As readers may know, Dr. Roy Spencer and I have had a long running disagreement with the group known as “Principia Scientific International” aka the Sky Dragon Slayers after the title of their book. While I think these people mean well, they tend to ignore real world measurements in favor of self-deduced science. They claim on their web page that “the Greenhouse gas effect is bogus” and thus ignore many measurements of IR absorptivity in the atmosphere which show that it is indeed a real effect. Rational climate skeptics acknowledge that the greenhouse effect exists and functions in Earth’s atmosphere, but that an accelerated greenhouse effect due to increased CO2 emissions doesn’t rise to the level of alarm being portrayed. Yes, there’s an effect, but as recent climate sensitivity studies show, it isn’t as problematic as it is made out to be.
I don’t plan to get into that issue in this thread, as this is an hands-on experiment showing one of the thermal premises of the “slayers” in action to prove or disprove it. Most of what that group does is to spin sciencey sounding theories and pal reviewed papers by a mysterious members-only peer review system, and I have yet to any one of them try to do anything at an experimental/empirical measurement level to back up the sort of claims they make.
What started the recent row was an essay by Dr. Spencer titled Time for the Slayers to Put Up or Shut Up, which I followed on with: The Spencer Challenge to Slayers/Principia.
In their response to Dr. Spencer, they made this essay…
…and in that response was this curious graphic from Dr. Alan Siddons:
To be honest, I laughed when I saw this, because for all their claims to be “experts” on thermodynamics while telling the world that “back radiation” has no effect, this is a clear-cut case of them not knowing what they are talking about when it comes to heat -vs- visible light. Clearly, you can indeed reflect/re-emit a portion of the visible and infrared energy back to the light bulb, energy which would have been lost to the dark surroundings. There is no “extra” energy per se, just a spatial redistribution of energy (a greenhouse atmosphere has higher temperatures near the surface, but lower temperatures at high altitudes). They also seem to fail to understand how a mirror actually works, bold mine:
“Does shining a flashlight at a mirror so that all the radiation comes back to the flashlight make the flashlight shine brighter?”
While the emissivity of a glass mirror is high, no mirror reflects 100%, and mirrors of course are not lossless, so it will also absorb some Visible and IR in addition to reflecting/re-emitting some of it back. You can see this loss of energy in the FLIR camera in the video just before the mirror is removed at about 16:30.
I put their claim of “a light bulb facing a mirror does not heat up” to the experimental test.
I did several spot experiments at home over the last couple of weeks to investigate the issue empirically (since talk is cheap), and to make sure it was repeatable, while discussing the design and results with Dr. Spencer. The first two designs of the experiment had weaknesses that I was not happy with, and so it has take time to devise an experiment in a way that was fully comprehensive and uninterrupted from start to finish. For example, in my first iteration, the experiment was shot from the side (similar to the diagram), but required rotating the bulb mount assembly away from the mirror to get the temperature of the bulb surface. This wasn’t always repeatable to get the same spot on the bulb surface and it introduced variances. Another problem was that standard household bulbs had odd temperature gradients across their surface due to the way the filament is placed. The flood lamp was much more repeatable at its center. Repeatability is important, because I want others to be able to replicate this experiment without significant variances due to the equipment and how it is setup.
After ensuring the experiment works, and is repeatable/replicable, and that the control run without a mirror performed as expected, today in this WUWT-TV segment, I present the entire experiment uninterrupted as one long video. It is almost 21 minutes long, but I had no choice, because at least 16 minutes of it were required to be non interrupted to show the experiment in progress. I didn’t want anyone to be able make silly claims that the experiment was faked that there were video edits going on to change the results, such as Al Gore did in his Climate 101 video.
In my case, I did some graphic overlays to illustrate points and data, but there was no discontinuity edits of the video or audio from start to finish.
Here’s the experiment equipment list and procedure.
Equipment:
- FLIR BCAM portable infrared camera
- 65 watt incandescent flood lamp (used due to mostly flat center target surface)
- clamp on ceramic lamp base and metal electrical base/stand
- small glass wall mirror from K-Mart
- video camera to record the event
Procedure:
- Setup equipment in similar fashion to Alan Siddons figure 3 above, using stands and clamps to allow for correct height and continuous recording of FLIR camera image and a timer image.
