Just came across this:
http://www.kickstarter.com/projects/andyrawson/ir-blue-thermal-imaging-smartphone-accessory

It's a 4x16 pixel that connects to a smartphone and gives a thermal overlay on the smartphones camera.  This appeals to my inner geek!  The field of view seems to suggest (from the above website) that:
1m = 65x65mm per pixel
10m = 65cmx65cm per pixel

Extrapolating this to 1km would be a 65m square that you get temperature information for.  So the question is: would this be any use for trying to locate entrances on expedition by walking up valley sides and looking at the opposite side and trying to find the cold bit?  How big is the cold bit around an entrance?

Of course the answer to all these questions is for me to buy the DIY kit for $145.....

the 'fiel of view' you mention there is for each temperature zone/sensor or the 64 on the device, so you would get the average temp for and area 65meters square at 1kilometer. If the enterance was big enough for that device then you would be able to see it anyway! (well thats how I read the article anyway)

You could also modify an old digital camera by removing the internal IR filter and adding a visible light filter.
You should be able to find some how-to guides online.
Try searching "DIY Thermal IR DSLR"
or check this website

http://www.lifepixel.com/tutorials/infrared-diy-tutorials

Warning - this is very serious DIY and may result in a dead or non-focusable camera.
Lifepixel can do the conversion for you.
Visible light filters can be as simple as exposed film or you can buy from lifepixel.

Interesting.

As it happens, I have a 'proper' TI camera and have wondered about using it to look in caves for 'cold' spots.

However....

As I'm sure you know, TI cameras pick up Infra Red wavelenths of radiation that are being raditated from the object into the camera.
So 'drafts' are actually quite tricky to pick up, as the moving air doesn't really radiate.  Easy to feel with a wet cheek.

the emmisivity of an object plays a big part of how well its picked up on a camera too.  So a shiny copper pipe at 100 deg C, in a cold garage will appear to be as cold at the garage since its hardly radiating any energy, and actually reflecting the surrounding temperature.

This does give rise to some very miss leading results!!

The temperature differentials will also be quite narrow I imagine so I'd imagine you want a fairly sensitive, hi res camera for this kind of project.

I might, however take mine out some time and experiment with it to see if it picks up known entrances.  Have to wait for either a very warm day, and look for cold spots, or a very cold day and look for warm spots (which is probably the best way round).

I recall (from deep in my memory) someone using infra-red photographic film to look for warm spots during cold weather to identify cave entrances.

I'm sure that there is an article on this in CRG Trans but as I can remember neither author, title, nor year it makes the search a bit tricky. I think it does mention that fact that Valley Entrance stood out as a 'hot spot' when blanketed in snow.

Dr Farrant was trying to arrange a flight of a high definition thermal imaging camera over Llangattock/Llangynidr but it kept getting cancelled and then the kit went back to N. America for use there.  I've played around with military spec kit  but it was problematic getting permission to take it outside of the locked laboratory where it was supposed to be kept so as not to fall into the hands of bad people - I did use it to check the insulation on my house though but since it was mains powered I needed to run an extension lead outside.  I have ready access to industrial thermal imaging cameras but I suspect that the resolution is not up to it (both spatial and thermal).  I'll see if I can borrow it overnight and report back...

kdxn said:

You could also modify an old digital camera by removing the internal IR filter and adding a visible light filter.
You should be able to find some how-to guides online.
Try searching "DIY Thermal IR DSLR"

Click to expand...

This give access to the near IR region (Si won't detect much beyond 1.3um and is most sensitive below 1um).  The sensor used in the above (MLX90614) looks at wavelengths from 5.5 to 14 um.

SamT said:

I might, however take mine out some time and experiment with it to see if it picks up known entrances.  Have to wait for either a very warm day, and look for cold spots, or a very cold day and look for warm spots (which is probably the best way round).

Click to expand...

This was how I envisaged it being used - I have done walking around in the snow looking for blow holes but thought this may be a quicker way.  R.e. the emissivity - I assumed that given the same vegetation the emissivity should be comparable.

Duncan Price said:

I have ready access to industrial thermal imaging cameras but I suspect that the resolution is not up to it (both spatial and thermal).  I'll see if I can borrow it overnight and report back...

Click to expand...

That would be interesting

Antwan said:

the 'fiel of view' you mention there is for each temperature zone/sensor or the 64 on the device, so you would get the average temp for and area 65meters square at 1kilometer. If the enterance was big enough for that device then you would be able to see it anyway! (well thats how I read the article anyway)

Click to expand...

