Re: Fit a combined spectra

I don't think you need this:

    Raul>  match_dataset_grids(data_vec1);
    Raul>  match_dataset_grids(data_vec2);

since it looks like the vec1 is MEG and the vec2 is HEG - the +1 and -1
orders are already on the same grid.


This isn't necessary, but could be nice if you want to look at the
fluxed arrays:

    Raul>  flux_corr(data_vec1);
    Raul>  flux_corr(data_vec2);


I think your fit might be spurious because you left it totally
unconstrained.  Try setting some reasonable limits on the parameters:

    Raul>  set_par("xaped(1).norm",1,0);



    Raul>  set_par("xaped(1).temperature",1e+7,0);

set_par("xaped(1).temperature",1e+7,0, 1.e6, 1.e8);

Since AtomDB is low-density only, this has no effect, so leave it frozen
at 1.0:

    Raul>  set_par("xaped(1).density",1,0);


The norm is % this is  1.e-14 * VEM / (4*PI*d^2), which will require an
estimat, or better, we'll renorm_counts when ready:

set_par("xaped(1).norm", 1.e-3, 0, 0.0, 0.1 );  



    Raul>  list_par;
    Raul>  exclude( all_data );
    Raul>  include(data_vec1);
    Raul>  group_data( data_vec1, 4 );
    Raul>  xnotice( data_vec1, 1.7, 25 );


Delta Ori probably has low nH.  If you have an estimate, set it and
freeze it.. NH will be degenerate with temperatures and abundances.  To
get and estimate using a isothermal plasma, you might do:


set_par( "phabs(1).nH", 0.1, 0, 0, 0.5);


Delta Ori also has slightly broadened lines, I think, so try the
following:

use_thermal_profile;  % to turn on thermal + turbulent broadening

 set_par( "xaped(1).vturb", 300, 0, 0.0, 1000);


Check the params to see if they have reasonable ranges, or are left at
some reasonable frozen value:

list_par;


Now get a better guess for the norm:

renorm_counts;
list_par;

popt.res = 4;  % for combined residuals
plot_counts( -data_vec1, popt );

And adjust the norm limits, if needed.

Now try the fit, and see if it converges on something reasonable, within
the limits of a low-density, isothermal model.  The emission measure
should at least be well constrained.  You might need to allow the
ovarall metal abundances to be free (you might give individual abundances can be
specified in the plasma state variable in a future refinement, once the
basics are working).

You will get a bad fit for some lines --- the forbidden lines are
photexcited away, and there is no density- or photoexcitation-dependence
in AtomDB. You might ignore the intercombination and resonance lines in
the triplets until you get a reasonable fit (which also might not be
possible with an isothermal plasma).  Then you could try the
photoexcitation models from
<http://space.mit.edu/cxc/analysis/he_modifier/>.


-- Dave

----
You received this message because you are
subscribed to the isis-users list.
To unsubscribe, send a message to
isis-users-request_at_email.domain.hiddenwith the first line of the message as:
unsubscribe