Cooling Stations
Over the course of the past few years I have taking a look at cooling stations a few times but never in much depth. A few other people have looked at them, but usually without much rigor. The standard approach is to find some cooling stations and then conclude that somehow global warming is thereby challenged as a theory. I supp that argument will come up in the comments, but my real goal here is to shed some light of the phenomena of stations that have long term cooling trends. The dataset I used here is the Berkeley Earth Data set version 2.0. I used Single value, with Quality control and seasonality left in the dataset.
The data selection process is pretty simple. We read the data in, window it to 1900 to 2011, and remove those stations that have no data
Data <- readAsArray(Directory = choose.dir())
Data <- windowArray(Data, start = 1900, end = 2011.99)
Data <- removeNaStations(Data)
In the next step I take notice of something that Carrick suggested over at Lucia’s and I focus on stations that have a high percentage of complete data. For this exercise I slected those records that had 99% of the data present for the time period. That’s 1332 months of data in this 112 year period (1344 months)
rsum <- rowSums(!is.na(Data))
dex<- which(rsum > 1331)
Data <- Data[dex, , ]
Next we align the station inventory with the data, do inverse weighting and calculate the least squares fit
Data <- intersectInvData(Stations,Data)
weights <- inverseDensity(Data$Inventory,Data$Array)
Temps <- solveTemperature(Data$Array,weights)
TIME <- seq(1900,2011)
plot(ts(Temps, start =1900, frequency = 1))
abline(reg = lm(Temps~TIME))
What we see should come as no surprise. The planet is getting warmer. The slope of the line is roughly .08C/ decade. The stations have the following geographical distribution:
The sum total of all stations is 478. The result, as we expect, is similar to the results we get if we use all stations, regardless of their length. However, we can note that the sample isn’t very representative so their could be bias due to spatial sampling. The overweighting of the northern hemisphere is likely to cause a trend that is slightly higher than a what we would see if the sample included more SH stations.
The next step I will merely describe in words because I will probably change the code I used here when I decide upon a final approach. After calculating the global average of all these long stations I then created trends and confidence intervals for every station. For trends I used a Theil-Sen estimator which may be a bit different than most folks are used to. Its described here: http://en.wikipedia.org/wiki/Theil%E2%80%93Sen_estimator. The zyp package has a quick implementation that spits out confidence intervals as well. I’ll play around with some of Lucia’s approaches and maybe look at Tamino’s approach, but for now, this will do.
Summarizing the trends we see the following:
summary(DF$Slope) Min. 1st Qu. Median Mean 3rd Qu. Max. -0.014450 0.001390 0.006038 0.006151 0.010850 0.036920
Note that the simple meanof all slopes is less than the area weighted global solution. And note that some of the slopes of individual stations are negative. These are so called cooling stations: A crude map below gives you an indication of their locations. Blue is cooling, red is warming. The warming were drawn first so they tend to get covered up.
Well there you have it. Some long stations cool. This is fairly well known but a week doesnt pass on the internet without somebody asking about it or pointing to it as proof of “something.” Before we dive into that there is a bit more work to do.
In order to look a bit deeper into the “cooling stations” one beneficial cut to make through the data is to isolate those stations that are coastal versus those that are inland. That’s interesting in and of itself. The effects of geography on climate are well known. Coastal stations will have a fewer extremes. In mathematical terms the standard deviation of monthly temperatures will be lower. Defining a range for coastal stations is somewhat problematic so I did some sensitivity analysis around that looking at the standard deviation as a function of range from the ocean. More on that in the future but for now Here are the results at 50km
And now for inland stations
An interesting question this raises is could their be a second process at work with a slightly negative mean trend. The first step in looking into that will be separating those stations which have a statistically significant cooling trend from those that do not. That is, before we set out looking for a “cause” behind the cooling stations it’s probably a wise move to look at those where the cooling trend is statistically significant. Of the 478 stations, 54 show a rate of cooling that is statistically significant (95%)
With these 54 stations and the metadata we have we can start to eliminate potential causes of the cooling. One of the first thoughts that occurs to people is “de urbanization” This has been raised several times over the past 5 years I’ve spent looking at the problem. In Our AGU poster I spent a short period of time looking at “de urbanization” and frankly didn’t find much. The number of stations that had decreases in population was very small. Nevertheless, we can look at it again very quickly for these 54 stations:
First, the historgram of year 2000 population density ( people per sq km), second the histogram for 1900, and then finally the histogram for the difference between 2000 and 1900:
As the chart shows there is only one station that had a decrease in population. That station started with 16 people per sq km in 1900 and slowly over time reduced to 14 people per sq km. Lets look at some of its metadata
Slope Upper Lower Id Lat Lon Altitude Months Dst
31137 -0.0088867 -0.005273885 -0.01257143 38048 42.3978 -90.7036 325.55 1343 1212
