PSS Analysis's reliability-Difference between time domain response in PSS and Tran analysis

 I'm designing a wide band matching network to drive an antenna.The output pulse at the output of pulse generator is delivered to one N1PORT as the antenna model thet refered to a S1P file with value from 1GHz to 8GHz. I want to use this interface between antenna(S1P port with n1port component) and pulse generator(transmitter) I did PSS and Transient analysis to plot output signal. Here are the defined parameters in both simulations:

1-PSS:fundamental tone in 40MHz and I'm going to calculate harmonics to 10GHz cause I want to see output PSD to 10GHz so I defined 250 as number of harmonics.

2-Transient: As input pulse is a unipolar 40MHz pulse rate with 1.2V amplitude, I set the transient time equal to 200ns to reach steady state condition (However, I think my circuit has no transient state).

After simulation, I plotted transient response of output at the positive port of N1PORT both in Tran and PSS simulation. ( to put matching circuit between them I grounded one input of 1-port network and directly connected another input to  matching circuit's output). But  the results are completely different, so:

a) I wonder to know which one is reliable based on using 1-Port instance, PSS or Tran analysis? (I think Transient is the most reliable analysis but as I have s1p  file, I'm not sure!!)

b) I want to Plot PSD(Power Spectral Density) of output signal, I used direct plot form to plot PSS>---Function: Power>--Modifier:dBm. I'm not sure that is it true or wrong? Indeed, dBmP is PSD of output? How can I plot output's PSD in 1Hz or 1MHz scale?

 Thanks in Advance.