Why the standard deviation of 1D power spectrum so large?

Yun Wang
  • 2
  • 6 Nov '20

Hi guys,

Firstly, I obtained a 1024^3 mesh from TNG100-1 data by CIC assignment scheme. And then I chose 10000 lines along the z-axis orthogonal to the x-y plane, so each line contains 1024 (\delta + 1).

Next, I perform FFT for every line. Finally, I compute the 1D power spectrum by averaging over 10000 lines.

uploading CodeCogsEqn (3).gif

Here is my result of dark matter

uploading dm_power_spectrum_1d_z0.0.png

However, the standard deviation of this power spectrum is so large, the maximum of std is a prop to 10^8. Why is it happen?

uploading dm_power_spectrum_1d_std.png

Dylan Nelson
  • 6 Nov '20

Hello Yun,

Have you compared your results to the power spectra computed and presented in Springel+18 (e.g. Fig 4)?

Yun Wang
  • 7 Nov '20

Thank you for your reply. I don't know why I failed to upload my figures.

The shape of my power spectrum is similar to Springel's.

  • Page 1 of 1