The Earth's Temperature Since 1884

Wow, it seems like the yellows sometimes come up and than fade away again, like a natural flow, but that the real global warming problem really started kicking in only 10-15 years ago.

What is 'usual' defined as?

I'm assuming the GISS baseline period for the temperature anomalies, which off the top of my head is 1961 to 1990 or thereabouts.

It's interesting what the previous poster said about the real upturn in global temperatures happening in the last decade or so. I was always under the impression that the warming trend had been occuring for at least a century. It's curious that the planet was so cool in 1884. Was the planet undergoing a "little ice age" at the time? Could there be a link to global warming and popluation size? The global popluation in 1884 was less then a billion wasn't it? It's now approaching seven billion, more people, more factories, more cars...

> This streaming video reveals the temperatures of the
> Earth's surface since 1884.

It is visually impressive, but in fact bullshit. It displays changes at the *whole* earth surface, in the Arctic and Antarctica, in desert, oceans and the rain forest. May I sincerely ask how they can display the temperature distribution *there* ?
I don't know when IR satellites were firstly available, but my educated guess is from 70's to 80's. So how can they "display" the temperature anomaly from 1884 ?!
If done right, it should display the distribution of measuring stations (which aren't constant by the way) and their averaged values: red for hotter, green for colder and yellow for constant. White means no data. Then display in the 70's and 80's "satellite data" so that people know why suddenly the whole globe can be observed.

TSK,

There are many temperature proxies used to reconstruct past temperatures (ice isotopes, tree rings, marine sediment, etc., etc.). For starters, take a look at the National Academies report - all sections are linked for free from this page:

http://www.nap.edu/catalog.php?record_id=11676

TSK, if you wonder how they get the temperature data for those areas without satellites, you can feel free to read GISS' documentation. It's all public. They even show where all the surface stations are and have long-lasting datasets for each one of them -- for instance, in less than one minute, I was able to find three stations near the Sahara that have been operating since 1880.

Suffice it to say, GISS has a method of determining temperatures over a thousand kilometers away from other stations, which is why GISS is able to include the arctic in their analysis while the Hadley center doesn't. This method can be applied to earlier temperature measurements, such as those predating satellites (which were launched in 1979).

(Speaking of the satellites, GISS doesn't use satellite data -- it's just the instrumental temperature record. If you want the satellite record, you'll have to look at the UAH or RSS analyses.)

If you still distrust GISS, feel free to look at Hadley's. I don't have a pretty Hadley-based animation for you, though you can still look at the data yourself. Or the UAH/RSS satellite analyses. Just be advised that they use different zero points.

Or, you could let others do it for you: Tamino compared the three instrumental records here, and GISS/HadCRU/UAH/RSS over their period of overlap here.

a) The varves (marine sediments) are only acquired in
depths of 0-1000, max. 2000 m. As the ocean basins are
much deeper, there is no coverage of the oceans at all
(if we exclude islands). The time resolution is dependent
on the sedimentation rate; only few have annual
resolution, most sediments have resolutions of 10-50
years.

b) The tree ring proxies are concentrated in the Northern
hemisphere which is the reason that most papers display
the climate changes ("hockey stick") there. The southern
hemisphere is a problem: You need temperate climate for
good ring quality and you need temperature values to
calibrate your proxies. Both are abundant.

c) The number of ice core stations can be counted with your
fingers, they don't cover a big area. You must use ice
on continents (Greenland, Antarctica) because you need
the exact position. Pack ice in the polar sea drifts and
is therefore unusable.

d) It is very likely that the "Earth temperature" video is
in fact the 2007 GISS modelE of Hansen. It is a
computer *model*, not a data display.
http://pubs.giss.nasa.gov/abstracts/2007/Hansen_etal_3.html

Showing that as "earth temperature" is bullshit and bad science.

The upturn appears in more recent years partly because, as guthrie said, the base period is 1961-1990. The data shown are anomalies from that base.

TSK, you can read all about the NASA GISS global temperature data set here.

If I read correctly, GISS uses a 1200 km / 750 miles (!) smoothing filter to cover temperature gaps. The surface sea temperatures (SST) are given by HadISST1; its data in former years was acquired by ship measurements.
They interpolated over vast areas to cover the whole sea area, especially the Southern Hemisphere, because data is sparse. The ice field data from 1871-1939 is based on the 1929-1939 atlas climatologies. No joke.
Building high resolutions by interpolations and educated guesses is ok for checking climate model and give forcings and comparative values, but it is by no means a valid representation of temperature measurements. Duh.