Sdev Anthrome2000 Anthrome1900 Lcover UrbArea Slope2 Upper2
31137 2.332492 32 32 14 0.0004538235 -0.008970948 -0.005294326
Lower2 popd_1900AD popd_1910AD popd_1920AD popd_1930AD popd_1940AD popd_1950AD
31137 -0.01267606 16.7313 15.5423 15.7447 15.2845 14.9356 15.4139
popd_1960AD popd_1970AD popd_1980AD popd_1990AD popd_2000AD
31137 15.5131 15.3346 15.1406 14.1625 14.3402
The station is 1212 km away from a body of water; and we can see that its standard deviation is high. Stations near the coast have standard deviations close to 1.4. A standard deviation of 2.33, in fact, is quite high in the last quartile of all 487 stations. Other things to note. The anthrome has not changed. Anthrome 32 is residential rainfed croplands. Lcover 14 also indicates rainfed croplands ( 300 meter data). My UrbArea paramater captures the amount of “built” area within 10km, so it looks like there is some human structure in the area. In addition to this metadata, my other sources indicate an airport location nearby, so it looks like it would be an airport built in a farming community. Checking with google earth. There you go:
So, how do we explain the cooling at a station that is located at an airport? Without looking at the station history in detail, you can probably figure that the station has a station move in its history. That move may be documented or undocumented. If there is a move, that might show up in the temperature record as a discontinuity. This is a good lesson in using Berkeley Earth Surface Data. The Single value data has not been through the scalpel. The scalpel performs the analysis to find structural breaks. Series are not homogenized or adjusted, but if there is a break then a single series is replaced with two series. Let’s see what this series looks like:
Looking at the chart the step down post 1950 looks like it may be the cause of the cooling. 1950 would be the start date of the dubuque regional airport series. When I get some time I’ll load up structchange and see if it can find the discontinuity. We can continue this process for the other 53 stations, but the speculation that all the cooling is due to “de population” really doesnt have much merit. The actual number of stations that have suffered de population or de urbanization is small. The more likely explanation for the coolings comes down to undocumented or documented station changes. If you spend some time looking through all the monthly series you’ll begin to see some patterns that usually turn out to be either station moves or station splices. At some point when the “post scalpel” data is released this will be a bit easier to understand.
Steven Mosher's Blog 
May I suggest an explanation for some of the cooling stations in the mid to eastern US.
Depopulation. Deindustrialization. Negative UHI.
There are 694 counties that have shrunk in population since 1900, many of them in the same region where most of the cooling stations are.
http://sunshinehours.wordpress.com/2012/03/17/depopulation-and-cooling-at-the-county-level/
I would suggest that first determine that the stations are ACTUALLY cooling. You know when people show warming charts that the first thing people ask for is error bars! Doh.
You’re suppositions are all testable. we will see how they pan out. In any case for the US you can ( I have ) got US population data down to the sq km, so no need to mess about with the
nonsense you tried.
This is one of the stations I think is cooling since 1900.
As for the “nonsense”, it isn’t the population, but the population change since 1900 that is important. And it may be that county population change does make a difference to all kinds of things like traffic, industry, etc etc.
I’ll repeat my advice. First determine which stations have a statistically significant rate of cooling. Then, you can start to look for cause. “depopulation” can be a cause. So can undocumented station moves. So can undocumented changes in instruments. The Single-value Quality controlled
data does not correct for all known quality issues. All you have to do is read the “readme” and check the flags data.
The reason for that is that the “scalpel” in best algorithm does
a check for undocumented changes in the data and then
series are split accordingly. Bascially, with Single Value QC you still have QC work to do. BEST has their approach to this final QC. If you cannot run BESTs algorithms (matlab) then you have to do the final QC yourself. Final QC would involve
checking the flags.txt file for every station and then looking
at the flag descriptions to ascertain if you want to apply that QC flag or not.
So, rather than just jumping into the middle of the data looking for answer to back up your speculation, the better approach is to understand the data you have. The processing steps it has been through. The processing steps that you are expected to make, and then build a hypothesis and test it. You will of necessity find anything you want to look for. A cooling airport?
yup. A warming rural site? yup. 36K stations is a large pile for cherry pickers. Look Mann was able to find a bristlecone pine to support his idea. Briffa was able to find a tree in Yamal. If you like that company, continue in your approach.
Dubuque County is only about 25×25 miles and has a density 10x what your data shows.
And Dubuque itself is only 4 miles away.
Yes, so if you want to use the population density for a city 4 miles away, you will have some physics to over come. but don’t let physics get in the way of your bias. press on.
Sorry, you are looking at county density. I’m looking at the density around the exact site.
As you should know the local zone around a thermometer is the most important. So, we look at the density where the thermometer is. Have a look at the google map where the station is located. Not de population bruce. Station move. See if you checked dubuque, you’d see
that the airport starts reporting in 1950.