- Focus FLIR on flat front surface of 65 watt bulb
- Start video camera to record experiment, simultaneously start digital timer
- Apply AC electrical power to 65 watt bulb
- Note FLIR temperature of bulb center surface at intervals, record that data.
- Run until equilibrium temperature is reached, which I defined would be when temperature no longer increases after a period of about 60 seconds, note that temperature, note how long that takes with timer. Record that data.
- Leaving all equipment in place and operating, place mirror perpendicular to 65 watt bulb surface, at about 3 inches away to fit scale of Alan Siddons Figure 3. This will obscure surface of bulb from FLIR camera but is required so that distance/position between bulb and FLIR is not changed, which could result in altered readings.
- Continue experiment.
- Show with video camera how equipment remains in place.
- Wait for the same amount of time as previous equilibrium temperature took to reach.
- Remove mirror, note on the FLIR camera what the surface temperature of the 65 watt light bulb is at that time.
Premise of the experiment:
If the temperature recorded by the FLIR camera is the same after the mirror has been left in place for the amount of time that it took to reach equilibrium temperature, then the Principia/Slayers claim is true.
If the temperature has risen, it falsifies their premise that “a light bulb facing a mirror does not heat up”.
Video of the experiment (with conclusion) :
Note that this is not a big budget production (it was done in the dining room of my home) so I apologize for less than perfect audio quality. BTW, the clothes iron I used as a prop was not turned on, which is plainly evident in the FLIR image. It just so happend that the tabletop ironing board and iron worked out well to position the mirror…. and I had no budget beyond a few dollars for light bulbs and lamp bases. Where’s that big oil check when we need it? /sarc
Plotted temperature data:
[Note: per a suggestion in comments, this graph was updated to show the data after the “mirror added” as dashed line, since only one datapoint (228F) was measured. – Anthony]
Supplemental information:
In a PDF file here: Slayers_lightbulb_experiment
- Temperature data recorded from the experiment to reach equilibrium temperature
- Graph of the data recorded from the experiment showing data including after removal of mirror.
- I also ran a separate control experiment for 2x of the tested equilibrium temperature time to see if bulb can reach same temperature without mirror. I’m satisfied that the experiment is properly functioning.
I have another experiment planned for part 2 that will test another claim that the Principia/Slayers routinely make. I’ll have that in a few days.
UPDATE: In the claim by Joe Postma at Principia where they stated a couple of days ago that we’d “cut and run” (obviously not, just taking our time to be careful) Alan Siddons makes this claim:
As PSI’s Alan Siddons laments:
“All of us on our side have researched and deeply pondered the actual principles of radiative heat transfer. On the other side, however, the “experts” we argue with, like Spencer, Lindzen, Monckton, Watts, just insist that a body’s radiant energy can be doubled by directing that energy back to it — even though the simplest of experiments will shows that this is false.
I’ve never made a doubling claim like that, nor am I aware that any of the others named have claimed a doubling, only that some energy will be returned, as I have just proven in the “simplest” mirror experiment postulated by Siddons.
I have to think these folks aren’t operating with a full understanding of what the physical basis is when I read things like this. This is an excerpt of this comment left in the thread below by Joe Olson where he confuses a microbolometer with doppler radar:
“Remote read IR thermometers are also used to ‘explain’ this back-radiation warming effect. These instruments work be sending out an IR signal and measuring the shift in the returned signal. ” (bold mine -A)
No, sorry, you are 100% wrong. it is a passive sensing device. No active signal is emitted.
FIGURE 1. One pixel in a microbolometer array. An infrared-absorbing surface is elevated above the substrate and thermally isolated from adjacent pixels. Low mass increases the temperature change from heat absorption. Read-out circuits typically are in the base layer, which may be coated with a reflective material to reflect transmitted IR and increase absorption of the pixel. http://www.laserfocusworld.com/articles/print/volume-48/issue-04/features/microbolometer-arrays-enable-uncooled-infrared-camera.html
Gosh, I didn’t think your misunderstanding of an IR bolometer was that distorted. No wonder you guys make the sort of way out claims you do.
