Using The Sdss Spectral Data Of Star Formation Of Galaxy Metal Abundance

Our aims are to study the oxygen abundance of the star-forming galaxies , to study the estimating methods and the calibration relations. We select a large sample of ~40,000 star-forming galaxies from SDSS DR2 database as the metal-rich sample I, and select 531 star-forming galaxies from the SDSS DR4 database with strong emission lines, including [OIII]λ4363 detected at a signal-to -noise ratio higher than 5σ, as well as 164 galaxies and HII regions from the literature with electron temperature measurements. They are our metal-poor sample II.We determine the gas-phase oxygen abundances for the sample II of 695 galaxies and HII regions with reliable detections of [OIII]λ4363, using the reliable and direct temperature-sensitive (Te) method of measuring metallicity. We compare our (O/H)Te measurements of the SDSS sample with the abundances obtained by the MPA/JHU group who used the photoionization models(Charlot & Longhetti 2001) Bayesian technique. For roughly half of the sample the Bayesian abundances are overestimated by ~ 0.34dex, possibly due to the treatment of nitrogen enrichment in the models they used. The R23 and R23-Pmethods systematically overestimate the O abundance by a factor of ~0.2dex and ~0.06dex, respectively. The N2 index rather than the O3N2 index provides more consistent O abundances with the Te-method, but with some scatter.The observed relations between [NII]/Hα, [OIII]/[NII], [NII]/[OII], [NII]/[SII],[SII]/Hα,[OIII]/Hβand log(R23) from sample I are consistent with those expected from the photoionization models of Kewley & Dopita(2002). However, most of the data spread in a range of ionization parameter q from 2×107 to 8×107 cm s-1, narrower than that predicted by the models. The relations of N2, O3N2 and S2 with log(O/H)Te are consistent with the photoionization model calculations of Kewley & Doptita (2002), but R23 does not match well. Using a large sample of ~40,000 star-forming galaxies selected from SDSS DR2, we derive analytical calibrations for oxygen abundance from several metallicity-sensitive emission-line ratios: [NII]/Hα, [OIII]/[NII],[NII]/[OII],[NII]/[SII] and [OIII]/Hβfor the metal-rich galaxies(8.5<12+log(O/H)<9.3).This consistent set of strong-line oxygen abundance calibrations will be useful for future abundance studies. Among these calibrations, [NII]/[OII] is the best for metal-rich galaxies due to its independence on ionization parameter and low scatter. Dust extinction must be considered properly at first. We also derive analytical calibrations for log(O/H)Te from R23, N2, O3N2 and S2 indices on the basis of the sample II, including the excitation parameter P as an additional parameter in the N2 calibration. We also estimate the nitrogen abundances of the large sample ~ 40,000 galaxies, which are very useful to reflect the origins of primary and secondary component of nitrogen element.