4

I am processing a big meteorological data set, which is a global forcings dataset derived by other datasets and observations, and calculating relative humidity from qair,temperature, and pressure according to these formulas How do I convert specific humidity to relative humidity?

I am noticing that I have a lot of occurrences where RH is slightly higher than 100% and I am wondering if this makes sense or not considering this is air temp from 2m, I would expect that near surface RH over 100% is some errors in the data set.

*I am not a climatologist nor a meteorologist, I use this data to study plants, so I hope for a not too technical answer, thank you.

Herman Toothrot
  • 667
  • 3
  • 7
  • 15

2 Answers2

2

It is reasonable to assume that air (atmosphere) acting as as a solution could become supersaturated with water vapor. Supersaturation condition is where the air literally contains more water vapor than is needed to cause saturation. How common these conditions might be, I do not know.

Earth Science Expatriate
  • 5,949
  • 2
  • 13
  • 29
1

The relative humidity is the ratio of the partial vapour pressure of air, to the equilibrium vapour pressure over pure water, at the same pressure - yes, it's slightly pressure dependent.But I would put an emphasis upon 'Equilibrium'. In both physics and chemistry one can get all kinds of bizarre results under non-equilibrium conditions, and >100% RH is one of them. It is just a transient feature caused by rapid changes in temperature, pressure, or absolute vapour pressure, and as such it is not really a genuine meteorological observation.

Also, I advise caution when interpreting RH results. There is often a strong diurnal variation in RH, so if there is a met site taking only one reading a day, it could be unrepresentative of true average conditions. This is not so much a problem in advanced countries, where the measurements are (usually) reliably taken at set times. But it is often wildly misleading in developing countries where the observer may take a reading when/if he/she gets around to it, or maybe fakes the data altogether!

Gordon Stanger
  • 14,238
  • 23
  • 44