A microbolometer is a specific type of bolometer used as a detector in a thermal camera. It is a grid of vanadium oxide or amorphous silicon heat sensors atop a corresponding grid of silicon. Infrared radiation from a specific range of wavelengths strikes the vanadium oxide and changes its electrical resistance. This resistance change is measured and processed into temperatures which can be represented graphically. Source: http://en.wikipedia.org/wiki/Microbolometer
You should really quit while you can Joe, you are making a fool of yourself when you make such claims that are so easily disproved. – Anthony
UPDATE3: The Principia/Slayers group has posted a hilarious rebuttal here:
http://principia-scientific.org/supportnews/latest-news/210-why-did-anthony-watts-pull-a-bait-and-switch.html
Per my suggestion, they have also enabled comments. You can go discuss it all there. – Anthony
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I wonder why flashlight manufacturers put a mirrored surface behind the bulb?
A WUWT-TV movie without Benji?
REPLY: Kenji was in the room, observing while maintaining a look of “why are you wasting your time doing this when we could be chasing balls?”. Given responses from Slayers so far, he knows better than I – Anthony
Anthony, interesting experiment. I see you too are part of the 97% of the real consensus (as opposed to the SkS manufactured consensus). 😉
I think there’s a potential confound here though between what you’re visualizing and what may be happening.
Here’s my take:
An IR bulb produces most of its radiation in the infrared. Mirrors are not very reflective very far into the infrared, so most of that infrared is getting absorbed and is heating up the mirror, which in turn is radiating infrared, part of which is heating up the bulb.
You can test for the importance of this effect by putting a second bulb (not plugged in) an equal distance from the backside of the mirror. It’d be interesting to see what happens if you replace the mirror with a black surface.
Both mechanisms (scattering and reemission) play a role in the inappropriately named GHG effect, so the function is still served in terms of refuting slayer quackery.
Nice experiment, Anthony. I noticed the FLIR instantly returned to room temp when the mirror was placed in front of the bulb, and it briefly hit the high 80’s while your arm was in front of it.
I can’t personally testify that there were no camera tricks, however. I skipped over both equilibration segments. Sorry. 🙁
REPLY: No tricks, this experiment is fully replicable as you see it – Anthony
What if, that mirror reflects back to, that is, it reflects the energy from the source back at the source?
What of several mirrors? Each one reflecting energy back at the source? Thinking now of, say a Tungsten filament in a bulb, in a vacuum, with a fixed amount of energy per unit of time (Watts) applied via a current through said filament.
Oh, never mind,. I see. It *will* make it cooler. I see the error in my thinking … /s
.
By putting the mirror in relatively close proximity to the heat source you would be restricting the air circulating around it.This would increase the temperature of the source. Nothing to do with ‘Back radiation’
REPLY: LOL! Complain to the slayers then, they designed the experiment and published it first as a diagram. I simply followed their example premise, bulb was ~3 inches away, like in their diasgram. Heated air rises, no constriction. – Anthony
Chico still has a K-Mart?
How does it stay in business with its nuclear weapon components aisle empty?
OK, consider this: Your experiment is measuring the temperature of WHAT? The only significant temperature here is that of the actual resistance wire heater in the IR lamp, usually around 1,100 F. Did the wire get warmer when the mirror was put in place? I guarantee it did not. To determine this you would need a Fluke, measure current and voltage, see if the IsquaredR changed.
My original post was not off-topic, as the physics of the Greenhouse Effect are in question here, and most on this site have profound misunderstanding of the basics.
REPLY: And yet, in the diagram proposed in the essay by Postma, such fine details were not mentioned nor required. Demanding them now post facto doesn’t fly – Anthony
Isn’t using an incandescent lamp bad for the climate?
Why have you snipped the comment by Michael Moon above? He was perfectly on topic discussing the theoretical aspect of the experiment. In addition what he (correctly) said can be calculated relatively easily. Censorship is very bad.
