INNAME Input UV file name (name) INCLASS Input UV file name (class) INSEQ 0.0 9999.0 Input UV file name (seq. #) INDISK 0.0 9.0 Input UV file disk unit # SOURCES Source list ' '=>all. CALSOUR Reference source list. Only one is allowed STOKES Stokes type to process SELBAND Bandwidth to select (kHz) SELFREQ Frequency to select (MHz) FREQID Freq. ID to select, 0=>all BIF 0.0 100.0 Lowest IF number 0=>all EIF 0.0 100.0 Highest IF number 0=>all TIMERANG Time range to use. ANTENNAS Antennas to correct. SUBARRAY 0.0 9999.0 Subarray; 0 => 1. GAINVER Input CL table 0=>high GAINUSE Output CL table: not = GAINVER -> high+1 OPCODE Operation code. CLCORPRM Parameters (see HELP DFCOR). BADDISK 0.0 9999.0 Disks to aviod for scratch CALIN Input file with the list of antennas, times and relevant atmosphere vertical delay. Used only with OPCODE='ATMO'

DFCOR Task: This task makes a differential corrections to a CL table. Adverbs: INNAME.....Input UV file name (name). Standard defaults. INCLASS....Input UV file name (class). Standard defaults. INSEQ......Input UV file name (seq. #). 0 => highest. INDISK.....Disk drive # of input UV file. 0 => any. SOURCES....list of sources to process. '*' = all; a "-" before a source name means all except ANY source named. CALSOUR....Reference source list. Only one is allowed STOKES.....The desired Stokes to correct. 'R', 'L', 'I', ' '=> all available FREQID.....Frequency identifier to select (you may determine which is applicable from the OPTYPE='SCAN' listing produced by LISTR). BIF........First IF to process. 0=>all. EIF........Highest IF to process. 0=>all higher than BIF TIMERANG...Time range of the data to be used. In order: Start day, hour, min. sec, end day, hour, min. sec. Days relative to ref. date. ANTENNAS...A list of the antennas to be modified. If any number is negative then all antennas listed are NOT to be modified. All 0 => use all. SUBARRAY...The subarray to modify. Do only one at a time. GAINVER....Input CL table version. If GAINVER is equal zero or greater than the total number of the CL tables then GAINVER is equal to the last existing CL table. GAINUSE....Output CL table version. If GAINUSE not equal GAINVER (after GAINVER default applied), then a new CL table is created and the input CL table version (GAINVER) is copied to version=(high+1). This adverb allows you to correct a CL table in place or to create a new one, but not to overwrite an existing one other than the input. OPCODE.....Operation code (see also EXPLAIN DFCOR) Right now only optype 'ATMO' works at the differential mode: 'POLR' => Modify Right-Left phase difference using phases in CLCORPRM (deg); up to 20 IFs may be processed at a time. Note that with 'POLR' the antenna table is also modifed. 'POLR' should only be used once. 'PHAS' => Rotate phase of residual gain by CLCORPRM (deg); up to 20 IFs may be processed at a time 'RATE' => Rotate phase of residual gain versus time. CLCORPRM(1) degrees constant term. CLCORPRM(2) = rate of change of phase (degrees/day) CLCORPRM(3) - (6) = day, hr, min, sec at which the "zero" phase (CLCORPRM(1)) is specified. 'OPAC' => apply atmospheric opacity amplitude corrections using zenith opacity of CLCORPRM(1) nepers. 'ADEL' => Correct phases, delays and rates for neutral atmospheric delay. CLCORPRM(1) = total pressure (mbars) at station, NOT at sea level. CLCORPRM(2) = partial pressure of water. CLCORPRM(3) = Temperature (C) CLCORPRM(4) = Tropospheric lapse rate (K/km) (should be negative) CLCORPRM(5) = Height of tropopause (km) CLCORPRM(6) = Scale height of water vapor (km). 'GAIN' => Correct using polynominal gain curve for antenna gain as a function of the zenith angle (ZA) in degrees. correction = CLCORPRM(1) + ZA * CLCORPRM(2) + ZA * ZA * CLCORPRM(3) ... 'CLOC' => Correct residual delay and model parms for the effects of a linear clock drift at a particular antenna. CLCORPRM(1) = rate of change of station clock (nanosec/day) CLCORPRM(2) = clock value at the "zero" time specified by CLCORPRM(3)-(6) (nanosec) CLCORPRM(3) - (6) = day, hr, min, sec at which the "zero" clock (CLCORPRM(2)) is specified. CLCORPRM(7) : correction has three modes, if = 0 then the clock drift is added as a small correction and CLCORPRM(2) is ignored. if = 1 then the total correction set by the CLCORPRMS is added. if = 2 then the values present in the CL table are replaced by those defined by CLCORPRM(1)-CLCORPRM(6). 