When you look at county density you skew the numbers based on the urban center. I look at the exact density around the site. So if the
site is outside the city you dont impute city population to a residential cropland area.
Again, don’t be driven by your bias when you look at data. be driven by the data and you will find more interesting things.
Even if the city is 4 miles away, the plane traffic in and out of the airport is related to the size of the city as is the traffic that drives by the weather stations.
As for 1950 … BEST QC shows data in 1900.
It appears warming counties grew much, much faster than the country as a whole, while cooling counties grew slower than the country as a whole.
http://sunshinehours.wordpress.com/2012/03/17/county-population-statistics-and-coolingwarming-stations-since-1900/
Bruce you seem to have forgotten your argument and are now making stuff up.
Cooling stations: without looking at any station specific metadata you concluded that it must be de population or de urbanization. You left out the following possible explainations:
1. an undocumented station move
2. a documented, but uncorrected for station move
3. an instrument change
4. random chance.
So, we begin looking for a cause by first eliminating things likely to happen by chance. That means we have to screen the cooling trends by looking at those that are statistically significant. Its the same approach we use in warming stations or the warming globe. Same
method. That way our bias ( your bias ) doesnt play a role. The we start by look at de pop
ulation. The best measure of population around the site shows that it has only changed
by 2 people per sq km. Further its population is rural not urban. That matters. It matters
because when people live in high density places they change the surface: they build tall
buildings ( a cause of UHI ) they change the surface ( turning grass into concrete which
changes heat storage which drives UHI) they change the skyview. they change the surface roughness. Rural population doesnt change those things that cause UHI.
So, much for you de population hypothesis. However, what the record does show is this.
Its happens a few times in the us record. We’ve talked about it for 4-5 years. There are
ocassions where a thermometer can move from an urban setting to a “rural” setting.
For example, when a thermometer moves from the city to a outlying airport. This can
cause a discontinuity in the record. Why? because airports are not universally warmer.
In fact, airports can be ( but are not ALWAYS) good places to take measurements.
in fact in about 50% of UHI literature airports are used as rural pairs. Why are aiports
good
1. Skyview factor is 1 ( no tall building and radiative canyons
2. aspect ratio is good
3. Albedo is higher than cities, thats good for less UHI
4. Roughness factor is low. Airports are built in areas with long fetches.
a long fetch is good and combined with low building promotes convection
of heat away from the local surface.
5. Winds tend to be from a prevailing direction
6. heat storage is changed greatly. Look at the percentage of land that
is turned into concrete. very low compared to the 100% you see in cities.
The UHi effect starts to kick in at 25% of the surface transformed. basically,
by changing the surface you alter evapotranspiration.
7. Population is low: the residential population is low and the ambient population
fluctuates. high during the day, low at night.
Basically, you are too eager to find what you believe and dont want to listen to data.
“As you should know the local zone around a thermometer is the most important.”
Thats really an interesting position you’ve taken … a thermometer is only capable of being a temperature proxy for a square mile or less around it. Even with 40,000 or so total thermometers, that means we only know the temperature of 40,000 square miles of the earth surface and no more.
You could be right.
Its the most important for understanding the potential bias. As far as using a station to estimate the TREND, the governing factor is the spatial coherence of the temperature field.
The more coherent the field the fewer stations you need. If the field were in coherent you would get a different answer with 478 stations than you get with 36000. But you dont.
You dont because of the physics governing the distribution of heat. Again, you cannot learn if you dont open your mind. I suggest you do a bunch of reading. Start with physics and then
do some stats. Then read the “readmes” that come with the files. Download the matlab and
work your way through that. When I got hansens code that is what I did. Steve Mcintyre did
the same thing as have other people. When you put in the effort you will find that your focus
will shift to things you can prove. And then you can start to assemble arguments that have a basis in data and sound methods
Maybe you think any result that doesn’t confirm your CO2 theory is caused by flawed thermometers or random chance.
That holds for your theory too.
But I’m looking at 1000+ stations and the counties that cooled or warmed the least also grew the least (on average). The difference in growth rates is striking.
I’ll be looking at the next batch (1910). You are free to do your own calculations.
You always come back to the argument that increasing CO2 must warm the planet.
Quadrupling the population in the same area, along with all the other factors like more industrialization, more roads, cars, airplanes also MUST warm the planet.
I’ve just found evidence for it.
I wouldnt use population data at the county level. its far too course.
The effect of UHI falls off dramatically with actual distance from population centers.
You can see this if you look at LST data or any number of UHI studies
and don’t be too sure of your evidence. always double check and provide your code.
If you look in our UHI poster you will see that we also found an effect for changes in population. That’s nothing new. It’s also besides the point. The POINT is this.