[it is getting too far off topic, he’s welcome to resubmit being more on-point about the experiment itself -mod]
Carrick says:
May 27, 2013 at 9:30 am
…
Here’s my take:
An IR bulb produces most of its radiation in the infrared. Mirrors are not very reflective very far into the infrared, so most of that infrared is getting absorbed and is heating up the mirror, which in turn is radiating infrared, part of which is heating up the bulb.
—————-
Interesting point, but doesn’t this also refute the slayer position? Reflection or back radiation, I thought their position was that cooler objects could not cause warmer objects to become warmer still.
Michael Moon says:
May 27, 2013 at 9:16 am
Heat transfer is always from warm to cool, proven by centuries of experience with steam in utility boilers.
Michael, that is incorrect. Heat is transferred in all directions. Sure, net transference is always from the warmer to the cooler, i.e. as shown in Anthony’s experiment where we know there is of course more energy moving from the bulb to the mirror than vice versa. But there is still some energy being radiated back from the mirror to cause the bulb to be even hotter.
Or did you miss that part of the video? Skip to the end if time is short, and watch.
Ha! One of those little Kill-A-Watt meters appliance power ‘meters’ (they measure apparent and true/RMS quantities as well) should indicate this …
I suggest a slight variation to the experiment. Continue the experiment past 1000 seconds at the removal of the mirror. The temperature as measured by the FLIR ought to decline towards the 210 degree level when the mirror was put in place.
REPLY: Yes, you can see the beginning of that after the mirror was removed. And I comment on it in the video – Anthony
[REPLY: then do that experiment and report back, we are discussing this one – Anthony]
Oh well, the slayers are becoming a distraction, they did bring up some good points though. What remains is that the Earth’s surface is predominantly cooled non-radiatively (evaporation and convection) and most of the energy is transferred to the atmosphere, not the space. The atmosphere of course can lose this gained energy only by LWIR to space. The bottleneck is this atmospheric radiation, not the very efficient surface cooling. On the face of it, even a net CO2 cooling effect cannot be ruled out. But the Null Hypothesis is NO effect of the trace CO2.
Bravo Mr. Watts! When I looked at the light bulb mirror claim Figure 3, I thought that no one could be so foolish as to set themselves up with so asinine a claim. I don’t see how they can recover from such a blunder. The best that can be done is to admit error on at least this point and maybe blame it on an unpaid summer intern. It also confirms an observation that I have made over the years which is that few people understand radiant heat. There is an area where I work that has a long hallway banked on one side with very large windows facing roughly north. In the winter this area is very cold. Clearly the architect did not understand radiant heat loss nor does any administration since. I tell them to cover the windows with aluminum foil (a cheap solution) but I only get blank stares. This works in summer as well for south facing windows and I used this in an apartment I lived in years ago. It’s like magic! A more direct example of the light bulb figure is the tin reflectors placed in front of open hearth ovens to aid in roasting a section of beef. This is technology going back to, at least, the middle ages. The PSI is well and truly cooked.
Joseph E Postma says: “Anthony, I don’t need to repeat the experiment because the experiment is incorrect, and so I’m just going to state why it is so.”
So where is your stated “why it is”? All I see in your comment here is a bunch of yapping. Absolutely nothing to back up your statement that Watt’s experiment is incorrect.
Have not been able to watch the whole video yet. Agree with your approach. One real experiment is worth any number of thought experiments.
Your results are different to the result I achieved with a different but similar experiment, I did not get any additional increase in temperature when I used a reflector around the sides of an insulated tin which was exposed to the sun at the end. Therefore I will try your experiment.
I think your experiment would be better if you were measuring the temperature of the filament, not the glass surface of he bulb. You are not trying to control convective or conductive heat loss in any way, and the mirror might have an effect on that, since it party blocks air flow around the bulb
Peter Champness
Anthony,
Wouldn’t a proper control for this experiment be a non reflective surfaced board of some sort in place of the mirror? Maybe just tape black paper over the face of the mirror. If the experiment was done again with the “control” board in place of the mirror we could see if there was still an incread in temperature, and if there was if the increase in temperature was due to the reflective nature of the mirror, or just due to to its obstructive nature. Is it really the mirror that raises the temperature?
