UV Filter Flux Calibration

PROCEDURE

Eight white dwarf stars were observed with the UVW1, UVM2, and UVW2 filters on the OM: BD+284211, BD+332642, BPM16274, G93-48, GD50, HZ2, HZ4, and LB227. All observations were processed with SAS version 5.4, and the flux of each star was calculated using the four different methods. Count rates (as produced by SAS v 5.4) for LBB227, BPM16274, HZ4, HZ2, G93_48, and GD50 in all three UV filters were provided to Bing Chen along with HST spectra. Bing used the count rates and HST spectra (downloaded from ESO) to calculate new spectral type ratios, new Vega spectral ratios, and new zero-points.

New calculations	UVW1	UVM2	UVW2
White dwarf spectral ratios	4.396e-16	2.027e-15	5.797e-15
Vega spectral ratios	4.6797e-16	2.0326e-15	5.2240e-15
Zero-points	17.284	15.868	15.053
Predicted results	within ~ 5%	within ~ 5%	within ~10%

OBSERVATIONS

Star	B mag	V mag	Spectral Type
BD+284211	10.22	10.54	White Dwarf
BD+332642	10.65	10.81	White Dwarf
BPM 16274	14.18	14.16	White Dwarf
G93-48	12.79	12.73	White Dwarf
GD50	13.82	13.98	White Dwarf
HZ2	13.81	13.86	White Dwarf
HZ4	14.62	14.47	White Dwarf
LB227	15.44	15.35	White Dwarf

One of the calibration targets, BD+284211, was discarded from the flux calculations. This star was brighter than most objects observed with the OM (B magnitude of 10.22 ), and the coincidence loss for observations in all filters was high. Many counts that should have been included as "source counts" were either subtracted as "background counts" or not included at all. In addition, the SAS did not identify any sources for BD+332642 in the UVW1 filter, so observations of this star were excluded from UVW1 calculations only.

FLUX CALCULATION METHODS

1. Magnitude to Flux

0.025 - Mag(UV) = -2.5 log [Flux(Vega)/Flux(source)]

Mag(UV) = MAG value from SAS source list file

Flux(Vega) = flux of Vega specific to each UV filter:

Flux(Vega) UVW1: 3.7391873E-09 erg/cm/cm/s/A

Flux(Vega) UVM2: 4.4609810E-09 erg/cm/cm/s/A

Flux(Vega) UVW2: 5.3884371E-09 erg/cm/cm/s/A

Flux(source) = UV flux of the target source

2. Count Rate to Flux

0.025 - (2.5 log (Count rate) + Zero-point) = -2.5 log [Flux(Vega)/Flux(source)]

Count rate = CORR_RATE from SAS source list file

Zero-point = zero-point value specific to each UV filter:

Zero-point UVW1 = 17.284

Zero-point UVM2 = 15.868

Zero-point UVW2 = 15.053

Flux(Vega) = flux of Vega specific to each UV filter, as above.

Flux(source) = UV flux of the target source

3. Vega Spectral Ratio to Flux

Count rate * Vega spectral ratio = Flux(source)

Count rate = CORR_RATE from SAS source list file

Vega spectral ratio = factor for Vega spectral type specific to each UV filter:

Vega spectral ratio UVW1= 4.6797E-16

Vega spectral ratio UVM2= 2.0326E-15

Vega spectral ratio UVW2= 5.2240E-15

Flux(source) = UV flux of the target source, as above

4. Source Spectral Ratio to Flux

Count rate * Source spectral ratio = Flux(source)

Count rate = CORR_RATE from SAS source list file

Source spectral ratio = factor for source spectral type (in case of all calibration stars, the spectral type is white dwarf) specific to each UV filter:

White dwarf spectral ratio UVW1= 4.396E-16

White dwarf spectral ratio UVM2= 2.027E-15

White dwarf spectral ratio UVW2= 5.797E-15

Flux(source) = UV flux of the target source, as above

Fluxes of Vega can be viewed on the SAS watchout page. The zero-points and spectral ratios were determined by Bing Chen and released in an email to the Vilspa OM calibration mail list on 27 February 2003.

COMPARISON OF OM FLUX TO HST FLUX

Once the UV flux had been calculated for each star in all three UV filters, the calculated OM flux was compared to the HST standard flux for each star specific to the effective wavelength of each UV filter: UVW1 = 291 nm, UVM2 = 231 nm, and UVW2 = 212 nm. HST standard fluxes were obtained from ftp.eso.org in the directory /pub/stecf/standards/hststan. The references for these fluxes are listed in the readme file as Bohlin, Colina, & Finlay, AJ, 110, 1316, 1995 and Bohlin & Lindler, STScI Newsletter, Vol. 9, No. 2, pl 19, 1992.

RESULTS

The magnitude to flux method calculates OM flux values that are within +/- 8.08% of the listed HST standard fluxes, with a standard deviation of 6.23 over all filters. This flux calculation method gives fluxes of the following accuracies for specific filters:

Flux Method	UVW1	UVM2	UVW2
Magnitude to flux / HST flux:	1.034	1.084	1.099
Std. Deviation:	0.00365	0.00827	0.0949
Count rate to flux / HST flux:	1.045	1.027	0.921
Standard Deviation:	0.00381	0.00784	0.00795
Vega spectral ratio to flux / HST flux:	0.985	1.014	0.842
Std. Deviation:	0.00348	0.00774	0.00727
Source spectral ratio to flux / HST flux:	1.048	1.017	0.917
Std. Deviation:	0.00371	0.00776	0.00791

CONCLUSIONS

Using the "source spectral ratio to flux" method provides the most accurate fluxes across all three ultraviolet filters. Because the CORR_RATE and CORR_BACKGROUND_RATE values have been corrected for coincidence and deadtime loss by SAS version 5.4, they provide more accuration calculations of flux than using the RATE and BACKGROUND_RATE values and applying coincidence and deadtime loss by hand.

White dwarf spectral ratios, Vega spectral ratios, and zero-points last updated 5 May 2003.