Good Day Everyone...
I have a (more or less) a technical question as related to the Mav 2ED and the thermal camera.
Recently I am was trying to interpret thermal imagery of a building's roof.

Sunrise...sun not of roof yet...
Outside temp was approximately 40F...
I am approximately 75 feet above ground level / approximately 55 feet above building / a flat roof building...
I expect to see varying colors based on the roof's radiative temperature and the particular color palette I am using...
(Essentially the brighter the color, the warmer it is -- relatively speaking)...
Isotherm setting is off...
ROI set to 33%...
Gain set to auto...
Auto FFC is enabled...

Using the "Rainbow" pallet or the "Hot Spot", I see what I would expect.  Varying colors based on how the roof was radiating in the cool morning.
Found varying hot spots (almost white) which suggest a "warmer / hotter" area(s) than orange or red 10 to 20 feet away.
Found varying insulated ductwork that is a deep shade of blue to almost black which suggests that this particular item was near to the current outside temp of 40F -- foil insulated duct and pipe.

All of which I would expect.

I enabled the  "spot temperature" feature to take a spot temperature reading.  I was curious to know how "warm" the roof is in varying spots.
I expected the roof to be warmer than the current outside temperature.

However, I received a spot temperature in the mid to upper 20's -- essentially below freezing on a cool day with the outside temp approximately 40.
The lowest temp was 39 or 38 during the night.

And to add "insult to injury", there is a roof-mounted temperature sensor located approximately 20 feet above the roof.  I checked it and it read 38F / 39Fdurng my operation.

I do not understand.  I would expect a spot temperature above the current outside temperature.  It hasn't been below freezing for several days.  The roof is "dry" -- no snow or rain recently.  

Later at my desk, I held the Mav and aimed it at my forehead -- 96 and some change -- close enough.  That suggests to me the sensor is probably ok?

Would someone be kind enough to explain why a roof's radiative temperature would be below freezing under this set of circumstances?
Thank You

=====================================================================

JOHN TRACEY BENNETT – SIU / PSO Unmanned Aerial Vehicle Administrator

FAA Certified sUAS / UAV Operator (active)

Chief Engineering Draftsman

CAD Systems / Archive Systems Coordinator

PLANT & SERVICE OPERATIONS

MAIL CODE 6727

SOUTHERN ILLINOIS UNIVERSITY

210 PHYSICAL PLANT DR

CARBONDALE, ILLINOIS 62901

E:  JohnB@PSO.SIU.edu / JohnB@SIU.edu

P:  618/453-6766

F:  618/453-6784

PSO.SIU.EDU

Was the roof wet?

Good Day!

No.  The roof was not wet.  It has been dry for several days.  It is a ballasted (gravel) roof, membrane, tapered roof insulation attached to a metal roof deck.  Roof drains.  Rooftop units (RTUs), insulated ductwork, conduit and pipes, hot pipe and vent penetrations.  Standard roof stuff.  It is somewhat old, damaged insulation, patch, leaks, etc.

I felt the roof would be at its coolest / radiating the most warmth just at sunrise --- the point before the sun just starts warming the roof.  I expected a lot of heat loss through the roof since it is an older building.  I saw what I expected (within reason) when I looked at the video.  Warmer segments of the roof "warm colors", cooler parts a darker blue.  Some of that was recently new insulation.  Some of it was bonafide damage.  I would have expected the radiative temperature of an older roof of a building that is being used at the same temp or slightly warmer than the surrounding air temp.  Thinking  I know (rather knowing I know), I thought I would take several spot temps and do a guestimate of the whole roof, trying to figure out where water damage might be hiding and how the roof insulation was performing overall.

The spot temp reported colder than the surrounding air and on a roof that is supposedly warmer than the surrounding air.  This implies to me (among other things) that:

1)  I am doing something obviously wrong...
2)  The roof was below freezing (I currently refuse to believe that unless someone could prove otherwise)...
3)  The spot temp reading is off -- how to calibrate, if possible...
4)  There is a distance vs accuracy relationship. ( although that wouldn't explain it reporting sub-freezing temperature on, presumably, an above freezing surface)
5)  If its a distance vs accuracy problem, what is the relationship function?  (for example, 1-degree error for every ten feet -- assuming I was approximately 35' (10 meters - ish) above the roof, that would imply a 3 to 4 degree error. Potentially, I had a 15 degree or greater spot temperature error in my readings)

With that said, I spent some time looking at this forum and some else made a similar comment that the spot temperature function was possibly reporting an erroneous temperature.

I cannot figure out how a warm roof, above 40 degrees, would report a sub-freezing temperature when the last several days' temperature has been above freezing.  That s why I am asking someone to explain it to me.  I feel I am missing something "obvious"...maybe.
Thank you for taking the time to respond
John
=====================================================================
JOHN TRACEY BENNETT – SIU / PSO Unmanned Aerial Vehicle Administrator
FAA Certified sUAS / UAV Operator (active)
Chief Engineering Draftsman
CAD Systems / Archive Systems Coordinator

PLANT & SERVICE OPERATIONS
MAIL CODE 6727
SOUTHERN ILLINOIS UNIVERSITY
210 PHYSICAL PLANT DR
CARBONDALE, ILLINOIS 62901
E:  JohnB@PSO.SIU.edu / JohnB@SIU.edu
P:  618/453-6766
F:  618/453-6784
PSO.SIU.EDU