NASA Faking Climate Data: Discussed Already?

[Virgil Showlion]

From the Telegraph:

...

Puzzled by those “2014 hottest ever” claims, which were led by the most quoted of all the five official global temperature records – Nasa’s Goddard Institute for Space Studies (Giss) – Homewood examined a place in the world where Giss was showing temperatures to have risen faster than almost anywhere else: a large chunk of South America stretching from Brazil to Paraguay.

Noting that weather stations there were thin on the ground, he decided to focus on three rural stations covering a huge area of Paraguay. Giss showed it as having recorded, between 1950 and 2014, a particularly steep temperature rise of more than 1.5C: twice the accepted global increase for the whole of the 20th century.

But when Homewood was then able to check Giss’s figures against the original data from which they were derived, he found that they had been altered. Far from the new graph showing any rise, it showed temperatures in fact having declined over those 65 years by a full degree. When he did the same for the other two stations, he found the same. In each case, the original data showed not a rise but a decline.

Homewood had in fact uncovered yet another example of the thousands of pieces of evidence coming to light in recent years that show that something very odd has been going on with the temperature data relied on by the world's scientists. And in particular by the UN’s Intergovernmental Panel on Climate Change (IPCC), which has driven the greatest and most costly scare in history: the belief that the world is in the grip of an unprecedented warming.

...

The article goes on into lengthy detail about how specifically the data is being fudged and its impact on climate science. Mr. Booker goes on record calling it "one of the most extraordinary scandals of our time" that "has never hit the headlines".

It's gotten the attention of the US House Science Committee.

Was this discussed during my absence? I didn't see anything about it in Paul's serial global warming thread.

[happyhoix]

Here is an explaination of the data homogenation by a science fact checking website:

Monitoring organizations like NOAA use data from other stations nearby to try and adjust for these types of issues, either raising or lowering the temperature readings for a given station. This is known as homogenization. The most significant adjustment around the world, according to NOAA, is actually for temperatures taken over the oceans, and that adjustment acts to lower rather than raise the global temperature trend.

An October 2011 paper in the Journal of Geophysical Research provides an overview of the entire Global Historical Climatology Network’s temperature data set, including detailed information about adjustments. In total, at least one “bias correction” was applied to 3,297 of the 7,279 stations in use at some point since 1801, though most of these occurred from the 1950s through the 1980s. As the chart below shows, there are approximately equal numbers of adjustments in the positive and negative directions.


www.factcheck.org/2015/02/nothing-false-about-temperature-data/

Guess Homewood only found those particular data points that were adjusted in the positive direction - funny how that works.

And, if anyone cares about what climate scientists think, there is this:

Even as these claims of data manipulation have resurfaced, there is now a general consensus that 2014 was the hottest single year since temperature record keeping began. This same conclusion has been reached by NOAA and NASA, the Japan Meteorological Agency, and the World Meteorological Organization. The United Kingdom’s Met Office said that 2014 was among the warmest along with 2010, but it is impossible to say for sure that 2014 was hotter. According to NASA, nine of the 10 warmest years have all occurred since 2000, with 1998 the lone exception.

We've beaten this dead horse many times, this won't change anyone's opinion on global warming, they will either think it's happening or they won't.

[Virgil Showlion]

Mar 12, 2015 15:03:36 GMT -5 happyhoix said:

Here is an explaination of the data homogenation by a science fact checking website:

Monitoring organizations like NOAA use data from other stations nearby to try and adjust for these types of issues, either raising or lowering the temperature readings for a given station. This is known as homogenization. The most significant adjustment around the world, according to NOAA, is actually for temperatures taken over the oceans, and that adjustment acts to lower rather than raise the global temperature trend.

An October 2011 paper in the Journal of Geophysical Research provides an overview of the entire Global Historical Climatology Network’s temperature data set, including detailed information about adjustments. In total, at least one “bias correction” was applied to 3,297 of the 7,279 stations in use at some point since 1801, though most of these occurred from the 1950s through the 1980s. As the chart below shows, there are approximately equal numbers of adjustments in the positive and negative directions.


www.factcheck.org/2015/02/nothing-false-about-temperature-data/

Guess Homewood only found those particular data points that were adjusted in the positive direction - funny how that works.

