I keep a dish of water outside on the back deck for the birds/squirrels. Currently, there’s a layer of ice forming on the surface of the dish, but ambient temp is 39F on the nearby thermometer. Is evaporative cooling causing the ice to form? Dew point at the airport (6 miles from here) is 15F.
I know frost or ice can form on your car’s windshield when the air temperature is greater than 0 °C and there are no clouds. It is due to the windshield radiating its energy into deep space. Not sure if it’s the same phenomenon as your water dish or not.
One factor is a temperature near the ground is often much lower than a thermometer will read several feet higher. If there’s wind it might contribute to evaporative cooling but with the temperature that low not much evaporation is occurring in calm air. Not sure how your deck is built but there could be a mass of cold air underneath it or a cold wind blowing air from below around the dish.
My first thought would be to use a “known good” thermometer (say: one of the ‘laser-aimed IR’ types) and measure temp of the water and/or the dish it’s in and/or the thing on which that water dish sits.
The “nearby thermometer” may be off, and/or – for any of a host of reasons – located in a spot where the temp is a fair bit different from that of the water dish.
And that’s before I’d start thinking of the meteorologic/physical issues like @Crafter_Man appropriately raises 
Thanks everyone. More info:
Using one of the laser IR thermos, the surface of the outdoor thermometer itself actually reads 42F. The deck floor and deck railing are 42F as well. The little platform that the dish sits on (outside of the railing) is 37F and the surface of the water in the dish is 33F. That platform got soaked this morning when I filled the dish with room temp tap water.
Was it (likely to have been) colder this morning … before you noticed this?
The water in the dish being just above freezing means it could simply have been 1+ degrees colder and frost could have begun to form … not to disappear quickly at a temp of only 1*F above freezing.
Y’know … maybe
ETA: also, the location of that outdoor thermometer – say: attached to siding of a building – might cause the thermometer to be slow to adjust to ambient temperature, since it’s in very close proximity to a lot of thermal mass that’s probably not warming up or cooling down all too quickly.
If that makes sense
I’d probably check the thermometer’s temp, the thermometer housing temp, and the temp of whatever surface it’s attached to.
One other semi-obvious thing: the ‘geography’ of the various items could make a fairly big difference, particularly when it comes to sun exposure and particularly first thing in the morning.
You may pick up a few useful pieces of info by playing with this:
Unless this is all North exposure (and you’re in the Northern Hemisphere) – shaded in the winter – the sun could be a big wild card – even for something as seemingly trivial as a shadow that gets briefly cast on the water dish
Indeed. Things were frozen this morning. Though not frozen solid. I used tap water to thaw, pour out, and refresh. But there was still some ice chunks stuck to the dish as I refreshed with the tap water. The dish is terracotta, and probably holds some (cold) water within its structure. I felt that my addition of room temp water would have abated any cooling effect from the dish. But, as I think of it, the mass of terracotta is probably equal to the mass of water it holds.
Buuttttttt. This happened yesterday, too. And yesterday, I started with completely dry vessels (2 dishes at different points of the deck. One of them is plastic). Both of the dishes frosted over after an hour or so. I wish I had collected more data yesterday.
Location is Raleigh-Durham.
ETA… I just looked again, and it’s only icy along the edge of the terracotta. So I’m assuming the terracotta was chilly enough to affect the temp of the water. I’ll keep an eye out and take better data if it happens to the plastic dish again.
Thanks to everyone for your time and attention!
Today I learned:
The frost point is between the temperature and dewpoint.
If you’re bored, and since you’re demonstrably handy with that IR thermometer, you could just try putting some numbers to all of these variables over the next few days.
I think – exposure being similar across the items – you’re likely to find that thermal mass and starting temp of the vessels, the temps of the surfaces that the vessels are sitting on (which would add to the relatively low thermal mass of, say, a plastic planter pot), and the temp and quantity of the added water are your biggest drivers.
to your ETA
This might be a factor too, if evaporative cooling is involved. Unglazed terracotta is used for many traditional (even ancient) evaporator cooling systems.
Big factor I think. Air movement may have stopped ice from forming earlier but once calm it could have started turning to ice around the edges before extending over the whole surface of the water. Evaporative cooling would be a contributing factor though maybe small like other things mentioned.
More about terracotta in this context:
The Persians and others used ice pools to produce ice even if the temps didn’t get down to freezing.
Many of the ancient cooling methods are making a comeback due to the simplicity of technology, no power needs, etc.
Don’t forget that the reading on an IR thermometer will vary with the IR emissivity of the object you’re aiming at. My IR thermometer has a chart with correction factors. Here a link to a full explanation.
Right. Before getting too deep into the science of cooling, I’d first consider that you might not be reading the right temperature. In addition to emissivity corrections, there’s also general inaccuracy of the IR thermometers - consumer-grade are typically only accurate to 2-4°F.
Right on. I have a high $$ Fluke, yet it’s only good for general guesstimations, even with an experienced user. If the OP needs to split hairs (as in this case) when a few degrees make a big difference one way or the other, I’d suggest using a “K” thermocouple.