Observations


6.1 Observation Shelters

          Dobson ozone spectrophotometers are commonly wheeled outside from storage shelters for observations. To minimize temperature changes, the instruments may be covered with thermostatically controlled electric heating blankets and quilted covers.

          A preferred procedure, which avoids instrument transport and possible consequent instrument jarring, is to house the spectrophotometers in permanent observing shelters. In the U.S.A., astronomical dome-shelters are used that are rotatable and have hatches that open and allow direct sun, moon and zenith sky observations to be made. Such domes are mounted on concrete pads, insulated, heated, air conditioned and equipped with dehumidifiers in tropical climates. For operating convenience, the spectrophotometers inside the dome shelters are mounted on rotatable pedestals.

          In polar regions, spectrophotometers are often housed in well insulated, rectangular, flat-roofed structures equipped with windows that open for sun or moon observations and a roof hatch that opens for zenith sky measurements. The need to open windows in cold climates when making direct sun or moon observations may, however, be avoided by using a periscope, devised by R. A. Olafson, to conduct light from above the roof of the spectrophotometer shelter into the instrument. Specifications for fabricating the periscope are available from the Canadian Atmospheric Environment Service, 4905 Dufferin St., Downsview, Ontario.

6.2 Observations of Total Ozone Amount

6.2.1 Observation Types

          AD-wavelength observations on direct sun (AD-DSGQP observations) are fundamental. However, a variety of other kinds of routine observations, listed in Table 2, can be made with the Dobson instrument. The kinds of observations to be made at any time depend on sky conditions, the elevation of the sun, and instrument characteristics. For example, the fundamental AD-DSGQP observations can only be made when the sun is unobscured by clouds and when it is fairly high in the sky (mu is less than 3.0 or Z is less than about 70 degrees). When the elevation of the sun is greater than 80° (Z <10° or mu < 1.015), the instrument's sun director becomes unusable and AD-DSGQP observations cannot be made.

          In general, DSGQP type observations should not be made at high mu since at low sun the brightness of the sky in the vicinity of the sun may be comparable in brightness to the sun's disc for short wavelengths. Since skylight has a different spectral composition from sunlight, spectrophotometer readings may be adversely affected. (Also, for A wavelengths, errors due to scattered light within the instrument occur at high mu.) Limiting dial readings for DSGQP type observations may be determined by taking a series of observations on the rising or setting sun, computing the ozone amounts and plotting them against N. At some value of N that should not be exceeded during routine measurements, the ozone amounts will apparently begin to decrease. Focused image (FI) type observations, on the other hand, should not be made on very low sun since errors arise due to light scattering within the instrument. At high mu the intensity of the short wavelength is extremely small compared to that of the long wavelength of each Dobson instrument wavelength pair, and may be comparable to the intensity of spurious light scattered by the instrument's optical surfaces. Finally, ZB and ZC type observations should not be made when the sun is low in the sky since deduced ozone amounts then become unreliable.

Table 2. Possible Types of Total Ozone Observations

Type of Obs.    Wavelengths    Light Source             Observing Range

AD-DSGQP A and D Direct sun, using GQP 1.15 < mu < 3.0 AD-DSGQP* A and D Direct sun, using GQP* 1.015 < mu < 1.15 CD-DSGQP C and D Direct sun, using GQP 2.4 < mu < 3.5 AD-DSFI A and D Focused image of sun 2.5 < mu < 4.0 CD-DSFI C and D Focused image of sun 2.5 < mu < 6.0 AD-ZB A and D Clear zenith 1.15 < mu < 4.0 CD-ZB C and D Clear zenith 1.8 < mu < 5.8 CC'-ZB C and C' Clear zenith 1.0 < mu < 4.4 AD-ZC A and D Cloudy zenith 1.15 < mu < 2.4 CD-ZC C and D Cloudy zenith 1.8 < mu < 5.8 CC'-ZC C and C' Cloudy zenith 1.0 < mu < 4.4 AD-RMFI A and D Focused image of moon 1.15 < mu < 3.0 CD-RMFI C and D Focused image of moon 1.15 < mu < 3.5 D-RMFI D Focused image of moon** 3.0 < mu < 5.0

*[Lens should be removed from the sun director (see Section 6.2.4.1).]
**[This observation made only in polar regions (see Section 7.1).]

