Frequency Determination on Alouette & ISIS Digital Ionograms

(4 May 2004 summary prepared by RFB from files created by the late William B. Schar; updated by RFB on 22 Dec 2006 based, in part, on files created by Daniel N. Emery)

1. Introduction

The process of analog-to-digital (A/D) conversion of Alouette and ISIS topside ionograms started with the ISIS-2 analog telemetry tapes. One of the challenges was to identify the frequency markers superimposed on the receiver video output and to interpret between these markers to obtain the frequency of the intervening receiver amplitude scan lines. Based on the experience gained with the ISIS-2 data, the procedure was modified for ISIS 1 and Alouette 2. In the case of Alouette 2, the frequency markers were not superimposed on the receiver video output on most of the telemetry tapes. In the discussion below, all times refer to the time within one ionogram file based on the frame-sync time as zero.

2. ISIS-2 Frequency-Marker Identification

Frequency markers are identified in the sounder video and times (of the scan line containing the frequency marker for all or nearly all of the scan line) are assigned to them.  The order in which the frequency markers are identified determines what frequency is associated with them.  The first frequency marker identified would be associated with the 0.1 MHz frequency marker, the second with 0.25 MHz and so on (see Table 1). In efforts to determine if any of these frequency marker times are identified in error they are compared to a pre-established frequency marker table (see Table 2). This table contains a list of accurate frequency-marker times as hand scaled from a representative ionogram. During the comparison if it is determined that more than one time is associated with a frequency marker then the time closest to the time in the table is used and the rest are discarded.   These discarded times may leave other frequency markers without associated times. Times from the table are then inserted for these frequency markers. A comment appears in the ionogram header portion of the pass header file describing these insertions. This comment consists of a line of "0"s and "1"s.  If a frequency-marker time is inserted then a "1" will appear in this line.  Otherwise a "0" will appear.  The position of these "0"s and "1"s identify the associated frequency marker.  The first position is the 0.1 MHz marker; the second is the 0.25 MHz and so on (see Table 1). These identified (or inserted) frequency marker frequencies and times are used to perform interpolation to determine the frequency of each scan line based on the scan line time. See Table 3 for a description of relevant pass header comments. An inspection of the pass headers, available from http://nssdc.gsfc.nasa.gov/space/isis/isis-status.html, will determine which (if any) frequency markers had associated inserted times from Table 2. If the user is working with a binary ionogram file, the list of times for the frequency markers (contained within the file) can be compared with the times in the frequency-marker table.  If any of the times of the frequency markers within the file are exactly the same as any of the times in the table, then there is an excellent possibility that those times correspond to inserted frequency markers. These times are also available from the line-plot option under "ISIS Options" in the "interactive tracing and inversion" program available from http://nssdc.gsfc.nasa.gov/space/isis/isis-status.html.

Table 1. ISIS 1 & 2 frequency-marker position number association table

Position #        Freq. Marker (MHz)

1 0.10

2 0.25

3 0.50

4 0.75

5 1.00

6 1.25

7 1.50

8 1.75

9 2.00

10 3.00

11 4.00

12 5.00

13 6.00

14 7.00

15 8.00

16 9.00

17 10.00

18 12.00

19 14.00

20 16.00

21 18.00

22 20.00

Table 2.  ISIS-2 frequency-marker times used for comparisons with auto-detection times of frequency markers. The times are based on the hand scaling of the exact frequency-marker onset times as seen on receiver amplitude scan lines containing frequency markers on the ISIS-2 ionogram shown on the first page of http://nssdc.gsfc.nasa.gov/space/isis/isis-status.html (Ottawa station, 71315 1824:41 UT).

  Freq    Time

 (MHz)  (ms)

  0.10   3309.875

  0.25   3728.550

  0.50   4425.100

  0.75   5122.675

  1.00   5816.350

  1.25   6511.675

  1.50   7204.675

  1.75   7895.800

  2.00   8585.650

  3.00   9551.325

  4.00  10514.950

  5.00  11486.850

  6.00  12000.050

  7.00  12523.300

  8.00  13059.800

  9.00  13610.825

 10.00  14179.300

 12.00  15375.025

 14.00  16670.475

 16.00  18104.250

 18.00  19761.725

 20.00  21787.150

Table 3. Possible comments, related to frequency-marker detection and interpolation between frequency markers, in ISIS 2 pass header files pertaining to individual ionogram files.

