FIRPO MESSAGE #8
Last modified on August 24, 1997 by
A. Richichi
The FIRPO photometric system
Jump at the end if you just want the conclusions!
In the nights of August 11-16, extensive photometry was carried out
both at TIRGO and with FIRPO at the 1.23m on CA. A total of
17 sources
happened to be observed at both sites in one or more filters.
This gives a nice opportunity to check the quality of the
FIRPO photometry against that of TIRGO.
A first analysis of the mag differences at the two sites gives
the following result:
diffj.dat: av = -4.00000E-03 -- sig = 0.194227
diffh.dat: av = 2.10000E-02 -- sig = 0.104924
diffk.dat: av = 0.12412E+00 -- sig = 0.193666
However, it is noteworthy to look at a
plot of
these differences against the spectral type.
Some editing has been done: a) a cluster of stars near K4.5-K5 in K
seems to be discrepant; it turns out that 3 of them were observed at
TIRGO at high airmass (3.5) and it is likely that the extinction correction
was not very good up to this limit. The fourth one was centered in the
IR without TV camera (dayight), Sam reports it is a binary. So these
4 stars were discarded. b) the A5 star seems uninteresting (although
more or less in agreement with G5), also discarded.
There result can be seen in this other
plot,
where the spectral types have been converted to temperatures.
It appears that although a constant fit is also not too bad, some
slope gives a better result. The values are
Program POLYFIT V 10/Jul/90 (A.Richichi) - 24-Aug-1997 13:38:25
File: diff_temp_jc.dat
Pol. degree: 1- N = 10- [ 2786.00- 4095.00] - [ 0.00- 1194.00]
1 -.10304E+01 0.26412E+00
2 0.29958E-03 0.74406E-04
sdev: 0.16063E+00 chi2:0.24486E+01
Program POLYFIT V 10/Jul/90 (A.Richichi) - 24-Aug-1997 13:42:59
File: diff_temp_hc.dat
Pol. degree: 1- N = 10- [ 2786.00- 4095.00] - [ 0.00- 1194.00]
1 -.43327E+00 0.16240E+00
2 0.13077E-03 0.45077E-04
sdev: 0.10084E+00 chi2:0.34024E+01
Program POLYFIT V 10/Jul/90 (A.Richichi) - 24-Aug-1997 13:44:31
File: diff_temp_kc.dat
Pol. degree: 1- N = 10- [ 2786.00- 4095.00] - [ 0.00- 774.00]
1 -.46648E+00 0.23764E+00
2 0.19942E-03 0.68655E-04
sdev: 0.14413E+00 chi2:0.29095E+01
Program POLYFIT V 10/Jul/90 (A.Richichi) - 24-Aug-1997 13:48:26
File: diff_temp_kd.dat
Pol. degree: 0- N = 3- [ 4095.00- 5085.00] - [ 295.00- 990.00]
1 0.21865E+00 0.40645E-01
sdev: 0.11161E+00 chi2:0.28824E+01
At K, the fit for stars hotter than K5 (4100K) was done with a constant.
How to interpret this values?
A definitive conclusion can be reached only for the K filter, since we
can compare the
transmission
curves for FIRPO and TIRGO. The K3 filter inside FIRPO is much
wider than the one at TIRGO. As a result, when we are in the
Rayleigh-Jeans regime, it tends to give an effective wavelength
shorter than at TIRGO. At CA, our standard stars are generally bluer than the
target stars, and we can expect that
the K magnitudes with FIRPO are fainter than the
true ones. This is shown well in the plot. The slope is due to the
fact that this effect is a function of the redness of the target star.
As soon as the target stars become less red, we can expect the
slope to become zero.
For the other filters, the situation could be the same although we
do not have a plot of the transmission curves of the FIRPO J and H filters.
In any case, the color difference between standards and targets still
remains.
In conclusion, the FIRPO JHK magnitudes will need to be corrected, and
at Arcetri we are in the process of rebuilding the FIRPO photometric
archive. It is my impression that without an extensive program of
standard stars, it will be hard to reach the intrinsic accuracy,
which is generally of 0.02-0.05 mag in the 3 filters. More likely,
we will have to deal with uncertainties of about 0.10 mag
unless the
observer takes care to use a standard star of the same color as the
target star.
For the future, we should plan to carry out at least another night
of common observations from TIRGO and FIRPO (Jan 98?). As a final
recommendation, pls always fill the "official" logsheets, and try
to conform to a standard pattern for the file names.
Andrea (with help from Sam)