6.2.2 Times of Routine Ozone Measurements

          Routine total ozone observations should be made each day near local apparent noon (L.A.N.) as well as during "observing windows" defined by the times of occurrence in the morning and afternoon of the mu ranges* specified below:

1.015 < mu < 1.2  
1.5   < mu < 2.0  
2.5   < mu < 3.0  
3.5   < mu < 5.8 .

As seasons progress from summer to winter, one or more of the observing windows approach the time of L.A.N., and should be omitted for approach times shorter than about one-half hour. As indicated in the previous section, an additional constraint on making observations is that direct sun observations are not possible for mu <1.015 since with the sun overhead, the spectrophotometer's sun director becomes unusable. Also, with high sun ozone observations made on the zenith sky become unreliable.

*[Observations during 3.5 < 5.8 need not be made if measurements were obtained during at least one of the other specified observing intervals of lower mu.]

          At stations where a regular schedule of observations must be followed throughout the year, or at very high latitude stations where mu changes slowly with time, it is less satisfactory but acceptable to make observations at L.A.N., and in the mornings and afternoons at times symmetrical about L.A.N. and spaced two to four hours apart in time. The reason that exact observing times may not be specified is that an observer may not be available to make observations at regular times because of other pressing work that he might be required to do. Furthermore, he may wish to advance or retard the time of an observation in order to make more precise measurements on, say, direct sun or clear sky, rather than on cloud.

          To ensure that observers do not waste their time in making total ozone measurements which cannot yield usable data, it is useful to provide each ozone observing station with a table entitled Times of Routine Measurements of Total Ozone Amount. This table, a sample of which is shown in Table 3, lists the approximate time intervals during daylight hours when various types of ozone observations may be made and classifies the different observations according to priority. Note that AD-DSGQP measurements are most reliable and should be made whenever possible. They are, therefore, assigned a priority 1. Other types of observations should be made in accordance with listed priorities. An observer should always consult his Times of Routine Measurements of Total Ozone Amount table before making an ozone measurement in order to determine what type of observation to make.

          Alternatively, computer produced mu tables (for every minute of each day) may be provided to field stations to aid observers in planning their observation programs. Such tables are especially useful where reduction of observational data is performed on site.

          AD-RMFI and CD-RMFI observations should be made in polar regions on the one-half full to full moon every two to three hours whenever adequate instrument sensitivity exists. In lower latitude regions, moon observations need not be made whenever reliable direct sun or zenith sky data are obtained, except when special investigations are undertaken. Moon observations on AD wavelengths are preferred. In general, it is found that CD-RMFI observations can be made on full moon with a sensitive spectrophotometer down to mu ~ 4.5 when the total amount of ozone is low. Useful moon observations in polar regions can also be made on D wavelengths alone.

Table 3


TIMES OF ROUTINE MEASUREMENTS OF TOTAL OZONE AMOUNT
(Observing Time Intervals in E.S.T.)
Station: Sterling, Virginia, USA.