1. is the effect large
2. is the effect uniform
3. is the effect statistically significant
4. how much bias does it add to the record.
Those are the questions of interest. Since 1960 the world population has doubled.
Compare two records:. the land record from 1979 to 2011. the UHA record.
The UHA record isnt bias by population. It measures temperature miles above the surface.
The slope of the UHA record and the land record are within .1C per decade. That
effectively puts an upper limit on all of the biases in the land record: UHI biases, spatial
sampling biases, bad thermometers, station moves, shaded stations, missing records
etc. Basically, there is evidence that a bias exists. That evidence suggest a small bias.
not zero. and not .5C. To understand the bias on a GLOBAL scale you have to look at
GLOBAL data. You can find UHI of around .1C per decade in japan, in china, in korea.
when you look EXCLUSIVELY at large cities. You can find .05C decade in the US,
again, looking exclusively at large cities. One of the reasons is the different pattern of
building you get in the US versus Asia. If you study the urban climate literature you’d
be miles ahead. Heck, even looking at Oke you’ll see that the role population plays
is modulated by : regional wind speed. wind direction, cloudiness, and the actual
region of the world. Such that figure you derive from the US and Europe do not
match those you will find in Asia or the southern hemisphere. That’s why finding about the US can’t be extrapolated to a GLOBAL answer. One hopes that you dont extrapolate
a US answer to the world. That would put you in the class of people ( say Hansen) who
you criticize for extrapolation 1200km.
1) UAH has the land today colder than 1980.
http://www.woodfortrees.org/plot/uah-land
It got colder after 1980 and it got warmer, but its colder than 1980 right now.
So when I trend from 1900 to 2011 and I get hundreds of US stations cooling, I think I’m right.
2) So far it looks like county population change does have a statistical relationship with cooling or warming.
3) “The effect of UHI falls off dramatically with actual distance from population centers.”
Not from the experimental evidence I’ve seen.
Then you should be surprise by the trend
http://www.woodfortrees.org/plot/uah-land/plot/uah-land/trend
its the TREND bruce. TREND.
Frankly, if you cant get that I would not trust any sort of statistical test you would do.
1. You cannot use all stations in the US from 1900 to 2011. You have to look at stations with complete records. If you have missing data the net result is a higher number of cooling stations.
2. You have to test the stations for statistical significance
3. you have to account for both documented and un documented changes. You’
ll need to look at the station site files to do that.
4. Have a look at any modis LST data and you will see the fall off. 100% certain of that
Click to access pkt258-59-6.0_UHI_newsletter_Issue_6.pdf
Click to access UHI_newsletter_Issue_5.pdf
Have a look at this one in particular and you can see the physical measurements.
Click to access UHI_newsletter_Issue_4.pdf
and you can look at brussels. Very nice map, interactive
http://www.leefmilieubrussel.be/Templates/Particuliers/Niveau2.aspx?id=3674&langtype=2067
And this nice picture which shows you how varied the phenomena is.
See those cool areas? guess what county population is the same for those cool zones
as for the warm zones..
http://earthobservatory.nasa.gov/IOTD/view.php?id=36227
When you have PHYSICAL MEASURMENTS that show that UHI does NOT vary with the distance from a population center, then you win a nobel and get some attention. Otherwise, you are just data mining and cherry picking.
The trend is towards many stations that are cooling.
http://sunshinehours.wordpress.com/2012/03/18/cooling-weather-stations-by-decade-from-1880-to-2000/
“When you have PHYSICAL MEASURMENTS that show that UHI does NOT vary with the distance from a population center”
“Using information from NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on the Terra satellite, Boston University, Boston, researchers discovered that city climates have a noticeable influence on plant growing seasons up to 10 kilometers (6 miles) away from a city’s edges. Growing seasons in 70 cities in eastern North America were about 15 days longer in urban areas compared to rural areas outside of a city’s influence.”
http://www.nasa.gov/centers/goddard/news/topstory/2004/0801uhigreen.html
Bruce. UHI also has a effect on precipitation within 20km downwind
VARY means changes with distance. That means is the UHI is say 2C in the city
it will VARY or CHANGE with distance.
Look at the LST images I pointed you at. CHANGING or DIMINISHING with distance.
Now, look at what they say. It has an “influence” up to 10km.
They dont say that it is a CONSTANT influence with distance. it VARIES with distance.
it changes with distance. So the effect is stronger at 1km than 6km
it VARIES with distance. get it. goes from HIGH to LOW to ZERO..
If it didnt VARY with distance the effect would go BEYOND 6km.
You didnt show that it was constant you showed the opposite with your example
Here: page 604
Click to access 23.pdf
See how it VARIES with the distance from the city. VARIES means changes.. diminishes the further you get away.
English. learn it
You are claiming Dubuque had no effect on the airport 4 miles away. I disagreed.