And, if anyone cares about what climate scientists think, there is this:

Even as these claims of data manipulation have resurfaced, there is now a general consensus that 2014 was the hottest single year since temperature record keeping began. This same conclusion has been reached by NOAA and NASA, the Japan Meteorological Agency, and the World Meteorological Organization. The United Kingdom’s Met Office said that 2014 was among the warmest along with 2010, but it is impossible to say for sure that 2014 was hotter. According to NASA, nine of the 10 warmest years have all occurred since 2000, with 1998 the lone exception.

We've beaten this dead horse many times, this won't change anyone's opinion on global warming, they will either think it's happening or they won't.

Agreed. But it's important to note that the actual changes in temperature over the world are apparently dwarfed by the "homogenization" of the readings. Hence any bias at all in the homogenization will lead to increasingly bad data.

It also invites the question of how exactly they determine the drift factors.

As the chart below shows, there are approximately equal numbers of adjustments in the positive and negative directions.

It does?

It looks like a mess of revisions that could easily introduce a net positive bias.

[happyhoix]

Hence any bias at all in the homogenization will lead to increasingly bad data.

If a computer is doing the homogenization, it will not bias the data. The computer is only looking for outliers, either too high or too low. As the report states, the homogenization created approximately equal adjustments to the high and the low values, but an ardent global warming denier is only going to jump on the adjustments he finds where a low value was changed to a higher one.

Somewhere, some scientist nerd could run a massive computer program to evaluate all the data points ever collected and determine the statistical variation of the data that was homogenized to determine the exact number that were nudged upwards or nudged downwards, to scientifically prove if there was any difference between the two sets of data.

However, most Americans are scientifically-phobic and prefer to trust their 'gut' feelings (e.g. - the whole vaccine drama where mommies believe they know better than scientists if their babies should get whooping cough). So everyone will ignore the climate scientist who runs that program, except for other climate scientists, who already think global warming data is accurate anyway, so why bother.

The GOP congresspeople will jump on this and have hearings to they can hold up some graphs showing adjusted data points, so they can grab some time on the evening news and show themselves to be friends of business, but that's all that will come of this.

[Tennesseer]

Cold in the eastern half of North America this winter.  But many seem to not know/forget/ignore/don't want to acknowledge the fact it was warm in western North America this winter.  

We had a number of days where my city was colder than Anchorage, Alaska was this winter.

[Virgil Showlion]

Mar 13, 2015 7:18:50 GMT -5 happyhoix said:

Hence any bias at all in the homogenization will lead to increasingly bad data.

If a computer is doing the homogenization, it will not bias the data. The computer is only looking for outliers, either too high or too low. As the report states, the homogenization created approximately equal adjustments to the high and the low values, but an ardent global warming denier is only going to jump on the adjustments he finds where a low value was changed to a higher one.

Somewhere, some scientist nerd could run a massive computer program to evaluate all the data points ever collected and determine the statistical variation of the data that was homogenized to determine the exact number that were nudged upwards or nudged downwards, to scientifically prove if there was any difference between the two sets of data.

However, most Americans are scientifically-phobic and prefer to trust their 'gut' feelings (e.g. - the whole vaccine drama where mommies believe they know better than scientists if their babies should get whooping cough). So everyone will ignore the climate scientist who runs that program, except for other climate scientists, who already think global warming data is accurate anyway, so why bother.

The GOP congresspeople will jump on this and have hearings to they can hold up some graphs showing adjusted data points, so they can grab some time on the evening news and show themselves to be friends of business, but that's all that will come of this.

I sympathize with your point, although I do think that dismissing the effects of homogenization isn't as simple as looking at whether the adjustments process has zero mean. For one thing, the data from each station is extrapolated over a different area, and the three stations Homewood looked at collectively represented a huge swath. They would have a significant weighting when computing the global mean.

For another thing, what on Earth are these adjustments based on? They're suggesting that the instruments need to be recalibrated due to new surrounding buildings, foliage, wind patterns, measurement times, etc., hence at some point in the process they have to speculate "well, if this building wasn't here and the wind currents weren't thus and such, what this station would actually be reporting is two degrees higher". But based on what? If a station is near a building, how do they calculate what temperature it would be measuring if it wasn't near the building (aside from assuming that any sudden changes in the temperature readings from pre-building years to post-building years are due to the building)?