Type Obs.
Priority
mu-range
Direct Sun Observations   Clear Zenith Observations Cloudy Zenith Observations
AD-DSGQP
1
(1.15 < mu < 3.00)
AD-DSGQP*
1a
(1.015 < mu < 1.15)
CD-DSGQP
2
(2.40 < mu < 3.50)
CD-DSFI
3
(3.00 < mu < 6.00)
AD-ZB
4
(1.15 < mu < 4.00)
CC'-ZB
5
(1.00 < mu < 4.40)
AD-ZC
6
(1.15 < mu < 2.40)
CC'-ZC
7
(1.00 < mu < 4.40)
Jan.  1
6
11
16
21
0945-1443
0943-1443
0940-1456
0936-1504
0931-1512
    0920-1039
0919-1035
0917-1030
0913-1023
0909-1015
1348-1507
1357-1513
1406-1519
1417-1527
1428-1534
0826-0945
0826-0943
0824-0940
0822-0936
0819-0931
1443-1601
1449-1606
1456-1612
1504-1618
1512-1621
0903-1524
0902-1530
0900-1536
0857-1542
0853-1550
  0853-1535
0852-1540
0850-1546
0848-1552
0844-1559
1039-1348
1035-1357
1030-1406
1023-1417
1015-1428
  0853-1535
0852-1540
0850-1546
0848-1552
0844-1559
26
31
Feb.  5
10
15
0925-1521
0918-1530
0910-1538
0902-1547
0854-1555
    0903-1007
0857-0958
0851-0948
0843-0939
0836-0929
1439-1452
1450-1550
1500-1558
1510-1605
1520-1613
0815-0925
0810-0918
0804-0910
0758-0902
0752-0854
1521-1631
1530-1637
1538-1644
1547-1651
1555-1657
0849-1557
0843-1604
0837-1612
0830-1619
0822-1626
  0839-1606
0834-1613
0828-1621
0821-1628
0814-1635
1007-1439
0958-1450
0948-1500
0939-1510
0929-1520
  0839-1606
0834-1613
0828-1621
0821-1628
0814-1635
20
25
Mar.  2
7
12
0847-1602
0837-1610
0828-1617
0819-1624
0810-1630
    0828-0919
0819-0909
0811-0859
0802-0849
0754-0839
1529-1620
1537-1627
1546-1634
1553-1640
1600-1646
0745-0845
0737-0837
0730-0828
0721-0819
0713-0810
1602-1703
1610-1709
1617-1715
1624-1721
1630-1726
0815-1633
0807-1640
0759-1646
0750-1652
0742-1658
  0806-1641
0759-1648
0751-1654
0742-1700
0734-1706
0919-1529
0909-1537
0859-1546
0849-1553
0839-1600
  0806-1641
0759-1648
0751-1654
0742-1700
0734-1706
17
22
27
Apr.  1
6
0810-1636
0753-1641
0744-1647
0736-1652
0728-1656
    0745-0830
0737-0821
0729-0812
0721-0804
0713-0756
1607-1652
1613-1657
1618-1702
1624-1707
1629-1712
0705-0801
0657-0753
0649-0744
0642-0736
0634-0728
1636-1732
1641-1737
1647-1742
1652-1746
1656-1751
0733-1704
0725-1709
0717-1719
0709-1719
0702-1723
  0729-1711
0718-1716
0710-1721
0702-1726
0654-1731
0830-1607
0821-1613
0812-1618
0804-1624
0756-1629
  0726-1711
0718-1716
0710-1721
0702-1726
0654-1731
11
16
21
26
May   1
0721-1211
0715-1142
0708-1112
0701-1057
0656-1046
1211-1701
1238-1705
1315-1710
1318-1714
1328-1718
1211-1211
1142-1238
1112-1315
1057-1318
1046-1328
0706-0748
0659-0741
0652-0735
0646-0729
0641-0723
1634-1716
1638-1721
1643-1725
1647-1729
1651-1734
0627-0721
0620-0714
0613-0708
0607-0701
0601-0656
1701-1756
1706-1800
1710-1805
1714-1809
1718-1813
0654-1211
0647-1135
0641-1112
0635-1057
0629-1046
1211-1728
1244-1732
1305-1737
1318-1741
1328-1745
0647-1735
0640-1740
0633-1744
0627-1748
0622-1753
0748-1211
0741-1135
0735-1112
0729-1057
0723-1046
1211-1634
1244-1638
1305-1643
1318-1647
1328-1651
0647-1735
0640-1740
0633-1744
0627-1748
0622-1753
6
11
16
21
26
0651-1037
0647-1029
0643-1024
0640-1019
0637-1016
1336-1722
1343-1726
1349-1730
1354-1733
1358-1736
1037-1336
1029-1343
1024-1349
1019-1354
1016-1358
0636-0718
0631-0714
0627-0710
0624-0707
0622-0705
1655-1738
1659-1742
1702-1745
1706-1749
1709-1752
0556-0651
0551-0647
0547-0643
0544-0640
0541-0637
1722-1818
1726-1822
1730-1826
1733-1830
1736-1833
0624-1037
0619-1029
0616-1024
0612-1019
0610-1016
1336-1749
1343-1753
1349-1757
1354-1801
1358-1804
0616-1757
0612-1801
0608-1805
0605-1808
0602-1812
0718-1037
0714-1029
0710-1024
0707-1019
0705-1016
1336-1655
1343-1659
1349-1702
1354-1706
1358-1709
0616-1757
0612-1801
0608-1805
0605-1808
0602-1812
31
June  5
10
15
20
0636-1013
0635-1012
0634-1011
0634-1011
0635-1012
1402-1739
1405-1742
1409-1746
1409-1746
1411-1748
1013-1402
1012-1405
1011-1407
1011-1409
1012-1411
0620-0703
0619-0702
0618-0702
0618-0702
0619-0703
1712-1755
1714-1758
1717-1800
1718-1802
1720-1804
0539-0636
0537-0635
0537-0634
0537-0634
0537-0634
1739-1836
1742-1839
1744-1842
1746-1844
1748-1845
0608-1013
0607-1012
0606-1011
0606-1011
0607-1012
1402-1807
1405-1405
1407-1812
1409-1814
1411-1816
0600-1815
0559-1818
0558-1820
0558-1822
0559-1824
0703-1013
0702-1012
0702-1011
0702-1011
0703-1012
1402-1712
1405-1714
1407-1717
1409-1718
1411-1720
0600-1815
0559-1818
0558-1820
0558-1822
0559-1824
25
30
July  5
0636-1013
0638-1015
0640-1017
1412-1749
1412-1749
1411-1748
1013-1412
1015-1412
1017-1411
0620-0704
0622-0706
0624-0708
1721-1805
1721-1805
1721-1804
0539-0636
0540-0638
0543-0640
1749-1846
1749-1846
1748-1846
0608-1014
0610-1015
0612-1017
1412-1817
1412-1817
1411-1816
0600-1824
0602-1825
0604-1824
0704-1013
0706-1015
0708-1017
1412-1721
1412-1721
1411-1721
0600-1824
0602-1825
0604-1824
      *Remove lens from sun director before making this observation.

