Group Index dependence on refractive index

[loginorsinup]

In N_g (the group index for a range of wavelengths), there is an index of refraction n used, but if the medium is dispersive, meaning n is a function of wavelength λ, which n is used? Is it some kind of an average? Or does n not change much over this range of wavelengths? if it doesn't change much, then

N_g = n - λ(dn/dλ)

would be approximately

N_g ~ n

So, it makes sense to me that dn/dλ must not be really small, because otherwise N_g ~ n and the concept of group index is redundant with the concept of the index of refraction. but if dn/dλ is not really small, then the choice of n (the first term in N_g) becomes crucial, does it not?
https://ssl.gstatic.com/ui/v1/icons/mail/images/cleardot.gif

 

[DrDu]

$N_g$ is a function of $\lambda$ too: $N_g(\lambda)$, so you should use the respective value $n(\lambda)$ in the formulas you gave.
Take in mind that the group index refers to a wavepacket which is very broad, so that the spread of $\lambda$ in Fourier space is arbitrary small. If n depends strongly on $\lambda$, the wavepacket will deform while traveling through the medium and it will be difficult to define a group velocity, or, you have to regard it as a superposition of broader wavepackets which travel with slightly different group velocities.

 

[loginorsinup]

Thanks! Makes total sense!