'PANG' => Add or remove parallactic angle corrections from CL table entries. CLCORPRM(1) > 0 => Add corrections CLCORPRM(1) =< 0 => Remove corrections 'PONT' => Correct for predictable pointing offset of an antenna. CLCORPRM(1) is the linear rate of change of antenna gain as the pointing drifts. 'IONS' => Make ionispheric Faraday rotation corrections using one of several models. CLCORPRM(1) = Model type: 1 = Chiu model, CLCORPRM(3) = Sunspot no. 'ANTP' => Correct antenna and/or source position; antenna / source corrections are values to be added to the old positions in meters / sec of arc. The antenna (only one) must be specifyed in the case of antenna correction. The source (only one) must be specified in the case of source correction. !!!!!!!!!!!!!!!!!!! ATTENTION !!!!!!!!!!!!!!!!!! Starting June 2001 DFCOR corrects the AN or/and the SU table, if the relevant correction of the antenna or/and the source is carried out. So the application of the corrected CL table is required to match the data. Once DFCOR has been run, then the CL table must not be deleted and the correction can be undone only by doing it again with opposite sign. If you might forget applying this, then use the task SPLAT to apply it immediately. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 = "X" correction in meters. 2 = "Y" correction in meters. 3 = "Z" correction in meters. 4 no longer used 5 = Correction in the picture plane towards the RA direction in sec of arc. So it is RA correction, muliplied by COS(DECL). 6 = Declination correction in sec of arc 7 no longer used 'PCAL' => MkIII manual phase cal; replace the gain correction in the CL table with unit vectors with phases given in CLCORPRM. CLCORPRM(1) corresponds to BIF etc. Phases are given in degrees. 'PCFX' => Patch up missing phase cals. CLCORPRM gives the expected relationship between phase cals and uses any non blanked values. 'SBDL' => Add a delay to the IF residual delays. Values given in CLCORPRM correspond to IFs BIF, BIF+1,... EIF in nanosec. 'MBDL' => Change the multiple band delay by introducing the corresponded slope at phase vs IF frequency dependence. Values given in CLCORPRM correspond to IFs BIF, BIF+1,... EIF in nanosec 'SSLO' => Correct the phase only for an incorrect frequency used to calculate the phase at the VLA. CLCORPRM(1) gives the frequency error in MHz. 'ANAX' => Correct the delay, rate and phase for antenna axis offset. CLCORPRM is an array of axis offsets in meters corresponding to the antenna list given. Note that with 'ANAX' the antenna table is also modifed. 'ATMO' => apply atmospheric delay corrections using zenith delay for the antennas and times given in the input file. The values of zenith delay should correspond to the desired correction. The corrections for the interferometer delay and phase are added (with sign) to the relevant columns of the CL table, which in turn are added to the data when calibration is applied. Example 1. The correlator did not provide any atmosphere correction. Since, the the atmospheric delays are positive, the corrections entered in the CALIN should be negative. Example 2. The correlator model used 210 cm for the zenith delay. The true zenith delay was 220cm. Thus, DFCOR has to make additional additive correction of -10cm, and the CALIN entry should -10cm. If CALSOUR.NE.'' then DFCOR reads the CL table twice. First time the data corresponded to the given calibrator (CALSOUR(1)) are read and the relevant delays and rate of delays recalculated from the input file of zenith delays. The found delays and rate of delays are stored at the array together with the arrays of time and antennas. Second time the data corresponded to the given source list (SOURCE) are read and the relevant delays and rates of delay are calculated from the input file of zenith delays. Then the stored delays and rates of delay for the