Since the county Dubuque is in is only 25m x 25m, county population may be a better proxy for temperature changes that say one square mile around a station.
The plot shows that when the station moved to the airport it cooled. If you want to argue that cool air from a city 4 miles away made its way to the airport, thats funny. If you want to argue that hot air from the city made its way 4 miles to the airport, I’ll remind you that hot air rises.
If you look at LST for the city and the airport, you’ll see your argumented busted.
Go get Modis LST. have fun
Do you have experimental evidence proving cities 4 miles away have zero effect?
Can you prove nobody drives their cars from the city to the airport?
What you have is inferences built on shoddy data.
I see bruce. Now the cars from the city are cooling the airport.
First your explaination was de population.
When I showed you that the series was actually made up of two series, an early part from the city and a later cooler part from the airport, you now conclude that cars driving from the city cool the airport, or is it that they warm the airport.
You realize of course that cars are not allowed out in the middle of the grass feild where the thermometer sits.
The point is to move the AVERAGE you have to do 1 of 2 things
A. Move TMAX up
B. Move Tmin up.
Now you could of course look at the literature on collecting temperatures next to roads and see how quickly the temperature impact falls off with distance from the road. You could, but you wont.
So, ya, I would say they cars driven from the city and parking have no measureable effect on T average.
Is there a dataset for “forest”?
The map pattern shown indicates to me a “re-forestation” hypothesis would be worth checking. Areas cut back in the 19th century (log cabins and farms and no towns) growing back around homes in towns and in parklands in the 20th.
Discussions of climate changes are, for me, much more interesting when they include anthropogenic changes such as soot, soil tillage, pavement and forest cover and don’t assume that CO2 is some sort of catch all explanation or proxy for all factors not active on Mars and Venus.
Yes there is a forest dataset. First of all one has to see if there is an effect worth explaining.
Natural Variation + UHI = Temperature.
You can postulate UHI = 0.
But UHI could still be a positive and natural variation could be cooling and there could still be a net cooling.
Or, UHI could have been higher in the past and with Dubuque’s economy stagnating and its population 5,000 lower than it was in 1980, the trend in UHI could be negative.
CO2 seems to cloud your mind.
I don’t postulate that UHI is 0. It’s positive in most cases, negative in others.
The point is finding it in the GLOBAL record is hard. In the next post I show you the beginning of understanding why the airport is cooler than the city.
You will see what the city does to sunshine. Funny. See, you started with bias. That bias means you dont go looking for things. Like, cloudiness over the city. And, you missed the river. You know what that does to nighttime temps? Guess.
Population? not that strong of a variable. To the extent that people BUILD things, it does matter. Its the building of things that changes the evapotranspiration. And that changes cloud formation, and that, changes sunshine. Your namesake, and you missed it.
Maybe the cars carry sunshine to the airport.
Oh I agree. More people in a county, more building, more change in climate.
But if a county stops growing, or as many US counties have down, shrinks in population, then cooling may occur.
It may well be that all warming/cooling is related to population growth.
First you need to look at the energy balance equation for the urban environment.
Lets’ review the causes of UHI
1. More absorption of SW radiation. You build buildings, those building provide
more surface area to absorb radiation. When people leave they dont take building with them.
2. Reduced skyview factor: Again, building related no population related.
3. reduced evapotranspriation: concrete, does nt leave with the people
4. reduced turbulant heat transfer. again building related.
5. reduced albedo. Cities have lower albedo. no population related.
6. Heat storage: building related
The ONLY factor in UHI that gets reduced with de population is Anthro heat. thats, heat from
buildings, industry, and traffic. In terms of Watts this amount is small compared to the other factors.
Mosher,
If the idea is to look at whether UHI or station moves are the most likely cause of cooling trends. Wouldn’t it be better to confine your search for cooling stations to those with a cooling trend during the satelite period? As I understand it, in the US (where the cooling stations seem mostly to be) we had a fairly hot 1920s and 1930s. It seems to me that a station move or observation change would be hardest to detect during the 1900 to 1940 period and if the record (or “good record”) started at a temp peak during the 1930s a “true” slight cooling trend might not be impossible in a particular area.
My first goal was to audit the claims that there were long series that had cooling trends. I’ve been looking at this problem off and on for some time. Check the posts here. In due course
I’ll look at it in more depth, for now I’m getting the methodologies and metadata in place.
That means doing work that others can replicate. So, first things first
Ok, I am not qualified to debate the subject and, obviously, you are doing the work, not me.
But it seems to me that the Midwestern US is one place in the whole world where you might find actual, honest-to-God, cooling stations compared to 1900. I think that if you were to graph non-adjusted US temps, you won’t see the strong trend you find in your top figure above. If the metadata for station moves and observation times are weak, you will have given yourself a very tough problem to solve.
