I'm analysing a time series with a Lomb-Scargle periodogram. Here's what the periodogram looks like:
As you can see, there are three peaks very close together (I've tried much finer frequency grids and they are still there). So there are several very similar frequencies in my data. I'd like to find a period estimate with error bars that capture all three peaks, but I'm not sure how I'd do that. I've tried fitting a gaussian to all three peaks, and taking the FWHM of all three, but neither method seemed to work satisfactorily (the smallest peak on the right tends to get ignored in the solution). How would you go about doing this?