calibrator are interpolated to the time of the CL table row and the interpolated values are subtracted of the calculated source delays and rate of delays. The differences are recorded to the CL table CLCORPRM...Parameters: see above. BADDISK....A list of disks on which scratch files are not to be placed. This will not affect the output file. CALIN......Input file (used at OPCODE = 'ATMO') The first (ascii) data record should specify NLINES, the number of the lines at the file. Then, NLINES records must follow, each with three numbers: The station number; The time in dd hh mm ss; for example 01 13 26 35.3 The zenith delay error in cm; The values given at the zenith delay error correspond to the desired correction. Therefore the given zenith delay error should be subtructed from the full zenith delay. The example of a line: 2 01 13 26 35.3 5.3 The line is for antenna 2; The time corrersponds to day=1; 13hours; 26 min; 35.3sec; The delay at zenith is 5.3cm

DFCOR: Task to apply corrections to a CL table corresponding to the difference of the source (SOURCE) and the reference source (CALSOUR(1)) Documentor: L.R. Kogan Related Programs: CLCOR, LISTR, SPLIT, TABED, LOCIT This task will compute various corrections and apply them directly to a calibration (CL) table. The operation to be done is determined by OPCODE. Details and/or additional information for the various models is given below. OPCODE='POLR' This option causes corrections to be made to remove the systematic phase offset between the right and left hand polarization systems. The phase offsets are passed in CLCORPRM in degrees and are typically determined from running LISTR on polarization calibrated data on a source with known polarization angle. Up to 10 IFs specified by BIF and EIF may be processed in a single run with the corresponding phase values in CLCORPRM(1) - CLCORPRM(10). Specified phase corrections are made to the left hand polarization. Because the feed polarization parameters must be modified as well as the Calibration table this option will cause the AN table corresponding to the specified subarray to be modified as well as the specified CL table. NB: the modification of both the CL and AN table are cumulative so a given correction should be made only once. Thus, if multiple CL tables in the same subarray are to be corrected OPCODE='PHAS' with STOKES='L' should be used for each IF separately for CL tables after the first run with OPCODE='POLR'. (If the AN table gets messed up, correct all relevant CL tables and then rerun PCAL; further runs of CLCOR to correct the R-L phase difference will be unnecessary.) OPCODE='PHAS' This causes the specified phases to be rotated by CLCORPRM(I) degrees. OPCODE ='OPAC' This causes the specified amplitudes to be corrected for atmospheric opacity using a zenith opacity of CLCORPRM(1) nepers. This operation does not modify the total model values. Exact values of the zenith opacity depend on the weather, especially at higher frequencies, but typical values are given in the following: 327 Mhz 0.007 610 Mhz 0.007 1.4 GHz 0.008 2.3 GHz 0.01 5.0 GHz 0.01 8.4 GHz 0.01 10 GHz 0.012 15 GHz 0.02 22 GHz 0.05 40 Ghz and up are highly dependent on the weather. OPCODE='ADEL' This causes the phases, delays and rates to be corrected for a model atmosphere. The model used is a two term round earth approximation. The parameters passed are CLCORPRM(1) the total atmospheric pressure in millibars at the station, NOT refered to sea level and CLCORPRM(2), the partial pressure of water vapor in millibars. Pressure in mm. of Hg. can be converted to millibars by multiplying by 1.33322. Measurments of Dew point can be converted to relative humidity using the following table: Relative Humidity ( percent) from temp and DP. DP depression Dew Point deg C temp-DP deg C -10 0 10 20 30