Not sure why you say any of that. My experience has been that if you trace down “cooling” stations you find one of the following
1. in complete or spotty station data. You will find missing months and the distribution
of which month is missing will not be uniform.
2. Station records which are ACTUALLY the merged product of two different stations
3. Stations that have not been TOBS corrected.
4. lack of statistical significance. ie its not cooling
But the search goes on.
For most of what I said, I say it because the post-1900 US trend does not look like the post-1900 world trend you posted at the top. Or am I wrong?
I assume you put the world trend at the top of this post because you wanted to show how “odd” cooling stations must be compared to the overall trend. Fair enough, but if my impression is correct and the US trend is flatter than the world trend and the number of stations is much larger an actual cooling station is no longer odd at all. It is my impression (although I have not seen a graphic of this) that the Midwestern US and Southern Midwest is even flatter than the US as a whole. If there are any honest-to-God cooling stations, that’s where they would be.
And if you ran into one, what would you do? What some would do is adjust the thing until the problem went away. Some people would put these stations into one or more of you four categories no matter what the metadata says. Is that the right thing to do? Depends on how good your metadata is. How good is your metadata all the way back to 1900? Are you sure it is that good? What if you find a non-significant cooling trend in a station which is likely affected by UHI? Are you going to adjust yourself into a problem or do you only adjust your way out?
If you were limiting your search to cooling stations post-1979 I would agree that the potential for an actual cooling station seems remote and I assume your metadata will be better. That’s all.
“For most of what I said, I say it because the post-1900 US trend does not look like the post-1900 world trend you posted at the top. Or am I wrong?”
Probably wrong.
“I assume you put the world trend at the top of this post because you wanted to show how “odd” cooling stations must be compared to the overall trend. ”
No I post the global trend to show that this method matches others. Cooling stations DO NOT have to be compared to others. wrong again.
“Fair enough, but if my impression is correct and the US trend is flatter than the world trend and the number of stations is much larger an actual cooling station is no longer odd at all.”
Does not follow.
” It is my impression (although I have not seen a graphic of this) that the Midwestern US and Southern Midwest is even flatter than the US as a whole. If there are any honest-to-God cooling stations, that’s where they would be.”
actually if you expect them anywhere you expect them in the SH on coasts.
“And if you ran into one, what would you do? What some would do is adjust the thing until the problem went away.”
1. Check the source data.
2. Check the statistical significance.
3. Check for inhomogeniety.
4. Look for physical causes.
“Some people would put these stations into one or more of you four categories no matter what the metadata says. Is that the right thing to do? Depends on how good your metadata is. How good is your metadata all the way back to 1900? Are you sure it is that good? ”
1. depends on the metadata source
2. depends on the calibration.
3. Requires work. Not merely hand waving.
“What if you find a non-significant cooling trend in a station which is likely affected by UHI? Are you going to adjust yourself into a problem or do you only adjust your way out?”
1. If its not significant ( statistically) there is not much one can do to “explain” it. I flip a coin 100 times. It comes up heads 52 times. Explain why!
2. “likely” to be affected by UHI is problematic. The phenomena is very complex. See my early posts on the spatial variablity.
3. I don’t think adjusting data is useful approach. You can correct for some known things ( like TOBS) but I wouldnt suggest going beyond that.
“If you were limiting your search to cooling stations post-1979 I would agree that the potential for an actual cooling station seems remote and I assume your metadata will be better. That’s all.”
Again depends on the metadata and the region. US population data prior to 1979 is WAAAY better than russian population data from 1979 on.
1. I do not understand your assertion that it is “probably wrong” to say the US trend is lower than the World trend. It seems to me that this is not wrong but, if it is wrong, why only “probably” wrong? Please explain.
2. Of course, “cooling stations DO NOT have to be compared to others.” Obviously, nobody is forcing you to do anything. I assume you are looking at cooling stations for a reason, but even if you have no reason (or refuse to admit to any) the selection of cooling stations as the subject of your study inevitably involves a comparsion to others you did not select. You are not looking at all stations and you are not looking at randomly selected stations. Do you have some reason to believe station moves, TOB changes, merged records, only occur at cooling stations?
3. Maybe I just don’t understand. I thought you were looking at cooling stations in an effort to understand why these stations are different than other stations. For that purpose, it seems to me you could also look at “trendless” stations to find out why they are different from others or even look at “excessively warming” stations to find out why they are different from what is expected. The absence of a statistically significant trend needs to be “explained” for the same reason that a cooling trend needs to be explained. Because it is not what you expect. As you point, 52 out of 100 heads does not require further explanation provided we assume the coin is fair. If we believe the coin is not fair, it might be interesting to find out why we got 52 out of 100 as a result.
4. “US population data prior to 1979 is WAAAY better than russian population data from 1979 on.” How is Russian population data circa 1900-1950? Are you limiting your study to areas with good population data?
