We present a detailed analysis of the intrinsic X-ray absorption in the Seyfert 1 galaxy NGC 4151 using Chandra /High Energy Transmission Grating Spectrometer data obtained 2002 May as part of a program which included simultaneous ultraviolet ( UV ) spectra using the Hubble Space Telescope /Space Telescope Imaging Spectrograph and the Far Ultraviolet Spectrographic Explorer .
Previous studies , most recently using ASCA spectra , revealed a large ( > 10 ^ { 22 } cm ^ { -2 } ) column of intervening gas , which has varied both in ionization state and total column density .
NGC 4151 was in a relatively low flux state during the observations reported here ( \sim 25 % of its historic maximum ) , although roughly 2.5 times as bright in the 2 –10 keV band as during a Chandra observation in 2000 .
At both epochs , the soft X-ray band was dominated by emission lines , which show no discernible variation in flux between the two observations .
The 2002 Chandra data show the presence of a very highly ionized absorber , in the form of H-like and He-like Mg , Si , and S lines , as well as lower ionization gas via the presence of inner-shell absorption lines from lower-ionization species of these elements .
The latter accounts for both the bulk of the soft X-ray absorption and the high covering factor UV absorption lines of O VI , C IV , and N V with outflow velocities \approx 500 km s ^ { -1 } .
The presence of high ionization gas , which is not easily detected at low resolution ( e.g. , with ASCA ) , appears common among Seyfert galaxies .
Since this gas is too highly ionized to be radiatively accelerated in sources such as NGC 4151 , which is radiating at a small fraction of its Eddington Luminosity , it may be key to understanding the dynamics of mass outflow .
We find that the deeper broad-band absorption detected in the 2000 Chandra data is the result of both 1 ) lower ionization of the intervening gas due to the lower ionizing flux and 2 ) a factor of \sim 3 higher column density of the lower ionization component .
To account for this bulk motion , we estimate that this component must have a velocity \gtrsim 1250 km s ^ { -1 } transverse to our line-of-sight .
This is consistent with the rotational velocity of gas arising from the putative accretion disk .
While both thermal wind and magneto-hydrodynamic models predict large non-radial motions , we suggest that the latter mechanism is more consistent with the results of the photoionization models of the absorbers