1. "NO FREQ INTERP (FRAME SYNC NOT IDENTIFIED IN VIDEO)"

    The frame sync pulse was not identified in the video.  Therefore no frequency interpolation was performed.  All scan line frequencies were set to the default value of –1.0E+31.

2. "NO FREQ INTERP (FREQ MARKERS NOT IDENTIFIED IN VIDEO)"

    The frame was determined to be a combined fixed/swept frequency frame (as determined from the PCM) but no frequency markers were identified in video.  Therefore no frequency interpolation was performed.  All scan line frequencies were set to the default value of –1.0E+31.

3. "NO FREQ INTERP (QUESTIONABLE DETECTION OF FREQ MARKER)"

    More frequency markers were identified in the video than expected.  The number of frequency markers expected was determined from the length of the ionogram (0.1 – 10.0 or 0.1 – 20.0 MHz) as determined from the PCM.  Therefore no frequency interpolation was performed.  All scan line frequencies were set to the default value of –1.0E+31.

4. "X OVERFLOW SCAN LINES"

    Where X (integer) indicates the number of scan lines considered to be "overflow" that were included in the ionogram file.  If a pre-determined number of samples are encountered for a scan line prior to the identification of a line sync pulse that scan line is truncated and a new "overflow" one is created.  This new scan line will contain the remaining samples from the previous scan line until a line sync pulse is identified.

5. "X FREQUENCY MARKER(S) NOT IDENTIFIED IN VIDEO"

    Where X (integer) indicates the number of frequency markers that were not identified in the video.  A line of 0's and 1's will follow this comment.  For a 0.1 – 10.0 MHz ionogram there will be 17 0’s or 1’s and for a 0.1 – 20.0 MHz ionogram there will be 22.  A "0" indicates the frequency marker was identified in the video and the scan line time associated with that frequency marker is used.  A "1" indicates the frequency marker was not identified therefore the time for that frequency marker is inserted from a pre-established table.  Each of the 0’s and 1’s corresponds to a specific frequency marker based on its position in the line as indicated in Table 1.

6. "SWEPT FREQUENCY SOUNDING"

     The frame was determined (from the PCM, GMODE off) to be a combined fixed/swept frequency frame.

7. "SWEPT FREQUENCY FRAME OF ALTERNATE FIXED/SWEPT"

    GMODE is on (as determined from the PCM) and at least one frequency marker was identified in the video.  Therefore it is considered to be a combined fixed/swept frequency frame.

8. "FIXED FREQUENCY FRAME OF ALTERNATE FIXED/SWEPT"

    GMODE is on (as determined from the PCM) but no frequency markers were identified in the video.  Therefore it is considered to be a fixed frequency frame.

9. "SOUNDING STATUS (FIXED/SWEPT) NOT DETERMINED FROM PCM"

 The sounding status could not be determined from the PCM.

NOTE OF CAUTION: If any frequency marker times were inserted from Table 2, caution should be used when using these times and the corresponding interpolated frequencies. After the processing of the ISIS 2 tapes was completed it was discovered  that this comparison test, designed to remedy missing or out of place frequency markers and their times, did not work as well as intended.  As a result, incorrect frequency marker frequency labels and incorrect associated times and/or non-monotonic frequency marker times and/or non-monotonic interpolated scan line frequencies can be encountered. These problems can be corrected by creating a new ionogram file based on hand-scaled frequency-marker onset times using the line-plot option under "ISIS Options" in the "interactive tracing and inversion" program available from http://nssdc.gsfc.nasa.gov/space/isis/isis-status.html. A different frequency marker-identification routine was used for the ISIS 1 data as described in the Section 4.

3. Interpolation between ISIS-2 Frequency Markers

Frequency interpolation is not performed on ISIS-2 ionograms if:

(1) the ionogram frame sync pulse was not detected,

(2) no frequency markers were identified,

(3) the ionogram is a fixed frequency ionogram or

(4) more frequency markers were identified in the video than expected.

If interpolation is not performed then all scan line frequencies are set equal to –1.0E+31.

For a fixed/swept ionogram the scan lines in the fixed-frequency portion (scan line times up to the 0.1 MHz frequency marker) are set to the fixed frequency as determined from the PCM data. For the swept-frequency portion, scan-line frequencies are determined by interpolation/extrapolation (using a subroutine called TAB) for each scan line time using the frequency marker times within the ionogram.

