NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Data > PlanOfRecordC42004 > StandardObservingModes > HowToReduceOTFData

Changes | Index | Contents | Search | Statistics | Jump to Topic

Data Reduction Case #7

OTF Map

Reducing OTF Map data proceeds as follows:

1. Reduce the spectrum for each integration, as you would for a point source observation
2. Assign a position to each integration (usually it is the position at the midpoint in time for that integration)
3. Average adjacent integrations, if necessary, and assign the proper position to the result
4. Grid the data, interpolate if necessary
5. Convolve if necessary
6. Display maps

An OTF Map is nothing more than a series of spectra observed at a number of positions spaced on a regular grid. So, the first step in reducing an OTF Map is to calibrate and baseline subtract data for each integration. Follow the data reduction case appropriate for the switching scheme used, e.g. if the data were taken in frequency switched mode, then the calibration and reduction should follow HowToReduceFSTrackData.

If the data are taken in total power mode, then the OTF Map can be reduced as described in HowToReducePSwitchSingleBeamData, but the user must be careful to define which spectrum represents the "reference" for each signal integration. Namely, each total power observation on the grid acts as a "signal" spectrum. The reference position, then, may be a point located near the edge of the map, or it may be a position located off the map and observed after every N seconds, or it may be determined by a customized process defined by the observer. Regardless of how it is determined, the reference position can then be applied to multiple signal positions. For example, integrations 1-10, each a unique pointing on the grid, may each use integration 11 (which may be on or off the grid) as the reference in a total power position switched calibration.

Once the observer has a calibrated, baseline-subtracted, averaged, and flagged spectrum for each integration, he can grid the data and convolve with a smoothing function if necessary. Channel averaging may then be applied, and the user may wish to produce countour maps or grayscale maps for each channel, or for a range of channels which have been averaged.

-- JimBraatz - 28 Jul 2004