Hence having a zero mean adjustments process (and we don't even know if the adjustments process is zero mean) is insufficient to prove the legitimacy of the homogenization process as well.

I'll read through the JGR article. Hopefully it bases its conclusions on something more substantive than "approximately equal numbers of adjustments in the positive and negative directions".

[djAdvocate]

does it matter if there is a displaced hair on the elephant's back?

it is still an elephant.

[OldCoyote]

Gee, wasn't one of the original studies of global warming (when it got down to the brass tacks)  based on the tree rings of a single tree??

Again, again, and again,  scientist fudging the numbers.

Got caught with they pants down,  you are trying to explain the plaid underwear.

[OldCoyote]

"they're", I will be caught again,  DJ,     typing to fast too much to do today.

[Virgil Showlion]

Mar 13, 2015 9:40:21 GMT -5 djAdvocate said:

does it matter if there is a displaced hair on the elephant's back?

it is still an elephant.

The effects of homogenization dwarf the actual long-term variability of temperature in this case. If the process is biased even slightly, it could easily lead to wrong conclusions.

Think of it like Rasmussen polling.

[Tennesseer]

Mar 13, 2015 10:11:31 GMT -5 OldCoyote said:

Gee, wasn't one of the original studies of global warming (when it got down to the brass tacks)  based on the tree rings of a single tree??

Again, again, and again,  scientist fudging the numbers.

Got caught with they pants down,  you are trying to explain the plaid underwear.

Study of 800-year-old tree rings backs global warming

They looked at the rings of thousands of ancient trees in the mountains above the most important rivers in the West.

Study of 800-year-old tree rings backs global warming

The jury is still out on climate change/global warming but proof global warming exists was not based on a single tree's rings..

[happyhoix]

For another thing, what on Earth are these adjustments based on? They're suggesting that the instruments need to be recalibrated due to new surrounding buildings, foliage, wind patterns, measurement times, etc., hence at some point in the process they have to speculate "well, if this building wasn't here and the wind currents weren't thus and such, what this station would actually be reporting is two degrees higher". But based on what?

I am not an expert in this area, but what I understood from reading the article was that they looked at data points provided from a lot of different sources for the same area, and sometimes they had some data points that were not in line with the others - for instance, if you had five field stations in one region, and four read the temperature x, and one read the temperature x -5, then they 'homogenized' the outlying datapoints to bring them more in line with the average from the other stations. Homewood looked at just 3 rural stations, but there may be other data the system was comparing that to, to different stations or sources, for the same area. 

It's actually pretty easy to disrupt what a thermometer reads. Place it in direct sunlight as opposed to shade, or out in the wind rather than in a sheltered area, or set it near something metal that absorbs and reflects the heat - even who the thermometer manufacturer is and what type of thermometer it is - all those things can influence readings, hence that's why they homogenize the data.  

[mmhmm]

Mar 13, 2015 10:14:29 GMT -5 OldCoyote said:

"they're", I will be caught again,  DJ,     typing to fast too much to do today.

In that particular post, the correct choice would be "their". 


* just pokin' a'cha* 

[Deleted]

And still, people don't seem to get the difference between climate and weather. 

sigh....

[Deleted]

Mar 13, 2015 11:51:51 GMT -5 happyhoix said:

For another thing, what on Earth are these adjustments based on? They're suggesting that the instruments need to be recalibrated due to new surrounding buildings, foliage, wind patterns, measurement times, etc., hence at some point in the process they have to speculate "well, if this building wasn't here and the wind currents weren't thus and such, what this station would actually be reporting is two degrees higher". But based on what?

I am not an expert in this area, but what I understood from reading the article was that they looked at data points provided from a lot of different sources for the same area, and sometimes they had some data points that were not in line with the others - for instance, if you had five field stations in one region, and four read the temperature x, and one read the temperature x -5, then they 'homogenized' the outlying datapoints to bring them more in line with the average from the other stations. Homewood looked at just 3 rural stations, but there may be other data the system was comparing that to, to different stations or sources, for the same area. 

It's actually pretty easy to disrupt what a thermometer reads. Place it in direct sunlight as opposed to shade, or out in the wind rather than in a sheltered area, or set it near something metal that absorbs and reflects the heat - even who the thermometer manufacturer is and what type of thermometer it is - all those things can influence readings, hence that's why they homogenize the data.  