There are 3 different sweep rates in an ISIS 2 ionogram (0.1-2.0, 2.0-5.0, 5.0-20.0 MHz).  Because of these different sweep rates a third degree interpolation is performed only within a single sweep rate region.  For example, if a scan line has a time that falls within the times of the 2.0 and 5.0 MHz frequency markers then only the 2.0, 3.0, 4.0 and 5.0 frequency markers would be used for the interpolation.

A third-degree interpolation is always used in each region. This interpolation can be based on a mixture of frequency-marker times identified during the A/D process and frequency-marker times inserted from Table 2 as described above.

4. ISIS-1 Frequency-Marker Identification

Frequency markers are identified in the sounder video and a time is associated with them.

After the frequency markers are identified a comparison is made between their times and times from a pre-established frequency marker table.  This table contains times based on the hand scaling of the exact frequency-marker onset times as seen on receiver amplitude scan lines containing frequency markers of a representative ISIS-1 ionogram. The comparison consists of calculating the time differences between adjacent identified frequency markers (the delta times) and comparing them to the delta times of the hand-scaled frequency markers from the table (see Table 4). This is a one-to-one comparison.  If the delta time between the 2nd and 1st identified frequency markers (0.25 and 0.1, respectively) is within plus or minus 20% of the table delta time then this identified frequency marker is considered to be "good".  This process is repeated for the rest of the delta times, 3rd and 2nd (0.5 and 0.25, respectively) etc., but using plus or minus 10% instead of 20%. If this delta time comparison fails, no further comparison is performed and the identified frequency markers prior to the failure are considered "good" and are used in interpolating for the frequency of each scan line up to the last "good" frequency marker.  If this comparison fails before reaching the third frequency marker (0.50 MHz) then no interpolation is performed.  If the 0.5 MHz marker is the last "good" frequency marker, then a second-degree interpolation is performed between the 0.1 and 0.5 MHz markers.

Table 4. ISIS-1 frequency-marker times and delta times (time differences between adjacent frequency markers) used for comparisons with auto-detection times of frequency markers. The times are based on the hand scaling of the exact frequency-marker onset times as seen on receiver amplitude scan lines containing frequency markers on a representative ISIS-1 ionogram.

   Freq    Time     Delta Time

  (MHz)   (ms)          (ms)

  0.10   3219.950  3219.950

  0.25   3690.550    470.600

  0.50   4469.700    779.150

  0.75   5250.775    781.075

  1.00   6034.625    783.850

  1.25   6820.250    785.625

  1.50   7600.750    780.500

  1.75   8387.175    786.425

  2.00   9159.925    772.750

  3.00  10528.475  1368.550

  4.00  11605.125  1076.650

  5.00  12697.850  1092.725

  6.00  13652.675   954.825

  7.00  14545.800   893.125

  8.00  15456.000   910.200

  9.00  16387.550   931.550

 10.00  17340.600   953.050

 12.00  19284.975  1944.375

 14.00  21282.600  1997.625

 16.00  23321.350  2038.750

 18.00  25373.775  2052.425

 20.00  27419.625  2045.850

5. Interpolation between ISIS-1 Frequency Markers

Frequency interpolation is not performed on ISIS-1 ionograms if:

(1) the ionogram frame sync pulse was not detected,

(2) no frequency markers were identified,

(3) the ionogram is a fixed frequency ionogram or

(4) the delta time comparison test failed prior to the 0.50 MHz frequency marker.

If interpolation is not performed, all scan line frequencies are set to a default value of –1.0E+31.

As in ISIS 2, there are 3 different sweep rates in an ISIS 1 ionogram (0.1-2.0, 2.0-5.0 and 5.0-20.0 MHz).  Again, frequency interpolation is performed only within a single sweep-rate region.  In the case of ISIS 1, however, a 3^rd degree interpolation is not always used. The frequency interpolation is performed as follows:

            Frequency range          Interpolation

                  (MHz)

                  0.1 - 2                    3rd degree

                  2 - 3                       linear

                  3 - 5                       2nd degree

                  5 - 6                       linear

                  6 - 20                     3rd degree

The linear interpolations between 2 & 3 and 5 & 6 MHz were used to comply with the John Jackson true-height analysis program where corrections are applied to the scaled frequencies in order to correct for the non-linear frequency sweep in these ranges.

