The GOODS-N field covers a roughly rectangular area of 10x16 arc minutes centered at about 12:36:55 +62:14:16 (2000.0) with the larger side oriented at a position angle of 45 degrees. See map for more details.

Instrument	DEIMOS @ Keck II
Observing modes	imaging and multi object spectroscopy
Instrument setup	Imaging: R filter Spectroscopy: 600 line grating + GG455 blocker filter, central wavelength 7200Å range=4600-9800 Å, resolution 4Å FWHM.
Mask Design	The slit masks were designed using software written by the DEEP Team at UC/Berkeley. Whenever possible, the slits were tilted to match the position angle of the object's major axis, or, in the case of two objects close on the sky, at a position angle where both objects would have their spectra taken. The prime catalog used in the mask design was magnitude limited at approximately R ~ 24.3 (in the AB system).
Observations	Imaging : 11 pointings, 300 sec exposure time each Spectroscopy: 18 masks, 3 x 1200 sec time each

Observing nights	March 26, May 2,3,4,5, 6 and 28, equivalent to two Director’s Discretionary Time nights + 8 masks observed using the DEEP team time. Observing log.
Object catalog	The object catalog was generated using Sextractor with the following parameters. Total number of objects (slits): 1909 Accurate Astrometry for the catalog (RMS of 0.137 arc sec in RA and 0.149 in DEC.) was obtained cross-correlating the USNO and the Sloan Digital Sky Survey (SDSS) catalogs.
Data Reduction	Imaging: the TKRS catalog was photometrically calibrated using objects classified as stellar both by the SDSS pipeline and by Sextractor for the DEIMOS imaging data. For a total of 130 objects, the RMS difference is 0.36 magnitudes. After the standard processing steps (bias subtraction, flat-fielding, etc) all images were astrometrically calibrated using stars in the Naval Observatory catalog, and subsequently combined into a mosaic. For elongated objects we also compared the position angles measured by Sextractor with independent measurements obtained by UCSC DEEP Team in the Hubble Deep Field and Flanking Fields (available through their archive). For galaxies brighter than R=23.5 and with major axes greater than 1.2 arc sec, the measurements agree to better than 10 degrees. Spectroscopy: we used the IDL pipeline developed for DEIMOS by the UC Berkeley Deep team. Three packages were used: spec2d takes a set of science images, plus corresponding arcs and flats, and extracts 1-D sky-subtracted spectra from 2-D images. spec1d was used to analyse each spectrum to estimate redshift and spectral type. zspec was used to manually check each redshift estimate. Zspec is an IDL-based GUI designed for evaluating the accuracy of the automatically determined redshifts outputs from spec1d. Zspec was written by Darren Madgwick at UCB.
Data products	Catalog of redshifts for 1810 objects. Quality of each measurement has been independently classified by at least two persons. Quality codes range from 4 (99.9% certainty) to 1 (not usable). Stars are classified as –1.

“Pie diagram” showing the spatial distribution as a function of redshift for all galaxies with good redshifts in our survey. Note the numerous walls and voids observed in this field due to clustering of galaxies.

The redshifts were measured using emission lines (Balmer Series, [NII], [SII], [OIII], [OII]) and the Calcium (H, K and other ) absorption features. Each redshift measurement was assigned a “quality” identifier according to at least two independent measurements . In the center the quality distribution is shown.

Quality codes were assigned on the basis of a set of criteria: a value of 4 is assigned to a “rock-solid” z measurement obtained from a spectrum with at least two clear features. A value of 3 indicates a good z value, with a 90% probability to be correct. At least two features are still required but ID is more doubtful than Q=4. A value of 2 indicates a non-usable measurement for the present catalog but the spectrum will be later revisited because a useful z value might be extracted from it. Q=1 means that the redshift is not usable and the data will never yield one. Q=-1 is reserved for stars.

This plot shows the z-magnitude diagram: magnitudes were obtained from Sextractor.