ITU-R recommendation P.530; staistical data; basic interpretation needed!

I am trying to translate a statistical database from ITU-R (the administration that sets out recommendations for the engineering of wireless systems) into programming code.

The databases describe the statistical probability that the refractive gradient of the air exceeds a certain value in a given location of the earth within a year.

Example for datasets:

- value for refr.gradient that is exceeded in 99.9% of a year
- value for refr.gradient that is exceeded in 99.0% of a year
- refr.gradient exceeded in 0.1% of a year

The problem for me is, that the recommendation says that a value from this dataset should be used in a calculation, that the probability of NON-exceedance is 1% of a year.
Is it correct to take the value of the 99% exceedance file?

And a second question, if I may: Exceedance for my understanding means always towards the positive side, matheatically speaking?
E.g. a value of -40 is exceeded by a value -39?

Thank you
Regards Sporex