STScI Logo

Hubble Space Telescope
WFPC2 Photometry Cookbook

Converting from STMAG to Johnson UBVRI and Cousins RI

The magnitude of an object in a WFPC2 field can be determined using the photometric zeropoint information in the header of your calibrated image file (.c0h) using the normal formula:

	m = -2.5 x log10(COUNTS) + ZEROPT

where COUNTS is the number of counts sec-1 from your image and:

    ZEROPT = -2.5 x log10(PHOTFLAM) + PHOTZPT

where PHOTZPT, and PHOTFLAM are header keywords. However, the resulting magnitude will be in the STMAG system (see p. 15 of the Synphot User's Guide), which is based on a spectrum with constant flux per unit wavelength. This is different than conventional photometric systems that use the spectrum of Vega to define the magnitude zeropoint.

The following table was generated using Synphot to provide rough conversions to the Johnson UBVRI and Cousins RI systems. Typical uncertainties are 5%, and probably much worse for the U filter. The correction depends on the spectrum of the object, hence the table was generated using a wide range of Bruzual models.

EXAMPLE: You want to convert to the Cousins I band for an object on WF4. Using the STSDAS command "imhead image.name" shows that:

     PHOTZPT = -21.1	(Note: This is always the same.)
     PHOFLAM = 2.6044 X 10-18

Hence:

   ZEROPT(STMAG) = -2.5 x log10(2.6044 X 10-18) - 21.1
                 = 22.861

The object you are interested in has a spectrum of an A0 V star, so the table indicates that:

     (Cousins I) - F814W = -1.21,

hence:

  
     ZEROPT(Cousins I)   = 22.861 - 1.21 
			 = 21.651

You can now convert from WFPC2 counts to magnitudes in Cousins I using:

     Cousins I = -2.5 log10(counts) + 21.651,

hence an object with 137 counts/sec has a value of:

     Cousins I = 16.309 mag.

Using this zeropoint would be appropriate when COUNTS is in units of counts/seconds. Some software packages expect a zeropoint which is in counts/exposure, hence if our exposure time were 600 seconds, the zeropoint would be 21.651 + 2.5 log10(600) = 28.596.


CONVERSION FROM STMAG TO JOHNSON UBVRI AND COUSINS VI:

       1. JOHNSON SYSTEM:

     U-F336W	  B-F439W     V-F555W    R-F675W    I-F814W
     -------      -------     -------    -------    -------
O5 V   0.53        0.67        0.05       -0.67      -1.11
B0V    0.46        0.66        0.05       -0.67      -1.13
A0 V  -0.08        0.67        0.02       -0.68      -1.22
F2 V  -0.03        0.62	      -0.00       -0.69      -1.28
G0V   -0.02        0.58       -0.01       -0.70      -1.31
K0 V  -0.10        0.53       -0.01       -0.69      -1.32
M0 V  -0.04        0.43       -0.00       -0.78      -1.48
M6 V   0.05        0.29       -0.03       -1.05      -1.67

       2. COUSINS SYSTEM:
                                         R-F675W    I-F814W
                                         -------    -------
O5 V                                      -0.71      -1.22
B0V                                       -0.70      -1.22
A0 V                                      -0.67      -1.21
F2 V                                      -0.63      -1.22
G0V                                       -0.60      -1.23
K0 V                                      -0.58      -1.23
M0 V                                      -0.54      -1.22
M6 V                                      -0.56      -1.21

NOTES:

1. The zeropoint for the STMAG system has been set to roughly match the Johnson system at V, hence the corrections are nearly zero at V.

2. The Cousins I filter is much closer to the F814W filter than Johnson I, as shown by the nearly constant correction as a function of spectral type (i.e., color term). Here is a Synphot script that plots the three passbands for comparison.

stsdas
hst_calib
synphot
plband "band(wfpc2,1,a2d7,f814w)" left=6000 right=12000 
      normali=yes ltype=solid device=stdgraph 
plband "band(johnson,i)"  normali=yes ltype=dashed 
      append=yes device=stdgraph 
plband "band(cousins,i)"  normali=yes ltype=dotted 
      append=yes device=stdgraph 

3. In some cases you will need to run synphot yourself (e.g., different filter, different spectrum, etc.). Here is the command used to generate the first entry in the table.

stsdas
hst_calib
synphot
calcphot  "band(johnson,u)"           crgridbz77$bz_1 vegamag 
calcphot  "band(wfpc2,1,a2d7,f336w)"  crgridbz77$bz_1 stmag 

Where "band(wfpc2,1,a2d7,f336w)" defines the F336W filter, crgridbz77$bz_1 defines an O5 V spectrum from the Bruzual library, and "band(johnson,u)" defines the Johnon U filter. The difference between stmag and vegamag is the result in the table.

4. This will provide typical accuracies of about 0.05 mag (much worse in the UV). More accurate photometry will require a variety of "corrections" (e.g., CTE effect, contamination and red leaks for the UV filters, variable gains on different chips, color terms, geometric distortions) which are discussed in detail in a paper submitted to PASP by the WFPC2 Investigation Definition Team. The new paper is by Holtzman et al., and is not yet in final form so it is still subject to revision.

5. Another approach to determining your zeropoint is using ground-based aperture photometry of objects in your WFPC2 field of view.

- Brad Whitmore (whitmore@stsci.edu)