Procedure

Figure:Flow-chart representing how hyperz works in its standard configuration.

The procedure followed in the code hyperz is illustrated in Figure 6. The first step consists in reading the observed magnitudes, with the corresponding errors and the limiting magnitudes for non detected objects. Take care that to obtain reliable results, the photometry must be carried out in all filters within the same aperture, i.e. the flux must be belong to the same physical region. If requested, a dereddening for the Galactic extinction is applied to the photometric catalogue. Hence the magnitudes and the errors are transformed in fluxes. On the other side, there is the handling of the template SEDs. The program builds an hypercube, consisting of the fluxes derived from the reference spectra in the following way. If the spectra are taken from the GISSEL library they consist of 221 age records. To save CPU time and memory, hyperz provides a rebinning, to reduce the number of considered ages to 51, with the values showed in Table 4. In the ``standard'' configuration 5 spectral types are considered, corresponding to the Burst, E, Sa, Sc and Im illustrated in the previous section. More SEDs can be redundant. The spectra are reddened following an extinction law chosen from those listed in Section 2.3, and then depressed with the Lyman forest, according to the considered redshift step. At this point, the spectra are convolved with the filter response functions (plus detector if needed, see Section A), to obtain the expected fluxes.

0.25cm

Table:Ages corresponding to the 51 records selected from the 221 available in the GISSEL library.

 

n	age [Gyr]	n	age [Gyr]	n	age [Gyr]	n	age [Gyr]
1	0.00032	14	0.02630	27	0.50880	40	8.50000
2	0.00100	15	0.03000	28	0.71870	41	9.50000
3	0.00158	16	0.03300	29	1.01519	42	10.5000
4	0.00251	17	0.03600	30	1.43400	43	11.5000
5	0.00398	18	0.03900	31	1.70000	44	12.5000
6	0.00631	19	0.04500	32	2.00000	45	13.5000
7	0.01000	20	0.05250	33	2.30000	46	14.5000
8	0.01148	21	0.06405	34	2.60000	47	15.5000
9	0.01318	22	0.09048	35	3.50000	48	16.5000
10	0.01514	23	0.12780	36	4.50000	49	17.5000
11	0.01738	24	0.18053	37	5.50000	50	18.5000
12	0.01995	25	0.25500	38	6.50000	51	19.5000
13	0.02291	26	0.36020	39	7.50000

After the hypercube construction, the program proceeds with the$\chi^2$ calculation, considering all redshifts, spectral types, ages and absorption values. Minimizing the reduced $\chi^2$ the program finds the better photometric redshift solution and the corresponding SED. Different output files can be chosen by the user, containing different informations, as the best $z_{\rm phot}$, the observed SEDs (mean fluxes and errors), the best-fit integrated SEDs and spectra, the minimum $\chi^2$ as a function of redshift and the secondary solutions.
 

2000-12-10