I researched this a little bit as I was dubious myself. I'm still trying to understand the mechanics of it all but it appears to be correct. You can even get different humidity levels with different salts. They do change with Temp but in the case of Table salt (Sodium Chloride) the difference between 0 and 80 Degrees is less than 1% in RH.
I did find this in the comment section of a youtube video on how to do the Salt test.
In pure water the molecules, energized by their temperature,
jump into the air until the rate at which they jump into the air
(fighting against their Van der Waals attraction to the other water
molecules in the liquid phase) and fall back into the liquid are equal,
which depends on the density of water molecules in the air. (The more
water molecules in the air the more likely they will hit the water and
stick. That rate has to equal the rate of leaping off.) That fixes the
relative humidity, for pure water that is 100% RH.
The water molecules in the salt paste are
bound by much stronger Van der Waals forces to the salt molecules. It
is harder for them to leap into the air. Lower density of
water molecules in the liquid phase balances the rate at which they
leap into the air (due to thermal energy fluctuations). if there is a
higher density of water molecules in the air then they will land more
than they leap until the two rates are equal, and similarly if there is
less than the equilibrium value. If you wait long enough, and it may
take a while since the surface area of the paste is very small compared
to the size of the container, the RH will reach the equilibrium value
determined by the degree of binding of the water molecules to that
particular salt at that temperature (and perhaps pressure.)
For each type of salt the molecular binding / Van der Walls force is
different and that causes a different relative humidity (at the same
temperature). There may be other forces involved than Van der Waals.
Some water molecules are probably bound by strong dative / covalent
bonds to some salts (even some dry salts have water molecules in their
crystals), but in a paste others may be glued by Van der Waals. My guess
would be only the weaker bound water molecules contribute to the RH.
For example MgCl salt gives about 33.3% RH and Na,2Cr20 gives about 55%
RH at room temperature. NaCl is just over 75.5%. (pure water = 100% RH, desiccators could give 0% RH).
It would make sense if you can only calibrate
one value to pick one near the desired RH. http://nvlpubs.nist.gov/nistpubs/jres/5 ... 19_a1b.pdf
What I do is use several references and put marks on less expensive analog
meters to show the correct values. For digital meters that don't
calibrate, or if you want accurate values over a larger range, you can
make a table of the true readings vs. the nominal readings.
If you want high accuracy, the reading can change quickly when you open the bag. For analog meters you can remember
where to put the mark. For digital meters you can hit the buttons through a flexible bag.
I have read in some places that you should not calibrate some digital
meters. However I don't think digital meters with calibration work
differently than ones that don't have built in calibration, and I can't
imagine how it could harm a digital meter to stick it in a bag with