0 100 100 100 100 100 1 92 93 94 94 94 2 86 87 88 88 89 3 79 81 82 83 84 4 73 75 77 78 80 5 68 70 72 74 75 6 63 66 68 70 71 7 59 61 63 66 68 8 54 57 60 62 64 9 51 53 56 58 61 10 47 50 53 55 57 12 41 44 47 49 14 35 38 41 44 16 31 34 37 39 18 27 30 33 35 20 24 26 29 32 22 21 23 26 24 18 21 23 26 16 18 21 28 14 16 19 30 12 14 17 The partial pressure of water vapor is obtained by multiplying the relative humidity (as a fraction) times the vapor pressure of water obtained from the following table: Temp (C) Pressure(mbars) Temp (C) Pressure(mbars) -------- ---------------- --------- --------------- -40 1.29 -30 3.81 -25 6.35 -20 10.35 -15 16.55 -10 26.00 -5 40.17 0 61.05 5 87.23 10 122.78 15 170.36 20 233.78 25 316.72 30 424.28 35 562.29 40 737.59 Information from Handbook of Physics and Chemistry published by The Chemical Rubber Co. 46th ed. The tropospheric lapse rate is the rate at which the atmosphere cools with increasing height. The model used assumes that the temperature declines linearly to the tropopause and then is constant. Accurate values can be derived from radiosonde data or approximate values can be obtained from the following table. An accurate value of the height of the tropopause may be derived from radiosonde data; typical values can be obtained from the following table (Davis et al. 1985, Radio Science 20, 1593): Latitude Lapse rate (K/km) Height of tropopause(km) 30 N -4.7 to -5.9 16 45 N -6.5 11.2 60 N -3.9 8 The scale height of the water vapor may be determined from radiosonde data (default = 2.2 km). OPCODE='GAIN' This causes the specified amplitudes to be corrected by a factor determined from a polynomial gain curve as a function of zenith angle (in degrees). The coefficients are specified by CLCORPRM and up to 20 terms are allowed. This allows for correction of amplitudes for the zenith angle (i.e. elevation) dependent behavior of the antenna gain. The amplitudes are modified by: FACTOR = CLCORPRM(1) + ZA*CLCORPRM(2) + ZA*ZA*CLCORPRM(3) ... OPCODE='CLOC' This causes the residual delays and model values to be corrected for a linear clock drift at a particular antenna. The parameters needed are : 1) clock rate (nanosec/day) ...... CLCORPRM(1) 2) "zero" clock (nanosec) ........ CLCORPRM(2) 3) time of "zero" clock measurement ........ CLCORPRM(3) - CLCORPRM(6) 4) correction mode ........ CLCORPRM(7) if = 0 then the clock drift is added as a small correction and CLCORPRM(2) is ignored. if = 1 then the total correction set by the CLCORPRMS is added. if = 2 then the values present in the CL table are replaced by those defined by CLCORPRM(1)-CLCORPRM(6). Mode = 2 gives the user the opportunity to set the residual delay manually. OPCODE='PANG' This option add or removes the phase of the parallactic angle to/from the residual phases of the specified CL table entries. The AIPS polarization calibration routines expect that this correction has NOT been made to the raw data; if the parallactic angle corrections has been applied, then this option with CLCORPRM(1) .le. 0 will remove it. The parallactic angle correction may be added using CLCORPRM(1) .gt. 0. Note: the above definitions of applying or removing the parallactic angle correction assumes that the phase of the first polarization (RCP for VLA, VLBA) decreases with increasing parallactic angle. This involves the definition of RCP. If the opposite definition of RCP and LCP is used, then the sense of applying or removing the parallactic angle correction given above is removed. OPCODE='PONT' This causes the specified amplitudes to be corrected for any gross predictable pointing error. The antenna gain correction factor is specified by a time at which the pointing was set correctly (i.e. the antenna gain is correct) and a linear drift factor. The TIMERANG parameters define the time at which pointing was done, CLCORPRM(1) specifies the rate of change of antenna gain (per hour). The inverse of this will be applied to the CL gain entries in order to correct the amplitude error. OPCODE='IONS' This option causes ionispheric Faraday rotation corrections to be computed and applied to the CL table. Several methods are available for determining the Faraday rotation correction; the method selected is indicated by the values given in CLCORPRM. The group and phase delays introduced by the ionospheric plasma are also computed and applied to the CL table. CLCORPRM(1): Electron density model type: CLCORPRM(1) = 1 => Chiu model. The electron