I have had my Tempest installed and in use since it was first available a couple years ago. When the sun is bright, the temperature reading is too high; I’ve seen it as much as 9 degrees too high. My ability to site the unit is limited due to my suburban location, HOA restrictions, etc. It is mounted to my deck railing using a 10’ long fence-top pole (as discussed in installation suggestions). The Tempest is in full sun for about half of the day after which the sun is blocked by my house roof. I’m comparing it to 3 other local (mine) temperature sensors. One is under a board of the deck, another under a bay window, third on northern side of house to block the sun. The three other sensors always read within a couple degrees of each other. When overcast or in the afternoon when the sun is blocked by my house, all 4 sensors are very close in readings. In bright sunlight, the Tempest is too high. I haven’t seen any discussion of this in the forums so I guess my unit is unique; not sure why. I’m bringing it up now because I’m recently trying to use the data in my Home Assistant automation and would like it to be more correct. Any suggestions how to correct this?
Can you post a photo of your installation?
Can you also post a URL link to your station so that “tech-gurus” can examine your Tempest’s temperature data? Upon looking at its location. . .it May Be too close to the house and getting a bit of “Heat Contamination” from the upper portion.
Here is the URL for my station ID 19232. Today was a very sunny day here in MD; tomorrow is predicted to be rainy.
always a tricky thing when you have instruments close to constructions.
The Tempest has a compensation algo for temperature to compensate the heating of the housing (elements used are wind, solar radiation …). This was measured in laboratory conditions during development. But this is again in ideal conditions.
From your writings all your other sensors are near or onto structures that influence directly the readings by radiation except to a point your Tempest (although I think your house must radiate enough to influence). Not to sure which sensor can be trusted
This is your Tempest temperature reading compared to nearby sensor on the airport (Flower Hill). Knowing that that sensor must be in open terrain and that wind has more influence on it compared to your Tempest. But the trend is similar (blue line is your Tempest)
I have this problem also.
My tempest is 5-10 degrees warmer than other tempest stations in my area during sunny afternoons. I didn’t notice it during the summer, only in the last couple of months.
I think I also use the same type of pole (galvanized fence pole). I’ve noticed that when the tempest is measuring these “high” temps, the pole itself is quite warm from the sun. My guess is that this heats up the sensors somehow.
It looks like the back of the house faces SSW.
My guess, along with the others who have replied, would be some heating of the backside of the house and also additional solar radiation from reflection from the siding (which is light colored).
The heating of the pole is another interesting theory.
Reflection from GLASS can fry (grill, whichever you think is healthier) the instrument
Unlikely - so many others use the same mounting that I’d expect it would have been noticed by now. That said, I’d suggest moving the Tempest at least temporarily to at least 10’ from the house and see if that helps. That might help identify if there’s heat reflecting off the house itself.
I appreciate everyone’s input on my issue albeit a bit discouraging; obviously hoping for a silver bullet solution. I hadn’t anticipated that there would be enough solar radiation/reflection from the wall to affect the Tempest about 6 feet away. In response to @george1 I’m standing pretty much South from the Tempest in the previous picture; the stair hand railing is pointing North to South. So the back of the house is looking to the East. So the Tempest gets direct-ish sun until noon-ish when the ‘sunset’ occurs over the roof peak. Based on the UV sensor, it looks like the house ‘sunset’ occurred around 1:30PM yesterday:
I’ve also included the UV reading of another Tempest a couple miles away of the same time frame verifying (at least to me) that the sun was still strong in the sky. (My recollection was that yesterday was a very sunny day all day with very few clouds.)
So, looking at the graph from @eric , what I think I’m trying to correct is what appears to be a false temperature of almost 5 degrees too high from 11AM to 1:30PM. I’ll try to fashion an experiment to verify the solar reflection theory. I’m thinking to temporarily lower my tower pole and put some sort of solar shade over top of the Tempest (not sure how yet). My thought being this should still exhibit the false reading in direct sun if the problem is coming from the house wall. I will also move the separate stand-alone temperature sensor from underneath the deck to be more co-located with the Tempest to see how it is affected. Of course, not sure what I will be able to do about all this since I have limited mounting options in my small backyard which also has some 20’ cypress bushes off to the right of the above picture.
Just for a test, relocate it for a day or two.