6.2.3 Recording Observational Data

          The spectrophotometer dial readings, as well as other pertinent information, should be recorded on a form similar to that shown in Figure 4. Computations of mu and the total ozone amount are made directly on the form.

          Under Notes indicate the condition of the sky at the time of observation according to the following code:

(a) Cloudless sky in the vicinity of sun, moon, or zenith:

 C - Clear H - Hazy VH - Very hazy 

(b) Cloudy sky (when making zenith observations only):

Cloud Height Cloud Thickness Cloud Texture

L - Low TN - Thin U - Uniform M - Middle M - Medium V - Variable H - High TK - Thick P - Patchy


Sample Observation Form GIF ( Kb)

6.2.4 Observing Methods

          This section outlines in detail procedures to be followed in making various types of total ozone measurements. Note that AD-DSGQP observations are fundamental; other types of measurements are made similarly, with minor variations.

          Regardless of the type of observations to be made, the ozone spectrophotometer should be prepared for use by executing operations (1) to (5) given below. Failure to perform these operations in the order presented may result in damage to the microammeter or photomultiplier.

(1) Unclamp the microammeter movement and zero the microammeter, if necessary.

(2) Check to ensure that the microammeter shunt is in the position of least microammeter sensitivity.

(3) Check to ensure that the photomultiplier step switch is positioned at step 1, the position of least photomultiplier sensitivity.

(4) Turn on the spectrophotometer power supplies.

(5) Install a newly smoked plate (or waxed chart) on the spectrophotometer R-dial and wind the recording mechanism clock.

6.2.4.1 AD-DSGQP Observations

(1) Perform operations (1) to (5) given above.

(2) Uncover the spectrophotometer inlet window and place the ground quartz plate over it.

(3) Place the sun director over the inlet window. The lens* within the sun director must be in its lowest position. IMPORTANT: When using the sun director, make certain that a tube within the unit that prevents sky light from entering the sun director through slots in its sides is in position; otherwise an appreciable amount of sky light may enter the instrument causing erroneous readings, especially when direct sun observations are made on low sun. For this reason, also, the ground quartz plate viewinq window built into the sun director must always be closed when observations are being made.

