I have encountered a spec of a laser diode given in dBm/nm instead of dBm:
If I want to know the average power between 2065nm and 2075nm, would the integral of that section give the average power in dBm? And what is the purpose of using dBm/nm and what does it mean?
How to interpret this type of dBm?
The same way you would interpret dBm × Hz, since wavelength is proportional to 1/Hz: dBm/wavelength = dBm/(1/Hz) =dBm × Hz.
It's not all that different compared to dBm/Hz, except that the "frequency" units are the reciprocal of what you got here (wavelength).
would the integral of that section give the average power in dBm
No, it will give the total power in dBm in that wavelength band.
The unit of average power would be the same as the unit integrated: dBm/nm. So, to get average power in dBm/nm:
what is the purpose of using dBm/nm
That's the spectral power density. How much power there is per each nm of wavelength. And indeed, if you integrate dBm/nm over wavelength, the unit of the integral will be dBm - exactly what you want. Simple way to think about it: the integral is the area, so an area of a rectangle would be [dBm/nm × nm=dBm].
The other two answers have reassured you that integrating the power is the correct thing to do, and why it's expressed that way.
Please bear in mind that 'integrating dBm' does not mean integrating the dBm number. It means converting dBm to power in watts, integrating the watts number, then converting back to dBm. dBm is a log unit, watts is a linear unit.
To be pedantic, what you will be doing is a summation (a discrete calculation) rather than an integral (a continuous algebriac calculation), as your original data is in discrete form in that graph.
would the integral of that section give the average power in dBm
Yes, it would give you the (total) power in dBm over the interval [2065nm ; 2075nm]
what is the purpose of using dBm/nm and what does it mean?
I think it has to do with the Optical Spectrum Analyzer's (the device that measure the laser power) finite resolution bandwidth \$ \lambda_{res} \$. Meaning, the OSA will not be able to measure the power received at \$ \lambda \$ but it will measure the power received on interval centered around \$ \lambda \$ with a width of \$ \lambda_{res} \$. Thus, a spectral density characteristic is more truthful to what has been measured.
