MHO_AdhocPhaseCorrection

Purpose

MHO_AdhocPhaseCorrection applies a time-dependent (and optionally channel-dependent) phase correction exp(-i*zeta) to visibility data. The operator supports three modes: sinewave, polynomial, and file-based. The phase correction zeta is computed per (channel, accumulation period) and applied uniformly across all polarization products and spectral points.

Control File Trigger

• Keyword: adhoc_phase

• Category: calibration

• Priority: 3.5

Parameters for adhoc_phase (sinewave mode)

Parameter	Type	Description
algorithm_type	string	Must be sinewave.
adhoc_amp	double	Sinewave amplitude in degrees.
adhoc_period	double	Sinewave period in seconds.
adhoc_tref	double	Reference time in seconds past the most recent hour.

Parameters for adhoc_phase (polynomial mode)

Parameter	Type	Description
algorithm_type	string	Must be polynomial.
adhoc_poly	list_real	1-6 polynomial coefficients in deg/s^n.
adhoc_tref	double	Reference time in seconds past the most recent hour.

Parameters for adhoc_phase (file mode)

Parameter	Type	Description
algorithm_type	string	Must be file.
adhoc_file	string	Per-station: path to the ASCII phase file.
adhoc_file_chans	string	Per-station: channel-label string defining column order.

Input Data

This operator acts on the visibility_type container.

Algorithm

The operator performs the following steps:

Initialization (``InitializeInPlace``):

1. If the mode is NONE, the operator is a no-op.

2. Extract the scan start time from the visibility container’s start metadata tag. Convert to fractional days since the beginning of the year, and compute the scan start in seconds past the most recent hour.

3. Derive the accumulation period duration from the time axis.

4. For file (PHYLE) mode only: load and parse both the reference and remote station adhoc phase files. Each data line contains a fractional day time followed by one phase value (in degrees) per channel. Comment lines (non-numeric first token) are skipped. If a file has only one data row, it is duplicated to ensure a valid interpolation interval.

Execution (``ExecuteInPlace``):

1. Iterate over every frequency channel in the visibility container.

2. For each channel, retrieve the channel_label (fourfit freq-code character). Channels without a label are skipped.

3. For each accumulation period (AP):

   1. Compute the AP center time in seconds from scan start: t_center = t_AP + 0.5 * t_acc.

   2. Compute the phase correction zeta (in radians) by calling ComputeZeta with the channel label and AP center time (details below).

   3. Construct the phasor Phi = exp(-i*zeta).

   4. Multiply all visibility values for all polarization products and spectral points at (channel, AP) by Phi.

Phase Computation (``ComputeZeta``):

The time argument used in the sinewave and polynomial models is:

\[\tau = t_{\rm center} + t_{\rm scan\_start\_fpday} \cdot 86400 - t_{\rm ref}\]

where \(t_{\rm ref}\) is the adhoc_tref parameter. This (legacy) convention measures time in seconds since the beginning of the year, minus the seconds-past-the-hour reference.

• Sinewave mode:

  \[\zeta = A \sin\!\left(\frac{2\pi \tau}{P}\right)\]

  where \(A\) is the amplitude (converted from degrees to radians by the builder) and \(P\) is the period in seconds.

• Polynomial mode:

  \[\zeta = c_0 + c_1 \tau + c_2 \tau^2 + c_3 \tau^3 + c_4 \tau^4 + c_5 \tau^5\]

  where coefficients \(c_0, \dots, c_5\) are in rad/sⁿ (converted from degrees by the builder). Up to 6 coefficients are used; missing entries default to zero.

• File mode:

  1. Convert the AP center time to fractional days since BOY: t_fpday = t_scan_start_fpday + t_center / 86400.

  2. For each station, look up the channel’s freq-code character in the station’s channel string to find the column index.

  3. Clamp t_fpday to the file’s time range, then find the bounding interval [n-1, n] by linear scan.

  4. Linearly interpolate the phase (in degrees) between rows n-1 and n:

     \[\phi_{\rm deg} = \frac{t_{\rm bound}(\phi_b - \phi_a) - t_a \phi_b + t_b \phi_a}{t_b - t_a}\]

  5. Convert from degrees to radians.

  6. The differential phase correction is \(\zeta = \phi_{\rm ref} - \phi_{\rm rem}\).

Effect on Data

For each (channel, AP) combination, all visibility values across all polarization products and spectral points are multiplied by a complex phasor exp(-i*zeta), where zeta is the mode-dependent phase correction in radians. In sinewave and polynomial modes, zeta is the same for all channels. In file mode, zeta varies per channel and is computed as the differential phase (reference minus remote) interpolated from station-specific phase files.