And here I thought they followed the same set of placement standards.                          Temperature sensor siting: The sensor should be mounted 5 feet +/- 1 foot above the ground. The ground over which the shelter [radiation] is located should be typical of the surrounding area. A level, open clearing is desirable so the thermometers are freely ventilated by air flow. Do not install the sensor on a steep slope or in a sheltered hollow unless it is typical of the area or unless data from that type of site are desired. When possible, the shelter should be no closer than four times the height of any obstruction (tree, fence, building, etc.). The sensor should be at least 100 feet from any paved or concrete  surface.  www.nws.noaa.gov/om/coop/standard.htm   wattsupwiththat.com/2007/06/21/standards-for-weather-stations/

[EVT1]

Mar 13, 2015 12:19:31 GMT -5 @joey said:

And still, people don't seem to get the difference between climate and weather. 

sigh....

[happyhoix]

Mar 13, 2015 13:14:57 GMT -5 @jma23 said:

Mar 13, 2015 11:51:51 GMT -5 happyhoix said:

I am not an expert in this area, but what I understood from reading the article was that they looked at data points provided from a lot of different sources for the same area, and sometimes they had some data points that were not in line with the others - for instance, if you had five field stations in one region, and four read the temperature x, and one read the temperature x -5, then they 'homogenized' the outlying datapoints to bring them more in line with the average from the other stations. Homewood looked at just 3 rural stations, but there may be other data the system was comparing that to, to different stations or sources, for the same area. 

It's actually pretty easy to disrupt what a thermometer reads. Place it in direct sunlight as opposed to shade, or out in the wind rather than in a sheltered area, or set it near something metal that absorbs and reflects the heat - even who the thermometer manufacturer is and what type of thermometer it is - all those things can influence readings, hence that's why they homogenize the data.  

And here I thought they followed the same set of placement standards.                          Temperature sensor siting: The sensor should be mounted 5 feet +/- 1 foot above the ground. The ground over which the shelter [radiation] is located should be typical of the surrounding area. A level, open clearing is desirable so the thermometers are freely ventilated by air flow. Do not install the sensor on a steep slope or in a sheltered hollow unless it is typical of the area or unless data from that type of site are desired. When possible, the shelter should be no closer than four times the height of any obstruction (tree, fence, building, etc.). The sensor should be at least 100 feet from any paved or concrete  surface.  www.nws.noaa.gov/om/coop/standard.htm   wattsupwiththat.com/2007/06/21/standards-for-weather-stations/

The standard you are citing was last updated in 2013. They don't have the historical standards on their website, so I don't know how long they've been providing temperature sensor siting guidance.

Do you think people were following this standard in 1990? In 1970? In 1950? (Let me help you out - NOAA was founded in 1970).  The three data stations that Homewood complained about collected data from 1950 to 2014 - in Paraguay. Do you think Paraguay was using NOAA siting guidelines in 1950? Do you think they use NOAA guidelines today?

Let me ask you this - why do you think NOAA issued such specific guidance on how to site a temperature sensor? Is it possible NOAA is aware a significant error might occur when the temperature sensor is not properly situated, due to many decades of data collection and study? 

Or maybe, being the wacky, crazy civil servants that they are, they just wrote these up for fun. I don't know any climate scientists, possibly they are all wild and crazy guys.

[Angel!]

Mar 13, 2015 11:51:51 GMT -5 happyhoix said:

For another thing, what on Earth are these adjustments based on? They're suggesting that the instruments need to be recalibrated due to new surrounding buildings, foliage, wind patterns, measurement times, etc., hence at some point in the process they have to speculate "well, if this building wasn't here and the wind currents weren't thus and such, what this station would actually be reporting is two degrees higher". But based on what?

I am not an expert in this area, but what I understood from reading the article was that they looked at data points provided from a lot of different sources for the same area, and sometimes they had some data points that were not in line with the others - for instance, if you had five field stations in one region, and four read the temperature x, and one read the temperature x -5, then they 'homogenized' the outlying datapoints to bring them more in line with the average from the other stations. Homewood looked at just 3 rural stations, but there may be other data the system was comparing that to, to different stations or sources, for the same area. 

