Sunday, December 06, 2009

My Science Project

One of the hallmarks of science is the ability of others to repeat experiments. Not to create consensus for it's own sake or for force of argument, but rather to ensure that experiments are done correctly. With this in mind, I thought it would be interesting to look at temperature data recorded in my neighborhood, that is used in calculating global temperature means.

Example 1, Banff Station
51.2 N 115.6 W 403711220000 rural area 1888 - 2009
(I went to this exact spot)
Here is the official temperature data. Under NASA GISS protocols, this station is considered a rural station because it contains less than 10,000 people. Technically correct when the station began recording, Banff was a sleepy village visited by intrepid train travelers. Over the course of the century, the ski hills opened, the Trans-Canada highway opened and the town began to attract year-round visitors. The international airport in Calgary opened and funneled international tourists into the town as well. What started as horseback traffic at the turn of the century grew into 5 million visitors a year.

I don't know what the heat signature is from 5 million people. I can only imagine that 25 Saskatoons, in a small valley like Banff, generates heat. Whether it's hotel air conditioners or thousands and thousands of RVs generating and reflecting heat, or millions of automobiles racing around on new 4-lane superhighways and interchanges, the heat signature around the recorder is totally different than it was at the turn of the century. Is it possible to factor these changes out? There may be generalized computer models whose accuracy is determined by how well we understand all the variables. Since this knowledge is still imperfect, the models are highly suspect. Certainly no one has made even an attempt to "correct" the data for this particular micro-climate at this particular recording station (and currently, this may not even be possible.) It seems to me that there are enough questions regarding changing the environment of the data recorder to rule out constructing a temperature trend from this data. Despite the obvious problems, this station forms part of the global temperature data.

Example 2 Calgary INTL station
51.1 N 114.0 W 403718770000 593,000 1881 - 1990

If you look at this data, you will see an increase that begins around 1970. Savy Calgary residents can easily tell from this graph when the OPEC crisis happened! Thats when oil prices soared and Calgary's urban development boomed. A picture is worth a thousand words so below is a before and after of Calgary in 1969 and later in the year 2000.

Pictured above is Calgary in 1969 before massive city expansion.

Pictured above is Calgary in 2000. Note the heat sink now called the downtown core that has enveloped the Calgary Tower. For a top down view see the picture below. I've coloured in the city expansion in the last 40 years in blue. Notice how not only did the core become a massive heat sink, but the location of the data recorder became completely surrounded by development. I placed the location of the data recorder myself but with the coordinates given and Google Earth, you can easily do it yourself.

With what you have seen, it seems pretty obvious that separating increased heat from the local environment is very difficult indeed. Clearly the post 1970 temperature increase in the Calgary data is an effect of urban development. But how much? Some non- zero amount that is closer to 100 percent than zero percent.

As in the first example, it seems hard to believe, but this data is part of the data set that is used to declare that the global temperature is increasing.

I use these examples not because they are the two bad apples of climate data, but because they are the climate recorders who's local environments I can verify for myself. In fact there are many really bad data recording stations out there whose data has been compromised by local building regardless of whether they are considered urban or rural stations. For a collection of really bad stations and further links to a project to visit all the weather stations to investigate local conditions that may affect climate data, see here.


Mark said...

Excellent research and commentary on Calgary's growth. You should also note that in 1969, Calgary's airport terminal was at what is now the South Hangars on McKnight Blvd.

The urbanization of the areas to the west of Deerfoot and to the east of 32nd has had a signifcant impact on the ability of land to absorb rainfall and cool by evaporation. Nowasdays, rain runs right into the sewers from any man-scaped roofs, streets and driveways, and those same features collect and reflect heat. That is on top of any impact of carbon emiting cars and homes.

Best wishes.
Mark Ruthenberg

Kevin Aschim said...


What magnificent work! When research confirms common sense and rejects outlandishness, I always think of Occam's Theory of a Lack of Imagination!

Mark's point above is also very important in that it is not just anthropogenic generated heat that affects the profile but anthropogenic changes to albedo both in terms of surface quality, colour and hydrological heat absorption and removal that also drives surface heat profiles.

and then the_doctor said...

Good points Mark and Kevin, thanks for the input.

and then the_doctor said...

After looking through NOAA a bit, these data sets get even more mistreated. It turns out that one of the corrections made is to ADD .2 degrees to what has been recorded. NOAA claims that siting has improved on USHCN sites such that, since 1900, their locations average 0.2C cooler due to being near more grass and less asphalt today than in 1900.

"During this time, many sites were relocated from city locations to airports and from roof tops to grassy areas. This often resulted in cooler readings than were observed at the previous sites."-NOAA

As we have seen, this added increase appears not just arbitrary (bad enough) but just plain wrong. It also appears that the bulk of global warming is due to data fudging coupled with local effects.

enthemic said...

A quick google for "impact of urbanization on temperature ipcc" barfed up:

"Urbanisation is another kind of land-use change. This may affect the local wind climate through its influence on the surface roughness. It may also create a local climate substantially warmer than the surrounding countryside by the heat released by densely populated human settlements, by changes in evaporation characteristics and by modifying the outgoing long-wave radiation through interception by tall buildings etc. This is known as an "urban heat island". The influence on the regional climate may be noticeable but small. It may however have a significant influence on long instrumental temperature records from stations affected by expanding urbanisation. The consequences of this urbanisation effect for the global surface temperature record has been the subject of debate. It is discussed in Section 2.2.2 of Chapter 2."

Here's the link to the debate page:

Note: I wonder how frequently the do reviews to determine when stations should have their designation changed - and whether that designation is taken into account in the models. It's probably documented somewhere but that's more google fu than I have inclination.

enthemic said...

Hmm there's a wikipedia article on urban heat islands too...

From the "global warming" section of that article the ipcc take on urban warming seems to be to add uncertainty (0.006°C per decade) to the assessment due to future urbanization rather than to identify an uncertainty within the data sets themselves.

The same article indicates that there have been reviews of the clasification of weather stations.

It would be interesting to get a breakdown of the number of stations which have required reclassification annually, and the amount of error that reclassification produced - er "mitigated" ;). This would let you compare their predicted uncertainty against the historic error to see if it really fell within their 0.006°C per decade.

Kevin Aschim said...

Hey Dave,

Just a suggestion to edit or supplement your blog with links so that visitors can go directly to the Goddard Space Centre and do their own checks with their own local weather recorders.

starcat said...

Ha. Finally found it.

The graph on p3 is the important part here. Taking just the few surface stations that Watts etal (from deemed good or best, not adjusting them, and comparing them to the full adjusted data set, we discover that they are basically the same.
Gave me confidence that the people studying UHI/siting difficulties actually know what they are doing.

Not to dissuade you from your project, though. I actually really enjoyed this post, as a fellow local data analyzer.

and then the_doctor said...

You make a good point Starcat about the people working on the surfaceproject. The photos they've taken speak for themselves though.

Actually they aren't about the same. The adjusted data set for the entire US shows a .6 degree increase and the unadjusted is .1 degrees

most of the increase is due to data adjustments. Thats a pretty awful signal to noise ratio! lol The noise is larger than the signal!

NOAA explains the adjustments here:
What bothers me about the adjustments is that the theoretical impacts they attribute to the stations. NOAA didn't actually inspect the stations to come up with their values. This oversight inspired the inspection project in the first place.