column density will be computed using a Chiu model using software provided by Chris Flatters. The model is described in Chiu, 1975, J. At. Terr Phys. 37, 1563). CLCORPRM(3) gives the sunspot number. Sunspot numbers may be obtained from Solar-Geophysical Data published by NOAA. Daily sunspot numbers are available and can vary considerably; actual daily sunspot numbers are preferable to smoothed, averaged or predicted values. The sunspot number for the day approximately 3 days before the observation is the appropriate value to use. Moderately accurate predicted monthly smoothed values are available a year in advance. For those who are in a hurry or don't care, the following table gives observed monthly averaged values until May 1989 followed by the predicted, smoothed values until 1990. The estimated uncertainty (90 percent) in the smoothed values is given in parentheses under the predicted value. International Relative Monthly Sunspot Numbers Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

1980 164 163 161 159 156 155 153 150 150 150 148 143 1981 140 142 143 143 143 142 140 141 143 142 139 138 1982 137 133 129 124 120 117 115 109 101 96 95 95 1983 93 90 86 82 77 70 66 66 68 68 67 64 1984 60 56 53 50 48 46 44 40 34 29 25 22 1985 20 20 19 18 18 18 17 17 17 17 17 15 1986 14 13 13 14 14 14 14 13 12 13 15 16 1987 18 20 22 24 26 28 31 35 39 44 47 51 1988 58 65 71 78 84 94 104 114 121 125 130 138 1989 142 145 150 153 157 163 166 169 176 181 183 185 ( 5)(11)(16)(19)(21)(23)(24) 1990 186 187 185 180 174 170 168 166 159 151 144 139 (26)(29)(31)(32)(31)(28)(27)(29)(30)(31)(29)(26) 1991 138 134 130 129 130 128 124 119 114 113 115 115 (27)(28)(29)(34)(35)(32)(29)(26)(22)(20)(21)(24) OPCODE='ANTP' This option will correct phases for an incorrect antenna and source position; delays and rates are also corrected. The corrections to be added to the old values for the x,y, and z components are given in meters as CLCORPRM(1), CLCORPRM(2) and CLCORPRM(3). If the coordinates are in a right handed system then CLCORPRM(4) should be 1.0 or larger. The corrections to be added to the source position at the picture plane are given in sec of arc as CLCORPRM(5) for RA direction and as CLCORPRM(6) for declination. The position of antennas are determined in different coordinate systems for VLA and VLBI. The X-axis is located in local meridian for VLA and in Greenwich meridian for VLBI and EVLA. So CLCORPRM(7) has to be >0 for VLA and =0 for VLBI/EVLA. Note: only one FQ ID is processed per run; if there are several FQ ID then DFCOR must be run several times. A correction for antenna position errors should be made before further calibration is done. First, run DFCOR with OPCODE='ANTP' using GAINVER=1; GAINUSE=2 to copy CL table 1 to version 2 and then correcting for the antenna position error(s). Then run CALIB applying CL table 2 (DOCAL=1; GAINUSE=2). Use SNPLT to examine the results to be sure that the correction was done properly. When using CLCAL the input CL table should be version 2 (GAINVER=2). If there are adequate observations of calibrators then LOCIT can be used to determine antenna position errors. OPCODE='PCAL' This option allows entering complex gain corrections directly into the CL table. This operation is mostly of use in the settiong of MkIII VLBI manual phase cals so only the phase can be specified. If the amplitudes also should be changed then TABED can be used to multiply the real and imaginary parts of the complex gains by the appropriate value. Note: the phases in the CL table are corrections to be added to the data and thus have the opposite sign from the Haystack convention. Up to 20 IFs can be entered with CLCORPRM(1) corresponding to IF BIF, CLCORPRM(2) to BIF+1 etc. Phases are given in degrees. Note: for some reason phase cals read by AIPS are 144 degrees more positive that those decoded by the Haystack software. To obtain the sedired results add 144 degrees to any manual phase cals used in the Haystack system. OPCODE='PCFX' This option is similar to 'PCAL' except that any unblanked phases are left unchanged. Any blanked phases are replaced by the value expected based