Confirmation of the Green House Effect

The Effect of CO2 on IR
The Setup
Volume of trapped air in Vacuum Flask 541ml
Air is in a vacuum thermal flask made of double skin stainless steel designed for isolation of liquid from external temperature influence

IR Absorber used to collect IR is 18g matt black anodised corrugated aluminium rectangular block 40mm by 44mm
Distance from end of tube to top of absorber is 128mm

IR Transmitter is 49mm diameter matt black aluminium
Thermal conduction isolation provide by 135mm tube with internal convection cooling internal diameter 49mm

Air exchange isolation provided by  low density polyethylene (LDPE “cling film”)
Test Set up

1       Temperature recorded at 1 second intervals using an eight channel USB thermocouple interface
2       Apparatus set up as shown but with the IR transmitter isolated from the system. The Internal volume of the vacuum flask was filled with room temperature air (RH approx 50%). The flask then had 60ml of CO2 injected (3 lots of 20ml) (This increases the concentration of CO2 but not to the full 60ml as each injection would displace some of the already injected CO2).
3       IR Transmitter maintained at greater than 100C whilst thermocouple temperatures stabilised
4       When stability reached IR transmitter cooled to 101C and placed on top of tube
5       Temperature of IR transmitter maintained as stable as possible at approximately 100°C during the heating of the gas to greater than 25°C
6       The IR source was removed and the vacuum flask air replaced with room air (using a small fan).
7       The system was then cooled by placing ice cubes in a glass on the LDPE film on the vacuum flask.
8       The system was then reassembled but without the IR source.
9       The air temperature was the allowed to stabilise.
9       The test method was repeated until the air temperature was above 25°C
10     The Air in the flask was then enhanced with CO2 and cooled again, repeating the same method as above

11     The same method was also run (with CO2 and normal done in reverse order) on a previous occasion. 
The Results:
1       Using the last 2 runs (room air then CO2 enhanced room air) and measuring the slope of temperature rise per second at around 25°C shows a significant increase with more CO2
2       Also the temperature array record a smaller spread with increased CO2.

CO2 in these two instances caused between 6 and 10% greater heating rate
The temperature gradient of the "air" is the reverse of what would be expected with convection or radiation (the higer up the flask the probe the cooler the air)

Radiant emmittance: 1097.64 W/m2
Radiance: 349.389 W/m2/sr
Peak spectral radiance: 29.5722 W/m2/sr/µm
Wavelength of peak: 7.76877 µm

Radiant emmittance: 447.186 W/m2
Radiance: 142.344 W/m2/sr
Peak spectral radiance: 9.62545 W/m2/sr/µm
Wavelength of peak: 9.724 µm

100C source is 135mm+125mm-45mm from top thermocouple
top thermocouple is 45 mm from re-radiator
The heating effect of the top 100C source is therefore reduced by 45^2/ 215^2

The green trace below is radiation from the 101C  source.
The Blue trace is the upward long wave radiation from the 25C collector.

CO2 3 times more absorption at 4um than 15um
BB radiation curve shows that 4um absorption and 16um absorption are similar when sensitivity is considered

101C curve always adds less energy to CO2 than 25C re-radiator at 45mm Sensor. The difference will increase as the sensor distance from the re-radiator decreseas


  1. fail. you should have eliminated water vapour as a factor.

  2. agreed
    but the air before additional co2 has h2o vapour and the added co2 has h2o vapour. Without measurements of humidity before and after addition of co2 it is difficult assigning the effect to h2o co2 or h20+co2.

    However, this was uk autumn so humidity would be reasonably high naturally, the added co2 had some h2o (but not a lot) so the main change was to c02 concentration but not h20. I would therefore say the effect measured can be mainly ascribed to either measurement errors or increased co2.

    A more controlled experiment is needed!