*[The purpose of the lens is to increase available light intensity and, hence, instrument sensitivity at low sun (mu ~ 3), and to permit focused image observations to be made. For Dobson instruments equipped with sensitive photomultipliers and electronics, and where, also, focused image observations are not made, use of the lens within the sun director may be omitted. Since, however, differential absorption of light at the A, B, C, and D wavelengths can occur within the lens depending on the optical quality of the quartz from which it is made, care must be taken to determine any differences in N values that may arise when observations are made with and without the lens. Appropriate corrections must then be applied when processing observational data.]

(4) Insert the S4 shutter rod all the way into the spectrophotometer. Set the wavelength selector rod to the SHORT position.

(5) Read the temperature of the instrument to the nearest 0.5 degree Celsius and set the Q1 lever stops for A and D wavelengths according to values given in the Table of Settings of Q. The Q2 lever stops are always set to the values given in the table corresponding to the temperature of 15° C.

(6) Orientate the instrument so that its long axis points toward the sun, the sun being on the observer's right hand. (It is important to align the spectrophotometer carefully with respect to the sun; otherwise, errors may result in the measurement. If several observations are made in succession, the orientation of the instrument should be periodically checked and corrected.)

(7) Switch on the motor which drives the sector wheel.

(8) Adjust the prism of the sun director so that the patch of sunlight falls centrally on the ground quartz plate.

(9) Set Q1 and Q2 levers for A wavelengths and increase microammeter sensitivity by rotating the shunt potentiometer fully clockwise. There should now be a small deflection of the microammeter. Turn the R-dial to bring the microammeter reading to zero and gradually increase the photomultiplier voltage with the step switch, if necessary, until sufficient sensitivity is obtained as indicated by a slight instability of the microammeter needle. Make a mental note of the position of the step switch.

(10) Rotate the step switch to step 1 to reduce the voltage on the photomultiplier and set Q1 and Q2 levers for D wavelengths. As in step (9), increase the photomultiplier voltage while zeroing the microammeter until suitable sensitivity is obtained. Again, note the setting of the step switch controlling sensitivity and the approximate R-dial reading.

          The spectrophotometer has now been readied for use in making an AD-DSGQP observation. Take actual observations as follows:

(11) Set Q1 and Q2 levers for A wavelengths and rotate the step switch to give suitable sensitivity as previously determined.

(12) Since it is always good practice to commence an observation at the beginning of a minute, glance at your chronometer (or other timepiece giving accurate time) and either make a mental note of the time you will begin the observation or record that time on a scratch pad. (To preserve accuracy in computed data, observations must be timed to within ±10 seconds of correct time when the sun is rising or setting fairly rapidly.)

(13) Several seconds before starting time, set the patch of sunlight centrally on the ground quartz plate. The patch of sunlight will shift during the course of the observation, and will require readjustment (see step (19)).

(14) At starting time, lower the recorder stylus on to the smoked plate.

(15) Carefully and slowly oscillate the R-dial so that the microammeter needle deflects just to one side of zero, then just to the other side, etc. Continue the manipulation of the R-dial for 20 seconds.

(16) Reduce sensitivity to that suitable for D wavelengths. Raise the stylus off the smoked plate.

(17) Set Q1 and Q2 levers for D wavelengths while at the same time adjusting the R-dial to the reading expected (see operation (10)).

(18) Lower the stylus again and record the R-dial reading for 20 seconds as in (15).

(19) Raise the stylus off the smoked plate. Set Q1 and Q2 levers for A wavelengths while rotating the R-dial to its correct position. Adjust the step switch for suitable sensitivity. Lower the stylus and again and make a 20 second recording of the R-dial reading. Adjust the sun director, if necessary, so that the patch of light falls centrally on the GQP.

(20) Repeat operations (16) to (19). You will now have completed an ADADA observation. Glance at the chronometer and record on a scratch pad the time, in hours, minutes, and seconds, when the observation was completed.

(21) To shut down the spectrophotometer, first set the sensitivity step switch to step 1. