If we take the Square-root of frequency domain what will be the impact on the time domain?
e.g. p(t) is our time domain function ,
its fourier transform is P(f) ,
If we take the square-root of P(f) ( P(f)^0.5 ) , what is the impact on the time domain p(t) ?
Unless there is some specific property including the root square, you can not determine it, because the root square operation is not linear, therefore you can not find an equivalent in time domain
If it's truly the square of the Fourier Transform (as opposed to the square root), that is equivalent to convolving the untransformed data with itself (or maybe its conjugate, if that's how the spectrum was squared).
In fact, this is the autocorrelation. If you recall, the autocorrelation and PSD are linked by the fourier transform (Wiener-Khintchine relation).
This assumes a stationary random process, of course.