hi

In all books and softwares that I know it is zero.

I'm sure Mr. Jamalpoor means one for the dielectric constant, zero for loss tangent.

Assuming you are asking for something with greater precision, Wikipedia gives:

er = 1.00058986 ± 0.00000050 (at STP, for 0.9 MHz)

and references a 1936 measurement by Hector and Schultz.

No loss tangent is given, but perhaps you can derive what you need from plots like these, attached, for specific attenuation in the atmosphere. Depends on humidity and other factors, of course. Its hard to find data that low in frequency, but these plots (and some others that I found) suggest it limits at around a few thousandths of a dB per kilometer.

Let's say 0.005 dB/km. That's about 0.5*10-6 Np/m. The loss tangent at 1 MHz is then:

tan d = (0.5*10-6 Np/m)(3*108 m/s)/(pi*106 Hz) = 5*10-5

At 30 MHz, it would be about 30 times lower, or 1.7*10-6

This is all very approximate of course. As Mr. Jamalpoor said, pretty close to er = 1 and tan d = 0.