Wilkinson Microwave Anisotropy Probe
The data made available through this page has been updated. The most recent version of this data may be accessed through /product/map/current/
Time-Ordered Data (archive with attitude, etc.)
Two complete time-ordered data archives are supplied with the latest WMAP data release; one contains calibrated data while the other contains uncalibrated (or raw) data. The uncalibrated archive contains all four radiometer channels for each observation; the calibrated archive has been coadded into two radiometer channels to reduce their size.
These data are provided through a set of compressed tar files. Each tar file contains twenty files; each of these files contains one day of WMAP data written in five binary FITS tables. These tables contain:
For a more complete description of the time-ordered data and the format of these data files, see the WMAP Explanatory Supplement. A sample of each FITS header is available here. IDL routines to read these files are available from the WMAP Software page. Given some recent confusion regarding time tags assigned to the data, we provide a more detailed description of the telemetry timing in the following paragraphs.
Each of the five telemetry segments tabulated above contain unique time stamps that are used to collate the data and to place it correctly on the sky. These data segments are collated from a variety of input telemetry streams, each of which have their own time stamps that were synchronized to a Master spacecraft clock.
The pointing and position data in the Meta-Data segment were collated from separate spacecraft telemetry streams that were recorded at 1 sec and 128 sec intervals, respectively, incommensurate with the instrument data. For user convenience, we have interpolated the pointing (quaternion) data to the start times of the science data major frames, recorded every 1.536 sec, and we have packaged them into 30+3 quaternions recorded every 46.08 sec. An IDL routine that interpolates the quaternion data to the center of each individual science observation, taken every 51.2 ms for W band and longer for the lower frequency bands, is available from the WMAP Software page. For bookkeeping purposes, the time stamp in the Meta-Data packet is copied from the AIHK packet, but it does not correspond to the start time of the first quaternion in that packet. The quaternions are specifically collated to the start of each science major frame.
The figure shows how times within the telemetry stream are collated in a time-ordered archive record (left) and how to interpolate the quaternions to obtain spacecraft attitude at the center of a given science observation (right). The left panel represents the incoming telemetry streams in light blue; the lines represent the start time of each telemetry record. The tan columns represent the data interpolated from the ephemeris and attitude telemetry stream; the ephemeris data is interpolated to the start of each 46.08 sec archive record while the attitude data is interpolated to the start time of each 1.536 sec science frame. These interpolated data are stored in the Meta Data table. The red polygon shows the contents of one archive record while the pale polygons represent the previous and next records; note the overlap in the stored quaternion data. The right panel represents a subset of the archive record, showing the individual K band observations within a few of the science frames stored in an archive record and the associated quaternions. Four of these quaternions (the two previous and the two following) are used to interpolate the attitude for a given observation. The WMAP interpolation software computes a quaternion given the dimensionless offset (ranging from 0-3) from the first of the four quaternions to the center of the observation of interest.
In addition to the spacecraft pointing information provided by the quaternions, separate line-of-sight vectors for each differencing assembly are required to place individual science observations on the sky. These are provided in the Line-of-Sight Vectors record, which contains twenty constant unit vectors, in spacecraft coordinates, for the A- and B-side bore-sights of the ten differencing assemblies. Each daily TOD file contains one such record for convenience: it is repeated throughout the mission. It is important to note that these vectors were determined by centroiding the Jupiter observations for each beam assuming the spacecraft pointing conventions discussed above. This self-consistent analysis provides immunity to any small timing offsets, since the spacecraft motion is dominated by a fast spin about its Z-axis. Thus, any small timing offset between pointing and science records would manifest themselves as a uniform shift in the azimuth of the bore-sight vectors. Further, we have verified that we obtain consistent bore-sight vectors (at the 10-20 arcsec level depending on DA) using Saturn observations, and that these bore-sight vectors do not drift with time over the first seven years of the mission.