This makes sense to me.  We do something similar with crash data - we do a regression to the mean calculation based on how far from the mean the data point is & how many years of data we have.  We have a methodology & reasons for doing so, but to an outsider who has no background in our work or statistics it could easily be viewed as fudging the numbers.

I would assume some sort of correction of outliers is fairly typical in any statistical work, either that or you have to throw out the outliers.  Taking them as is when you know the measurements are off would cause more problems.

[Deleted]

Mar 13, 2015 13:43:01 GMT -5 happyhoix said:

Mar 13, 2015 13:14:57 GMT -5 @jma23 said:

And here I thought they followed the same set of placement standards.                          Temperature sensor siting: The sensor should be mounted 5 feet +/- 1 foot above the ground. The ground over which the shelter [radiation] is located should be typical of the surrounding area. A level, open clearing is desirable so the thermometers are freely ventilated by air flow. Do not install the sensor on a steep slope or in a sheltered hollow unless it is typical of the area or unless data from that type of site are desired. When possible, the shelter should be no closer than four times the height of any obstruction (tree, fence, building, etc.). The sensor should be at least 100 feet from any paved or concrete  surface.  www.nws.noaa.gov/om/coop/standard.htm   wattsupwiththat.com/2007/06/21/standards-for-weather-stations/

The standard you are citing was last updated in 2013. They don't have the historical standards on their website, so I don't know how long they've been providing temperature sensor siting guidance.

Do you think people were following this standard in 1990? In 1970? In 1950? (Let me help you out - NOAA was founded in 1970).  The three data stations that Homewood complained about collected data from 1950 to 2014 - in Paraguay. Do you think Paraguay was using NOAA siting guidelines in 1950? Do you think they use NOAA guidelines today?

Let me ask you this - why do you think NOAA issued such specific guidance on how to site a temperature sensor? Is it possible NOAA is aware a significant error might occur when the temperature sensor is not properly situated, due to many decades of data collection and study? 

Or maybe, being the wacky, crazy civil servants that they are, they just wrote these up for fun. I don't know any climate scientists, possibly they are all wild and crazy guys.

I don't know of any climate scientist that would be concerned with any short term weather changes or how the data was collected.   If the standards were recently set or not.  Or standards were followed or not.  Or data is homogenized.  It has little to do with a possible climate change.  1950's ? It's recent weather.  Thanks, now I know where your error is.  Your confusing weather with climate. 

[Angel!]

Mar 13, 2015 14:20:11 GMT -5 @jma23 said:

Mar 13, 2015 13:43:01 GMT -5 happyhoix said:

The standard you are citing was last updated in 2013. They don't have the historical standards on their website, so I don't know how long they've been providing temperature sensor siting guidance.

Do you think people were following this standard in 1990? In 1970? In 1950? (Let me help you out - NOAA was founded in 1970).  The three data stations that Homewood complained about collected data from 1950 to 2014 - in Paraguay. Do you think Paraguay was using NOAA siting guidelines in 1950? Do you think they use NOAA guidelines today?

Let me ask you this - why do you think NOAA issued such specific guidance on how to site a temperature sensor? Is it possible NOAA is aware a significant error might occur when the temperature sensor is not properly situated, due to many decades of data collection and study? 

Or maybe, being the wacky, crazy civil servants that they are, they just wrote these up for fun. I don't know any climate scientists, possibly they are all wild and crazy guys.

I don't know of any climate scientist that would be concerned with any short term weather changes or how the data was collected.   If the standards were recently set or not.  Or standards were followed or not.  Or data is homogenized.  It has little to do with a possible climate change.  1950's ? It's recent weather.  Thanks, now I know where your error is.  Your confusing weather with climate. 

This thread is about data collected from the 1950s to today.  Discussing data collection techniques from the 1950s is exactly on point.

[Deleted]

Mar 13, 2015 14:25:07 GMT -5 Angel! said:

Mar 13, 2015 14:20:11 GMT -5 @jma23 said:

I don't know of any climate scientist that would be concerned with any short term weather changes or how the data was collected.   If the standards were recently set or not.  Or standards were followed or not.  Or data is homogenized.  It has little to do with a possible climate change.  1950's ? It's recent weather.  Thanks, now I know where your error is.  Your confusing weather with climate. 