I had some friends, years ago, that got a busy little business going selling what were essentially tin foil liners to people for their attics on the basis that these would act as insulating agents summer and winter. They had a cute little doll house that they used to demonstrate the thing by shining a light on the house and observing a thermometer in the living room. When they introduced the tinfoil the temp went down behind the tin foil (for a while) – that was the trick that was eventually discovered and undid their little enterprise. So they moved on to some other scam. I think it was health food related.
The experiment is basically correct. However, the source size is so small compared to the spacing, that the effect would be difficult to accurately measure, and also due to the fact that the source is not a constant input power device (voltage is fixed, but resistance changes slightly at higher temperature). Blocking some of the convection is also a source of the possible errors. The heating of the bulb also enters in the balance, complicating the experiment. You need a cleaner experiment.
However, anyone that understands radiation physics knows the effect you are striving to demonstrate is valid. The absorbed back radiation results in reduced direct radiation in the direction of the blockage, and thus slows energy loss (from a constant input power source) driving the source goes to a higher temperature. It is analogous to covering part of the radiating area with a layer of thermal insulation, which would produce exactly the same effect. i.e., the temperature would go up slightly, but the radiating effective area would go down slightly, and TOTAL power out would remain constant.
People commenting on lack of conservation of energy are just showing lack of understanding.
Joseph E Postma says: “The problem is that you don’t understand the data, and what you’re actually supposed to measure, and what you’re measuring.”
If all you have to say is We be right and You be wrong and We understand the data and You don’t; then you have zero credibility. I have looked at PSI/Slayer site and find it to be severely lacking in facts and science, being vague on PSI/Slayer reasoning, and requiring faith to accept PSI/Slayer position. Now where have we seen this before? Hmmm…
There does seem to be theoretical & even some experimental support for the hypothesis that increasing CO2 will measurably warm the lower atmosphere, all other factors being held equal, if only by a trivial amount (by slowing the rate of reflected radiation out to space). But all other factors cannot be held equal. An issue is therefore is whether net feedback effects be positive or negative. Under some conditions in some areas of the globe, the net effect of increasing CO2 could be cooling rather than warming, even if the effect globally might be to warm.
As for the man-made contribution, humans also add cooling agents to the air, so it’s possible that the net effect of peoples’ activities may at times be to cool the planet, IMO.
In any case, Antarctic glaciation commenced with CO2 concentration possibly as high as 1500 ppm, & most likely no lower than 750, at a time when solar radiation was practically as strong as now. So if humans decide we don’t want the present major ice sheets (Greenland & Antarctica) to start melting before the next glacial phase begins (when North American, European & Siberian ice sheets spread again & glaciers advance in the SH), the “safe” level of carbon dioxide isn’t 350 ppm but perhaps 700.
As I’ve noted before, 1000 ppm, as in real greenhouses, can give some people headaches, but IMO even burning all accessible fossil fuels over the next few hundred years wouldn’t by itself achieve that level. Humans, with our high oxygen need, would suffocate at Cambrian & Ordovician levels of 7-8000 ppm. As you know, there was a glacial Ice House phase in the Ordovician, with CO2 possibly 20 times higher than now, & the sun only some four percent weaker.
I’m part way through the experiment–why is the left side of the mirror consistently hotter than the right side?
REPLY: Perhaps the iron (used as a prop for the mirror) wasn’t exactly flush with the surface. Look at the IR pattern on the FLIR after the iron is removed, I’ve added it here:
http://wattsupwiththat.files.wordpress.com/2013/05/post_iron-removal_capture.jpg
– Anthony
@Anthony: you can see the beginning of [temp decline] after the mirror was removed. And I comment on it in the video – Anthony
Indeed you do: Another nail in the coffin.
Mirror removed at 16:35 –
228 at 16:35
226 at 16:43 (8 sec)
224 at 16:52 (9 sec)
222 at 17:04 (12 sec)
220 at 17:15 (11 sec) last frame.
At this point, I would just add a note under the chart:
“After mirror removed, temperature dropped 8 degrees in 40 seconds at end of video.”