-
|
As shown, we attempted to fit the SED of type-2 AGN with X-CIGALE. First we fit the SED without X-ray, and then we fit X-ray with this SED optimal parameter configuration (because this source is highly obscured and we can't make accurate correction to the X-ray data under the existing observations). It shows that the FIR/sub-mm does not fit well (the first figure). We want to know how setting parameters can solve this problem. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 20 replies
-
|
The 2nd figure looks better. Except that the 2nd figure doesn't use X-ray, is there any other difference between the two figures |
Beta Was this translation helpful? Give feedback.
-
|
You mean to change the redshift (0.0072) in file obs.txt to -0.0072 ? |
Beta Was this translation helpful? Give feedback.
-
|
No, the "redshift_decimals" parameter in pcigale.txt. |
Beta Was this translation helpful? Give feedback.
-
|
the red-shifting module setting? |
Beta Was this translation helpful? Give feedback.
-
|
No, under [analysis_params]. |
Beta Was this translation helpful? Give feedback.
-
|
Okay, thank you very much Mr. Yang. |
Beta Was this translation helpful? Give feedback.
-
|
The cause might be this: your X-ray flux is too low, and that single data point affects the normalization of the entire model. Imagine you move up the normalization of the model in figure 1, that would make the deviation in X-ray band worse and incur a huge increase in chi2. So, I guess the key issue is X-ray data is far too low (and it messed up the entire fit). What's your purpose? If you just need to visually show X-ray is heavily obscured, you can simply include X-ray module in your fits of Figure 2 (but without supplying the X-ray data). Then you will still have the full model (including X-ray) in the output, and just overplot the actual X-ray flux manually. |
Beta Was this translation helpful? Give feedback.
-
|
However, the signal-to-noise ratio of existing X-ray observations is insufficient. |
Beta Was this translation helpful? Give feedback.
-
|
Then just drop the X-ray data. That's not useful. |
Beta Was this translation helpful? Give feedback.
-
|
Okay, thank you very much Mr. Yang. |
Beta Was this translation helpful? Give feedback.
-
|
According to your suggestion, we reconstructed the SED of the target source, and the result is shown in the figure. |
Beta Was this translation helpful? Give feedback.
-
|
It seems the X-ray points is near the x-axis limit. You can change the plotted range by, pcigale-plots sed --xrange 1e-4:6e2 |
Beta Was this translation helpful? Give feedback.
The cause might be this: your X-ray flux is too low, and that single data point affects the normalization of the entire model. Imagine you move up the normalization of the model in figure 1, that would make the deviation in X-ray band worse and incur a huge increase in chi2.
So, I guess the key issue is X-ray data is far too low (and it messed up the entire fit). What's your purpose? If you just need to visually show X-ray is heavily obscured, you can simply include X-ray module in your fits of Figure 2 (but without supplying the X-ray data). Then you will still have the full model (including X-ray) in the output, and just overplot the actual X-ray flux manually.