murray answered that question see Peterson. In addition to that as long as the air conditioner is not new it doesn't materially affect the trend as again shown by Peterson. If you can show contra Peterson a poorly sited station and a nearby well sited one and that the anomalies are materially different outside the error bars and show how the homogenization process misses something then I will concede your point. Or you can do what I did and show how the Australian homogenization appears to be superior to the American one. The Australians have a key advantage as they have access to the metadata which showed that Darwin PO closed for temperature in 1941 but continued to collect precipitation data through 1962. The error further appears to be not in the homogenization but in the collection of the raw data, identifying a non-existent station. Neither the Australian BOM nor GISS made this error. A human being can see this but not a computer program. Eschenbach appeared to miss!
  this fact, however, because the error is in the raw data.

  You need to choose which is superior the raw or corrected data. Between what you have shown and Peterson I would say that the corrected data is superior. Any data stands to be improved, e.g. http://www.ncdc.noaa.gov/oa/climate/research/nature02524-UW-MSU.pdf where the mid-troposphere satellite data needed to be corrected since the cooling stratosphere was producing a cooling bias in the mid-troposphere data. So, the bad sites you mentioned are worthy of investigation and also the correction methodologies but you need to go to the next step and show whether the bias in the data produces a material difference or how a proposed correction procedure is superior to the current one.

    Ducking, dodging and evading is not becoming a scientist. here is another case. http://gallery.surfacestations.org/main.php?g2_itemId=29132

    Would you care to explain the air conditioner question and why the monthly raw data doesn't show siberian warmth,

  You have a bigger question since it's abundantly clear there is warmth from the buildings falling into the permafrost and the permafrost itself gaining two degrees and the coastlines eroding why doesn't your data show it? The effect in the Arctic is so patent now you don't even need a thermometer to show it.

    why
    the satellite data has a step function in the middle of it,

  It's called the El Nino of the Century and is why when anybody picks 1997/1998/1999 as a start point of any trend it should be viewed with extreme suspicion. In fact, the longer the trend the better. The trends I quoted were 31 years or greater.

    and why are
    modern thermometers made to read warmer than those of 1900?

  Again it doesn't matter since the very homogenization process you seem to hate deals with different thermometers. Where you need to deal with this is when you change the thermometer. As long as you have an existing thermometer whether it reads warm or cold for whatever reason it's still the same thermometer and thus you can then glean a decent trend from even a biased thermometer, particularly if you homogenize it with the more accurate nearby rural stations. This is not speculation on my part and it was proven, see Peterson. http://ams.allenpress.com/archive/1520-0477/87/8/pdf/i1520-0477-87-8-1073.pdf Peterson said, "Analysis of a small subset of U.S. Historical Climatology Network data does not find a time-dependent bias caused by current poor station siting."

  Now it is a small subset so maybe you might be able to find a counter-example. So, you could show a time-dependent bias on a different subset but as of yet I haven't seen you do that. I'm waiting.

    Is ducking and dodging and art form?

  Is an argument from silence still a logical fallacy?

  Rich Blinne
  Member ASA

To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Wed Dec 16 08:14:41 2009

This archive was generated by hypermail 2.1.8 : Wed Dec 16 2009 - 08:14:41 EST