1. “probably” wrong? Please explain.
why is it only “probably wrong?” because most statements are probably right or probably wrong. Looing at trends for all continents ( in review ) you will not see that great a difference
between US and others. So your assumption that US trends are lower than global, is
“probably” wrong. Not certainly wrong. and it’s largely besides the point.
2. Of course, “cooling stations DO NOT have to be compared to others.” Obviously, nobody is forcing you to do anything. I assume you are looking at cooling stations for a reason, but even if you have no reason (or refuse to admit to any) the selection of cooling stations as the subject of your study inevitably involves a comparsion to others you did not select. You are not looking at all stations and you are not looking at randomly selected stations. Do you have some reason to believe station moves, TOB changes, merged records, only occur at cooling stations?
a) I’ve been looking at cooling stations for about 4 years. Also look at warming stations.
There is no inevitability in comparing them to others. It depends entirely on what your
purpose is. One might, if one chooses to. One might not for a whole host of reasons
3. Maybe I just don’t understand. I thought you were looking at cooling stations in an effort to understand why these stations are different than other stations. For that purpose, it seems to me you could also look at “trendless” stations to find out why they are different from others or even look at “excessively warming” stations to find out why they are different from what is expected. The absence of a statistically significant trend needs to be “explained” for the same reason that a cooling trend needs to be explained. Because it is not what you expect. As you point, 52 out of 100 heads does not require further explanation provided we assume the coin is fair. If we believe the coin is not fair, it might be interesting to find out why we got 52 out of 100 as a result.
Looks like you don’t understand. Sorry I can explain things for you but I cannot understand them for you. The process is called abduction. google CS pierce.
4. “US population data prior to 1979 is WAAAY better than russian population data from 1979 on.” How is Russian population data circa 1900-1950? Are you limiting your study to areas with good population data?
I’m not limiting my study areas with good population data. I’m answering your uninformed claim about US population data.
1. So I have looked at the trends and agree that there is “not much” difference but that is not the same thing as saying there is no difference or that the difference favors the world trend. Did you mean to say “probably right” and accidently used the wrong word?
2. Ok. But why did you select cooling stations to study? Give any substantive response at all and you prove my point. Deny that you decided to look at cooling stations and I will admit both defeat and utter incomprehension.
3. Good one. I think I understand abduction. I just don’t understand why a statistically significant cooling trend is important to your search for station moves, etc. At one time, I thought maybe it was because you were looking for bad data and thought cooling stations would be a good place to look (as opposed to looking at all stations or a random selection of stations). I now see that this was incredibly foolish and stupid thing to think. I am profoundly ashamed of myself. Feel like doing the explaining thing now?
4. What uninformed claim about US population data? I don’t recall making any claim about population data, uninformed or otherwise. Do you have me confused with somebody else?
1. you claimed the us was the one place in the world to look
as i said thats probably wrong. the sh tends to warm less than the nh. so id expect to find them in the sh. and on coasts. and in less urbanized areas. finding them in the us was surprising. finding them inland even more so. so that looked interesting.
arrg phone posting is hard
2. your point was that i needed to compare warming stations to cooling stations. i see no reason to given what i am doing. first and foremost i build tools. thats it. second i wanted to .revist a claim that 30% of the record was cooling. to see if that was true i check cooling stations. make sense? somebody makes a claim i check. peroid. that means auditing the data. back through the merge process. so now im working on distance tools. why? needed for the merge process.
3. simple. i wanted to investigate a claim. 30% of stations cool
to do that i build some tools and test old tools. i expected to find them by coasts and in the sh. surprise. you see some of that but many in farmland usa. so why? here is where 4 years of lookibg at this comes in handy. is it real or statiscal artefact? is it spotty data data with the same season missing. is it station moves? that means looking at the merge process. 9 times out of 10
odd results stem from screwed up data.
4. you made a categorical statement about metadata being better after 79. not true. us pop is one example. simple avoid making catehorical statements.
Ok.
But actually I did not claim you “should” compare warming stations to cooling stations. As far as I know, I don’t get to vote on what you should do. My original point was only that if you were hunting for bad data you might be looking in places where it would be difficult to tell whether the data was really bad or not. Of course, that does not mean that the data wouldn’t actually be bad, just that the evidence showing it was bad might be hard to get.
I agree that the cluster of cooling stations in the US Midwest is interesting and somewhat counter-intuitive. But we are only talking about a tiny portion of the Earth’s surface and the fact that there is a cluster rather than a uniform dispersal implies (to me) that a real cooling might have taken place. With the metadata issues, I don’t see how you can reach a really satisfactory conclusion. But maybe I’m wrong. I would like to be wrong about this, best of luck.
