PSD from Acc vs Time data

I need to construct the power spectral density plot in G^2/HZ from recorded data of acceleration versus time. I also have the FFT data points from the Tektronics oscilliscope used to measure the original acceration vs time data. I know the x axis of the FFT is frequency in Hz, but it is not clear what the units on the vertical axis are. Assuming the vertical axis is Grms, how should I manipulate the data from the FFT to get the PSD?

I have 10,000 data points over a 2 sec time period for the Acc vs Time plot.
I have 500 data points from 0hz to 1250hz from the FFT of the Acc vs time data. All data is in a spread sheet where I can manipulate it.

I am mostly interested in energy at lower freq, below 100 hz. The FFT indicates considerable energy occurs in this range. Any info or advise is greatly appreciated!