on any unblanked phases and the relationship between the IF given in CLCORPRM. In the case that all phases are blanked then this option is the same as 'PCAL'. This option is intended for use when some phase cals are present for a given antenna but are missing for some IFs. This option allows using the good values and estimating the missing values. OPCODE='SBDL' This option allows adding values to the IF residual delays. This correction will not effect the current values of the total model delay and is intended primarily for correcting MkIII data for the antenna based difference between the multi-band and single-band delays. Individual IF values can be entered but it is usually assumed in the MkIII system that this is a constant value for all IF. In this case fill CLCORPRM with the desired correction. Values are given in nanoseconds and will be added to the current correction, i.e. subtracted from the data when this table is applied. The values in CLCORPRM correspond to IFs BIF, BIF+1, ... EIF. Having applied this correction the new CL table application will change the slope of phase frequency dependence inside the each IF. The change of the slope at each IF is determined by the value of corresponded parameter CLCORPRM(I). This option does not change the phase difference between different IF's. OPCODE='MBDL' This option changes the multiple band delay by introducing the corresponded slope at phase vs frequency dependence. The same phase is added for all channels of the given IF. Values of CLCORPRM are given at nanoseconds. The added phase is determined by the next formula FI[I]=TWOPI*CLCORPRM[I]*DF[I], where DF[I] is a difference between the frequency of the given IF and residual one. The values of multiple band delays CLCORPRM[I] can be different for different IF, although it is constant typically. The values CLCORPRM[I] for all IF have to be installed even if they are the same for the all IF. The delay corresponded to the first IF - CLCORPRM[1] is substituted at the modified CL table at MBDELAY1(2) collumn. OPCODE='SSLO' The VLA has four separate 'IFs'. In normal observing, the frequencies of these are locked together into pairs. Each pair forms an AIPS IF. In normal observing, the phase calculation, or finge stopping, is done correctly. However, for some spectral line observing, it is desirable to have the separate VLA IFs of a pair at the same center frequency, but with different bandwidths. In this case, the Signed Sum of the LOs is different and the two frequencies can not be locked together. The VLA has only sufficient hardware to control one IF of each pair, so the fringes for the other cannot be stopped correctly. In general, this leads to decorrelation. However, if the frequency difference is sufficiently small, it is merely results in a residual phase error, which changes slowly with time. This option is to allow a post facto correction to be made to the phase only. Note that the delay is calculated correctly. The VLA IFs that are affected are 'B' and 'C'. By convention, 'A' and 'D' are always fringe-stopped correctly. Since the error depends upon the antenna location, the possibility that the antenna coordinates are defined in a left- handed system is allowed for, even though the VLA uses a right- handed system. The correction for SSLO is calculated at each CL table entry time as follows: If A is the antenna position vector and S is the unit source position vector, then the calculated phase correction (dph) is dph = 2 * pi * [ integer part of ( A . S * dfq / c)] where dfq is the frequency correction in Hz calculated from CLCORPRM(1). The new value of the complex gain in the CL table (gnew) is calculated from the old value (gold) as: gnew = gold * exp (i * dph) or, in cartesian coordinates: xnew = xold * cos(dph) - yold * sin(dph) ynew = xold * sin(dph) + yold * cos(dph) where g = x + i * y OPCODE='ANAX' This option changes the delay, rate and phase in accordance with an antenna axis offset. CLCORPRM(I, I=1, NANTSL) is the axis offset of antenna I in meters. The AN table is modified also. Two type of antenna mounts (altazimuth and equatorial) are supported.