Hi!

I looked at the gamma-ray plots produced by galplot from the conventional and plain diffusion models fits files that I made using Galprop 50p. It seems that the low-energy EGRET data, which in my understanding should be independent of any galprop result, change from one model to the other in the plots.

I put the eps-files and galdef and galplotdef files for thw two cases in

http://accms04.physik.rwth-aachen.de/~h ... ots.tar.gz

Any comment would be appreciated.

Henning

Some output from galplot showing the differences in the EGRET spectra follows:

--- 599298h ---
EGRET spectrum *E^2: E1 E2 Ebar Ibar 30 50 38.7298 0.0204328
EGRET spectrum *E^2: E1 E2 Ebar Ibar 50 70 59.1608 0.0220291
EGRET spectrum *E^2: E1 E2 Ebar Ibar 70 100 83.666 0.0261807
EGRET spectrum *E^2: E1 E2 Ebar Ibar 100 150 122.474 0.032454
EGRET spectrum *E^2: E1 E2 Ebar Ibar 150 300 212.132 0.0457423
EGRET spectrum *E^2: E1 E2 Ebar Ibar 300 500 387.298 0.0577835
EGRET spectrum *E^2: E1 E2 Ebar Ibar 500 1000 707.107 0.0672297
EGRET spectrum *E^2: E1 E2 Ebar Ibar 1000 2000 1414.21 0.0685659
EGRET spectrum *E^2: E1 E2 Ebar Ibar 2000 4000 2828.43 0.0598313
EGRET spectrum *E^2: E1 E2 Ebar Ibar 4000 10000 6324.56 0.0338678

--- 999726h ---
EGRET spectrum *E^2: E1 E2 Ebar Ibar 30 50 38.7298 0.023998
EGRET spectrum *E^2: E1 E2 Ebar Ibar 50 70 59.1608 0.0248845
EGRET spectrum *E^2: E1 E2 Ebar Ibar 70 100 83.666 0.0285792
EGRET spectrum *E^2: E1 E2 Ebar Ibar 100 150 122.474 0.0344775
EGRET spectrum *E^2: E1 E2 Ebar Ibar 150 300 212.132 0.0476831
EGRET spectrum *E^2: E1 E2 Ebar Ibar 300 500 387.298 0.059519
EGRET spectrum *E^2: E1 E2 Ebar Ibar 500 1000 707.107 0.0686318
EGRET spectrum *E^2: E1 E2 Ebar Ibar 1000 2000 1414.21 0.0685217
EGRET spectrum *E^2: E1 E2 Ebar Ibar 2000 4000 2828.43 0.0598327
EGRET spectrum *E^2: E1 E2 Ebar Ibar 4000 10000 6324.56 0.0338675

I do not know how exactly they were produced, I have to download the plots, but I guess that the regions on the sky are exactly the same. The EGRET data can be somewhat different in case if the deconvolution has been applied. The deconvolution is model dependent. It works in such a way that the model is convolved first and then the coefficients derived from a comparison of the original model and convolved model are applied to the data. The deconvolution is necessary to compare the data from different instruments, e.g., COMPTEL and EGRET, on one plot. If only EGRET data are used then the deconvolution is unnecessary; it is enough to convolve the model. Andy Strong can probably say more.

* convolution usually means "convolution with the instrumental point spread function (PSF)"

_________________
Igor Moskalenko
Stanford University

yes Igor is right.

Plotting a spectrum comparing model with data, one can decide either to put the instrumental response in the model curve or the data points. So we could either keep the EGRET points 'as measured' and plot the spatially convolved model, or try to correct the data points for the convolution. The latter solution was adopted so that the spectra are 'real' photon spectra
and data from multiple instruments can be placed on the same plot.

Since the EGRET PSF widens to low energies, the effect is maximised there, but is present at all energies at some level. Since the PSF spreads the emission, it reduces the observed intensity averaged over a given latitude range. Correcting for this increases the intensity. You can see the effect explicitly by running galplot with convolve_EGRET=0
in which case the correction is not applied.

This is to be contrasted with our longitude and latitude profiles, where we DO convolve
the model for direct comparison with the original EGRET data - in this case only one instrument is involved, so it is the better procedure.

By the way this procedure is described in detail in our 2004 paper.

_________________
Andy Strong, MPE

I see.

So is there an option in the galplotdef file to convolve the model with the EGRET PSF? In other words, what is the best way to compare two different galprop models to EGRET data in the same plot?

Henning

yes, in the current version there is an option to output the PSF convolved models
which is done if
fit_EGRET = 2


However it is probably not in the public version - if you want I can make it available.


There is no possibility to show more than one galprop model on the same plot at present.

_________________
Andy Strong, MPE

From what I gather from the code, there seems to be an option, fit_EGRET=2, which leads to the subroutine convolve_EGRET_HI_H2() being called. If that is not what you mean, I would be grateful if you could make the new version available.

By the way, the files
counts.vp0001.g006
counts.vp0001.g007
counts.vp0001.g008
which seem to contain the EGRET data above 10 GeV, that you used in your 2004 paper, are not included in my Galplot version and I have been unable to find them on the website.

Henning

yes you are right: if it is already in your version please try it.

those non-standard EGRET energy range (>10 GeV) files are not public, have to
check with Olaf Reimer who produced them

_________________
Andy Strong, MPE