This thread is about data collected from the 1950s to today.  Discussing data collection techniques from the 1950s is exactly on point.

Exactly! So why would she reference climate scientists in her last sentence as maybe being wild and crazy guys? No climate scientist will barely give a passing glance at short term weather statistics. Or how they're not in conformity (collection methods)or homogenized (data).  Climate science is more in the realm of the physics of energy transfer.   Here is a cut/paste from her reply #3 of this thread with more references to climate scientists which you "liked".  <<< So everyone will ignore the climate scientist who runs that program, except for other climate scientists, who already think global warming data is accurate anyway, so why bother. >>>  Seems that pointing out the difference between weather and climate was spot on in regards to this thread. 

[Virgil Showlion]

Mar 13, 2015 11:51:51 GMT -5 happyhoix said:

For another thing, what on Earth are these adjustments based on? They're suggesting that the instruments need to be recalibrated due to new surrounding buildings, foliage, wind patterns, measurement times, etc., hence at some point in the process they have to speculate "well, if this building wasn't here and the wind currents weren't thus and such, what this station would actually be reporting is two degrees higher". But based on what?

I am not an expert in this area, but what I understood from reading the article was that they looked at data points provided from a lot of different sources for the same area, and sometimes they had some data points that were not in line with the others - for instance, if you had five field stations in one region, and four read the temperature x, and one read the temperature x -5, then they 'homogenized' the outlying datapoints to bring them more in line with the average from the other stations. Homewood looked at just 3 rural stations, but there may be other data the system was comparing that to, to different stations or sources, for the same area. 

It's actually pretty easy to disrupt what a thermometer reads. Place it in direct sunlight as opposed to shade, or out in the wind rather than in a sheltered area, or set it near something metal that absorbs and reflects the heat - even who the thermometer manufacturer is and what type of thermometer it is - all those things can influence readings, hence that's why they homogenize the data.  

I don't get the impression from anything in the articles that there was/is a whole lot of redundant sensing going on here, but you might be right.

I would love to have a few million dollars in research funds at my disposal to hire a team of statisticians completely outside the climatology sphere to actually dig through the data and figure out exactly how the recalibration was conducted. To borrow from a recent meme: doing that by oneself, ain't nobody got time for dat.

[djAdvocate]

Mar 13, 2015 10:16:23 GMT -5 Virgil Showlion said:

Mar 13, 2015 9:40:21 GMT -5 djAdvocate said:

does it matter if there is a displaced hair on the elephant's back?

it is still an elephant.

The effects of homogenization dwarf the actual long-term variability of temperature in this case. If the process is biased even slightly, it could easily lead to wrong conclusions.

Think of it like Rasmussen polling.

sure, but the whole point in scientific polling is to refine your methods to match the data.  so long as they are doing that, the predictive models should improve over time.

[djAdvocate]

Mar 13, 2015 10:11:31 GMT -5 OldCoyote said:

Gee, wasn't one of the original studies of global warming (when it got down to the brass tacks)  based on the tree rings of a single tree??

Again, again, and again,  scientist fudging the numbers.

Got caught with they pants down,  you are trying to explain the plaid underwear.

i am not aware of ANY data fudging, oc.

[OldCoyote]

Early on in the global warming thing was a research done on a tree or trees in Mongolia.

If I remember right, it came to light because of hacked emails, many researchers that did not agree with the out come of the study were blacklisted and intimidated.

Does anyone here remember that??

onlinelibrary.wiley.com/doi/10.1029/2000GL011845/abstract 

[OldCoyote]

Let's not forget this also,  you know all polar bears drowning because of no ice.


news.nationalgeographic.com/news/2007/12/071212-AP-arctic-melt.html

[OldCoyote]

In that article it notes that the weakened arctic blast will reduce rain and snow in some areas,  Just say thanks Global warming for the reduced snow and warming temps.

I can say have had less snow here in Phoenix.  

[weltschmerz]

Regardless of sensors, data, debates, documentations, charts and graphs....

The glaciers are still melting.

[djAdvocate]

still waiting for that data fudging.

[OldCoyote]

Data fudging,  How about I change it to cherry picking,  that way I can say the numbers are true. Right??

OK,  none of us here would ever cherry pick, would we?