title test optically thin model that extends to very low densities
c
c commands controlling continuum =========
table agn
ionization parameter 0
c
c commands for density & abundances =========
hden -2 to the -2 power
init file="ism.ini"
c
c commands controlling geometry  =========
sphere
c this will take density down to 1e-8, below stated limit
radius 10 13 
c
c other commands for details     =========
c
c commands controlling output    =========
save performance "limit_lowden.per"
save overview "limit_lowden.ovr" last
c
c commands giving the monitors    =========
// this model is optically thin in the Lyman lines and so pumping
// is very important - 
// >>chng 00 dec 06, change ionization parameter of model from 1 to 0
// old value was on unstable part of cooling curve, and small changes in
// cs of Fe13 lines changed temperature by 150% - with lower value
// results are themally stable
// >>chng 07 jan 25, from -19.014 to -19.045, use only Gu UTA data file
// do not assert H lines since cloud is optically thin and takes
// at least three iterations to converge optical depth scale, t
// that is not the purpose of this sim
// assert line luminosity "totl" 4861 -19.045 error 0.07
//
// >>chng 04 jun 12, from 0.5998 to 0.552 chng cont shield fcn
// >>chng 06 nov 29, rm since Badnell DR raises temp above case b table limit
// assert line "ca b" 4861 0.552
//
// >>chng 03 nov 23, from 4.410 to 4.431, Kraemer et al. guess of dr
// >>chng 06 feb 03, from 4.431 to 4.453, guess dr to Badnell mean
// >>chng 06 nov 29, from 4.453 to 4.480, Badnell DR now default
monitor temperature hydrogen 2 4.480
//
// model is highly spherical so these are not degenerate tests
// large error is to accomodate usual limited precision unix box
monitor ionization hydrogen 1 volume -5.359 error 0.25
//
// chng 00 jul 7, to get sgi and pc to agree
monitor ionization hydrogen 1 radius -5.359 error 0.2
//
// >>chng 06 feb 03, from -3.333 to -3.358, guess dr to Badnell mean
// >>chng 07 jan 25, from -3.358 to -3.384, use only Gu UTA data file
monitor ionization helium 2 radius -3.384
//
monitor ionization Carbon 5  -0.175 
//
// >>chng 07 jan 25, from -0.520 to -0.497, use only Gu UTA data file
monitor ionization Carbon 6  -0.497
//
// >>chng 06 feb 03, from -1.600 to -1.571, guess dr to Badnell mean
// >>chng 06 nov 29, from -1.571 to -1.542, Badnell DR now default
// >>chng 09 feb 24, from -1.542 to -1.52, botches due to r2763
monitor ionization Carbon 7  -1.52
//
// >>chng 06 feb 03, from -1.394 to -1.420, guess dr to Badnell mean
// >>chng 08 oct 18, from -1.42 to -1.446, include Badnell UTA data
monitor ionization Nitrogen 5 -1.446 
//
monitor ionization Nitrogen 6 -0.087
//
// >>chng 06 nov 29, from -0.870 to -0.847, Badnell DR now default
monitor ionization Nitrogen 7 -0.847
// 
// >>chng 06 nov 24, from -1.513 to -1.488, turn on Badnell RR
// >>chng 06 nov 29, from -1.488 to -1.440, Badnell DR now default
monitor ionization Oxygen   5 -1.440
//
// >>chng 06 feb 03, from -0.700 to -0.724, guess dr to Badnell mean
// >>chng 08 oct 18, from -0.724 to -0.749, include Badnell UTA data
monitor ionization Oxygen   6 -0.749
//
monitor ionization Oxygen   7 -0.144
//
// >>chng 06 jul 12, from -1.300 to -1.298, use Savin & Gu DR to make mean DR
// >>chng 06 aug 17, from -1.298 to -1.277, update to Badnell dr
// >>chng 08 oct 18, from -1.277 to -1.256, include Badnell UTA data
monitor ionization Oxygen   8 -1.256
//
c limit_lowden.in
c class limit 
c ========================================
c 

This model is optically thin, with density falling off as inverse square law,
so ionization and temperature should be nearly constant.
if outer radius increased by 2 dex problem with level3 will appear,
several li seq lines (OVI, NeVIII) will fluctuate when density about
1e-9

We do not assert H lines since the cloud is optically thin and takes
at least three iterations to converge optical depth scale, 
That is not the purpose of this sim