That is my understanding.  I reckon you will get is an averaged temperature for each pixel.

graham said:

I'm sure that there is an article on this in CRG Trans but as I can remember neither author, title, nor year it makes the search a bit tricky. I think it does mention that fact that Valley Entrance stood out as a 'hot spot' when blanketed in snow.

Click to expand...

I have found the below from CREG, but entering all the terms I can think of into the online index doesn't show anything else - may have to wait for me to get down to the library at SWCC.

Lyles John 2007 The Technology of Infrared Imaging No. 66 pp 16-20
The author describes the principles behind far-infrared imaging, reviews some commercially-available devices and answers the question "Can I build my own thermal imaging camera?". In a follow-up article he will explain how to use infrared imaging for cave detection and bat studies.
Lyles John 2007 Thermal Imaging Cave Detection No. 67 pp 8-12
In this second of two articles, the author examines the effectiveness of thermal imaging techniques for cave detection and illustrates their use in monitoring bat behaviour.

Took the camera out last night (a Flir i50) just to see what we could see on our way to Moorfurlong (a mine near bradwell).

Snow was on the ground, crisp cold clear night.  On the walk across the fields - not much showed up.  areas without snow appeard to be be slightly warmer than those with snow, as you'd expect.

Got the metal lid of the cave, which was clear of snow, and this is what we saw on the camera.



Not much obvious really..

on opening the lid



wow, mega difference.

Now, whist this shows that openings in the ground will stick out like a sore thumb, there are several places along the way that I suspect are run in shafts - and they didn't really show up, so my thoughts are that you are unlikely to see anything with the camera, that you cant see with your naked eye anyway.

I'm interested to experiment a bit more, perhaps in an area with lots of known open holes, small and large, see what shows up.

Can anyone think of anywhere - Hope Valley wise.. Cavedale??  one of the rakes - New Venture?? or perhaps stoney??

Awesome stuff mate - like looking into a furnace. And it was especially toasty down there last night in contrast to the bleak scene above. But I can see plenty of uses for this. One thing with the other shafts - perhaps the camera needs to pushed much closer, so keyholes or cracks around the edge gain a greater 'weight' in the overall sensor area?

Either way, it looks fantastic...

How about the back of Oxlow Hill, Mountbatten, Thistle area, or even Fawcett rake? although in these conditions you are likely to see any blow-holes without needing the camera.
A comparison between the same area with and without snow could give useful info on the practicality of the technique.

Thermal imaging has been used with some success to detect old mine shafts, not because they emit warm air, but because the soil around the shaft head is different or more compacted, and thus has a different moisture content. The amount of water in the soil affects its thermal properties quite markedly. This difference in the thermal signature of the soil shows up as a doughnut shaped ring around the shaft head.

As Duncan mentioned, we tried getting some thermal imagery flown over Llangattock, but alas we didn't get the data for various reasons. In hindsight, I suspect we might have found some warm(ish) bogs and hot sheep (this is Wales... ) rather than cave entrances. Incidentally the coldest part of Aggy was the first damp crawl at 5.5 deg C whilst the interior of the cave was a balmy 7.7 deg.

Mmmm .... Hot sheep.

Whilst it's bitterly cold it'd be great to capture a thermal image of that steam plume which comes roaring out of the cleft on the right in Cavedale above Puttrell's Aven in the Great Cave.

Apart from it being an interesting phenomenon in its own right it'd be useful to test the camera in a "best scenario" situation like that, which would then provide a good comparison when interpreting other less favourable situations.

This is excellent stuff Sam, well done.

Well as this thread has been dredged up by a sales rep...

With IR Thermography there is always the consideration of the emissivity value used on your scaling.

Sam's images above clearly provide a great indication of where the heat is coming from, but the you must be sceptical of the temperature value unless your e value is correct for the material in question. I'm sure he knows that, I've just seen it come up a couple of times in places of work where people have blindly trusted the default results. Use little spot stickers to help! (Pet peeve, apologies).

Anyway for caving purposes, i.e. finding blowing entrances, the indicative colour difference is rather lovely...

And Sam - those caps you gave me are potentially going into service at an undisclosed location in the coming weeks, you're more than welcome to join if you're feeling skinny.

[gmod]We are not in favour of this sort of 'sign up just to advertise my product' type of posting. Normally I'd just remove it but it does seem to be offering a guide to products rather than selling a product itself. I'm inclined to let it stand for now unless I receive any more complaints by members who are better informed.[/gmod]