Ideal would be to get the Tempest still facing South, not on anything that itself will heat up in the Sun (ie, not on the picnic table), at least 20 feet from the house or so, and waist-level height or above.
Less ideal would be to just loosen your hanger and let the pole slip down to the ground, as long as the Tempest doesn’t pick up reflections off the glass when the sun is in the wrong angle.
You probably have a siting issue, so don’t sweat the fact that moving it will alter your wind readings. Work the temperature question first.
I honestly wouldn’t think so either, but my pole is nowhere near the house, at least 40’ away.
When I grab the pole with my hand it is noticeably very hot, almost too hot to grab and hold. (2pm on a clear day). It is close to some trees, but that’s about it.
Right now the temp on 2 nearby Tempest stations is 71 & 73 while mine is 81.
OK on direction… I was basing my estimate on the direction of the solar panels.
Let us know how your experimentation goes,
Not being an expert but I can imagine when there is wind the heat is blown away from the pole. When there is no wind the heat goes up and I can imagine that in those conditions the hot air goes even into the shield and creates a hot bubble in there. Will ask someone that has knowledge about that if I can catch him.
I do not have the solution but I have carried out similar experiments trying to match my Tempest temperature to my ‘Air’ and other thermometers in an adjacent stevenson screen.
I find using an infrared thermometer gun useful for testing and I do find the Tempest solar panels become very hot when they are facing the sun.
I have not published my suspicions from my tests.
If you completely shade the Tempest including the solar sensor on top of the Tempest I believe that it will cause the Weatherflow Temperature compensation to be defeated and your Tempest should report the actual temperature from its sensor.
Without two Tempests for comparison I found it very difficult to compare changes.
The wind makes a difference both its speed and direction. If the breeze hitting your house causes a gentle updraft passing anything warmer like a dark coloured deck in the sun or the Tempest pole and solar panels that face the sun and then flows into your temperature sensor I believe that it will read hotter than the surrounding air.
As a paragliding pilot when I fly using thermals I often use galvanised steel high voltage electricity towers as a thermal source. The sun heats them and they disturb the wind causing a thermal updraft.
As the OP, I thought I should circle back and post where I am with my problem in case anyone in the future happens upon this posting. I decided it didn’t make a lot of sense to try and ‘prove’ that the Tempest algorithm was correct or incorrect since so many people are using the Tempest successfully. So, I took @vinceskahan suggestion to try and see if there was a better siting solution. Unfortunately I don’t have a lot of potential sites. It needs to be in my backyard due to HOA restrictions; my side & back neighbors house walls are less than 20’ from my walls and there is a group of tall cypress bushes that I need to work around. Due to all of the obstructions, there is very little air flow in my back yard; I knew the Tempest wasn’t going to be able to provide a very good wind indication (and that has been verified). My intent is to use the wind sensor from the local airpark about 3 miles away, the NWS forecast, and merge that data with the Tempest data and display all of the fused data in my Home Assistant UI (which is working well) instead of the Tempest App. As stated by @iladyman, it is very difficult to determine what is truly the real air temperature. I tried to pick similar cloudless days with similar UV strengths for comparison purposes. As indicated in my original post, I have 3 other sensors that I can monitor, several other Tempest within a few miles that appear to have a good sky view, and the airpark. In early testing, I was relying heavily on the airpark data as reported by NWS. Midway through my experiments, I determined that the NWS feed using the NWS API for the local airpark was typically delayed about 30 minutes (including the timestamp being off) from the real data as reported directly on their website (which I determined the website was correct). So during mornings of rapidly rising temperature, this 30 minute delay can have a huge affect. After this discovery, more experimenting. I mounted the Tempest with the same 10’ galvanized pole (to remove that variable) to a tripod that I could move around. As we moved from winter to spring and a higher sun, the UV intensity has gone from to 2’s to the 4’s. More variables. My conclusion after all this was that for me a location half-way between my house and my neighbor’s was the best that I could do and indeed it is a more consistent temperature location when comparing with all the other sensors. Don’t totally know why but I guess just various thermal reflections off of everybody’s walls. Not an ideal location, but I’ll use a mail-box post ground attachment to erect something a bit more permanent. Thanks everyone for their comments and I guess experimenting is the only way to figure it out.