Categorical statements maybe important and are as likely to be true as other statements. I still say, categorically, that metadata from 1979 to present is going to be better than metadata from 1900 to present. I agree that Russian population data circa 1979 is very possibly worse than US population data circa 1900. But you are not limiting yourself to geographic areas where metadata (or population data) is good so that observation does not undermine the validity of the categorical statement above. To show that the statement is wrong, you must show, categorically, recent metadata is worse than 100+ year old data. I think that will be hard to do.
This data strongly suggests that temperature change is regional, and driven by industrialization rates. As the NE USA de-industrializes, temperatures are decreasing regionally.
The data also suggests that this effect has likely been hidden due to global averaging of temperatures, making it appear that CO2 was the cause, not simply an effect of industrialization.
The question then becomes, what is it about industrialization that drives temperature? It cannot be CO2 production, because CO2 is well mixed and thus we would not see regional results such as the NE USA.
i would say the data doesnt show that at all. part of the problem is allmost all of the cooling stations have a discontinuity. ie statiom move.
Earlier it was said;
‘Quadrupling the population in the same area, along with all the other factors like more industrialization, more roads, cars, airplanes also MUST warm the planet.’
Has anyone done the calculations for how many watts humans add to the energy budget just by being alive and through their increasing use of the various benefits of industrialisation, some of which were cited above?
Just a question, I’ve no idea how significant it would be.
In addition. when talking of depopulation being a possible cause of cooling, I wonder how much effect insulation of properties, double glazing etc has in keeping heat from escaping from a building and thereby cooling the local environment compared to the leaky buildings of yesteryear?
tonyb
yes tony there are calculations for the watts humans add. It’s a kid peeing in a pool. It might be local warm but compared to the entire pool it is lost in the noise. That is the funny thing about how people remember about the heat capacity of air when discussing C02 but forget about the heat capacity of air when discussing UHI. There are measurments of the anthropological contribution to UHI from a few places. The numbers are all over the map. One easy way to think about it is to look atthe warming trend in the Ocean and the warming trend in the Lower trop. UHI excess heat doesnt mix into the ocean. and it doesnt show up in the trop. Just by looking at the trend at the trop and the surface you get an Idea of the BOUNDARY for potential UHI bias.
Buildings. There is literature and studies about building effects and where heat is lost.
mostly from the roof. The roof layer is the upper bound of what is called the canopy layer UHI. above the roof is the boundary layer UHI. Thermometers are in the canopy layer.
heat that comes off the building typically goes up the side of the building. it falls off with
distance from the building.
On quadrupoing population: Its not the population. Its what you DO to the surface to hold that population. compare the temp over a parking lot to the temp over a parking lot with one person on it. 2 on it, 3 on it, etc. The transformation of the surface is permanent and much more powerful than the heat given off by the person.
take 1000 people. Put them in a skyscraper . Measure the delta temp.
take 1000 people spread them over miles of land in tents. Measure the delta temp.
That should clue you into the relationship between building practice and UHI. Its not
about population ( mostly) its about what is done to the surface FOR that population.
In any case. google is your friend
Mosh
Thanks for the interesting reply.
tonyb
“What we see should come as no surprise. The planet is getting warmer. The slope of the line is roughly .08C/ decade. The stations have the following geographical distribution:”
I do not see that the shown geographical distribution is “the planet”, so the statement “the planet is getting warmer” does not follow. The stations are very much those in highly populated and built-up areas.
I like the reasoning in the post, but this too easy assumption destroys it for me.
yes the distribution shown is not the planet it is a sample of the planet, Since the planet is warming it is no surprise to find a subsample warming. You are taking my given for a conclusion. Next, the statement that the stations are in highly populated built up areas is also not exactly accurate. 40% of stations are in unpopulated un built areas.
This means you can go find stations in the other 60% to try to prove that 40% doesnt exist, but it does.
Hey Steve,
If, as you claim, the US is warming along with the rest of the world, then how do you explain this?
If catastrophic global warming was actually a thing then we should absolutely see evidence of this in the all-time maximum temperature records
–yet ONLY TWO U.S. states have set heat records since the turn of the millennium i.e. After the year 2000.
Out of the 50 U.S. state record high temperatures, 23 were set during the 1930s, while 36 occurred prior to 1960.
https://electroverse.net/temperature-extremes-show-no-sign-of-global-warming/
Even, your mate Hansen, wondered aloud why the best climate network in the world i.e. the US, was cooling while the much lower quality and sparse global network was warming.
If he were a proper agenda-free scientist he would have suspected the low quality global network to be wrong, not the other way round.
QUOTE: “How can the absence of clear climate change in the United States be reconciled with continued reports of record global temperature? … Indeed, in the U.S. the warmest decade was the 1930s and the warmest year was 1934. Global temperature, in contrast, had passed 1930s values by 1980 and the world has warmed at a remarkable rate over the last 25 years.”
https://www.giss.nasa.gov/research/briefs/hansen_07/