The frequency in the fixed-frequency portion (prior to 0.1 MHz) of a fixed/swept ionogram is determined from the PCM data; in the swept-frequency portion it is determined by interpolation/extrapolation using the ionogram frequency markers.

Unlike ISIS 2, where frequency interpolation was performed for all scan lines or none, frequency interpolation is performed on ISIS 1 ionograms for scan lines up to the last "good" frequency marker (as described above) and the remainder of the scan lines have a default value ( -1E+31) assigned as their frequency.

6. Alouette-2 Frequency-Marker Identification

As with ISIS 1 & 2, there were 22 frequency markers on Alouette 2. They occur at different frequencies in Alouette 2, however, and only cover the range from 0.1- 13.5 MHz (see Table 5). In addition, the frequency markers were recorded differently on Alouette 2 then they were on ISIS 1 & 2. There was the capability to have the markers imbedded in the sounder video data (as on ISIS 1 & 2) or to have the frequency-marker data on a separate channel from the sounder video data. Most of the data were recorded in the latter mode. As a result, most of the Alouette-2 digital ionograms will not display frequency markers as are displayed on the ISIS-1 & ISIS-2 digital ionograms. After the frequency markers are identified during the A/D operation, a large amplitude enhancement is placed after the amplitude calibration pulse at the end of the line scan associated with the frequency marker. These amplitude enhancements produces tick marks at the bottom of the jpeg images used for quality-control diagnostics and they prominently appear in the line-scan mode of the analysis program.

After the frequency markers have been automatically identified, a comparison is made between their times and the times from a pre-established frequency marker table (see Table 5) using the approach used with ISIS 1, rather that the approach of ISIS 2, i.e., the approach based on  the time differences between adjacent identified frequency markers (the delta times) as described in Section 4 except for some differences in the allowed tolerances.  If the delta time between the 2nd and 1st identified frequency markers (0.2 and 0.1, respectively) is within plus or minus 30% of the table delta time (rather than the 20% used for ISIS 1) then this identified frequency marker is considered to be "good".  This process is repeated for the rest of the delta times, 3rd and 2nd (0.5 and 0.2, respectively) etc., but using plus or minus 15% (rather than the 10% used for ISIS 1).

Table 5. Alouette-2 frequency-marker times and delta times (time differences between adjacent frequency markers) used for comparisons with auto-detection times of frequency markers. The times are based on the hand scaling of the exact frequency-marker onset times as seen on receiver amplitude scan lines containing frequency markers on three representative Alouette-2 ionograms.

   Freq    Time     Delta Time

  (MHz)   (ms)          (ms)

0.10 3681.500 3681.500

0.20 4644.688 963.188

0.50 7823.188 3178.500

0.55 8333.126 509.938

0.90 11612.001 3278.875

1.25 14462.689 2850.688

1.50 16296.064 1833.375

1.60 16999.564 703.500

2.00 19614.877 2615.313

2.50 20151.877 537.000

3.50 21120.690 968.813

4.50 22052.815 932.125

5.50 23003.065 950.250

6.50 24013.753 1010.688

7.00 24553.003 539.250

7.50 25123.191 570.188

8.50 26368.441 1245.250

9.50 27572.879 1204.438

10.50 28188.004 615.125

11.50 28866.504 678.500

12.50 29675.879 809.375

13.50 30667.567 991.688

7. Interpolation between Alouette-2 Frequency Markers

The conditions when frequency interpolation is not performed on Alouette-2 ionograms are the same as for ISIS-1 ionograms as discussed in Section 5 above. Assuming a good ionogram with all 22 markers, the frequency interpolation on Alouette-2 ionograms is performed as follows:

For scan lines between the 0.1 and 0.2 frequency markers, a linear extrapolation is performed.

For scan lines between the 0.2 and 2.0 frequency markers, a 3rd degree interpolation is performed using only the markers in range.

For scan lines between the 2.0 and 13.5 frequency markers, a 3rd degree interpolation is performed using upper range markers

Note: 3rd degree interpolation is performed only if 4 or more frequency markers are known within the range(0.2 to 2.0 or 2.0 to 13.5). If only 3 markers are known a 2nd degree interpolation is performed. A linear interpolation is performed if only 2 markers are known.