21.02.2013 Views

Gas Disks and Supermassive Black Holes in Nearby Radio Galaxies

Gas Disks and Supermassive Black Holes in Nearby Radio Galaxies

Gas Disks and Supermassive Black Holes in Nearby Radio Galaxies

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

3.3.3 Quantify<strong>in</strong>g error sources<br />

The STIS data h<strong>and</strong>book (Brown et al., 2002) gives the follow<strong>in</strong>g absolute <strong>and</strong> relative<br />

accuracies applicable to this work: A wavelength absolute calibration error (∆λ offset)<br />

of 0.1 to 0.3 pixels (2.6 to 7.7 km s −1 at 6500˚A) with<strong>in</strong> an exposure, <strong>and</strong> from 0.2 to<br />

0.5 pixels (5.1 to 12.8 km s −1 at 6500˚A) between exposures. An absolute photometry<br />

error of 5% <strong>and</strong> a relative photometry error of 2% with<strong>in</strong> a s<strong>in</strong>gle exposure assum<strong>in</strong>g<br />

a wide slit observation. 5 µm variations <strong>in</strong> slit width along the slit lengths could<br />

result <strong>in</strong> variations of up to 20% <strong>in</strong> flux along the 0. ′′ 1 slit.<br />

In Verdoes Kleijn et al. (2002a) Hα + [N II] fluxes were presented for each nucleus<br />

<strong>in</strong> the sample. The values presented there agree well with the values we f<strong>in</strong>d here,<br />

certa<strong>in</strong>ly given our limited ability to extract comparable apertures <strong>and</strong> with<strong>in</strong> the<br />

20% potential flux errors noted above.<br />

In Section 1 we <strong>in</strong>dicated that a 1.3 km s −1 error could be <strong>in</strong>corporated <strong>in</strong>to the<br />

f<strong>in</strong>al data as a result of shift<strong>in</strong>g the spectra for image comb<strong>in</strong>ation <strong>and</strong> cosmic ray<br />

rejection. This shift is <strong>in</strong>significant compared to other error sources.<br />

In Table 3.26 we showed that by allow<strong>in</strong>g different free parameters with<strong>in</strong> the<br />

s<strong>in</strong>gle-Gaussian-per-l<strong>in</strong>e fit produced changes <strong>in</strong> the measurement <strong>in</strong> velocity of ∼<br />

67

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!