Chapter 1 introduces active galactic nuclei (AGN) in a clear but formal way. There's a fair bit of detail on the historical development of how AGN are classified into Seyfert 1s and 2s, and later Seyfert 1.2s and 1.5s; read it if you like. The most important material is on pages 5 (20) and 7-8 (22-23), and study the diagram on page 6 (21). This introduces the "unification model" used to calculate the properties of AGN; the model describes AGN as a black hole surrounded by an accretion disk and obscured by clouds of material. The whole point of Barb's thesis is to analyze a set of observations and compare them to the predictions of the unification model; as we'll see, some predictions were borne out and some weren't.
Section 1.2 derives an estimate for the required mass of the central black hole; a bright high-schooler could understand the equations here, but the rest of the thesis doesn't really depend on knowledge of this section. Read section 1.3.1, pages 14 (29) to 17 (32), which is a nice account of X-ray observations of AGN over time and how the picture got steadily more complicated. Section 1.3.2 discusses how X-rays interact with matter (you'll need to think back to your chemistry; remember electron shells?), and 1.3.3 discusses how these interactions form different components that are added up to form AGN spectra.
The goal of the analysis is to tease out these different components by fitting the observations to a model. Section 1.4 describes what Barb is trying to do; in particular, 1.4.2 starting on page 29 (44) describes some predictions that the unification model makes. Scan through this section.
Chapter 2 on the RXTE satellite is the easiest chapter to read, and I found it fascinating. Read the whole thing, from page 31 (46) to Barb describes the Proportional Counter Array (PCA) instrument which took all of the data that she analyzed; RXTE has other instruments, but the PCA is the most relevant for her purposes. The PCA is made up of five units, and these units have been successively failing over time; units 3 and 4 have periodic discontinuities, and units 0 and 1 have lost their propane layer and are now unusable for Barb's sources. Compensating for the shifting conditions requires using calibrated models of the background; these models also have to account for RXTE's passage through the South Atlantic Anomaly, where radiation levels increase and add more noise.
Chapter 3 covers the data selection process. Barb started out with all sources that RXTE observed in a given timespan, and then adds conditions: they must be Seyfert galaxies, have a long-enough observing time, must not span observing epochs (where the background models change), have received at least 125K photons for each spectrum, etc. After all of this, she has 39 galaxies and 821 spectra.
You can skip most of chapter 3, unless the data-selection details appeal to you. Scan through pages 72 (87) to 77 (92), though, which explain the seven parameters for the mathematical model (called pexrav) that Barb is using. Some of these values are fixed or constrained, and others are left to be freely solved for. The rest of the thesis will be mentioning these parameters a lot.
Chapters 4 and 5 discuss the results in great detail; you can skip these and jump ahead to the results. Here Barb looks at the sample properties, comparing the Seyfert 1 and 2 classes, and tries to figure out if the properties are significantly different between the two classes.
Chapter 6 presents the conclusions. In short:
- For Seyfert 1s, there's a relationship between the radio properties (radio-loud versus radio-quiet) and the X-ray spectra; three of the parameters are different between the loud and the quiet AGN.
- For two parameters, the RL and RQ Seyfert 1s are different. In RL sources, the parameters are correlated; in RQ sources, they're anti- correlated, the exact opposite. This indicates the unification model is too simple; some other factor is involved here. Barb suggests that the presence or absence of a relativistic jet could be the cause.
- For Seyfert 2s, the data is more scattered, also suggesting that the unification model is not capturing some aspect of AGN.
- There's a strong correlation between two parameters, Γ and R, but Barb tried analyzing randomly-generated spectra and also got a strong correlation. This suggests the correlation is an illusion; the software solving for the parameters can trade off the values of these two to fit the data, so the correlation doesn't actually represent anything real.
- She observes one anticorrelation called the X-ray Baldwin effect, but doesn't see a predicted relationship between Γ and the luminosity ratio.
- For some Seyfert 1s subclasses, the distribution of Γ matches the prediction of the unification model; for others, it doesn't. The absorption column for Seyfert 2s also doesn't match the prediction of the unification model. Hmmmm...
Pages 148 (163) - 150 (165) summarize all this, listing relationships and whether the results found were consistent with unification or not. Section 6.3 discusses what this means, and looks forward to future work; instruments that can observe a wider range of X-ray frequencies will be needed to resolve these questions.
The appendixes, all 150 pages of them, are all data tables and graphs. Flip through them and be suitably impressed Now go back and check out the acknowledgements on page iv (7); I found them touching.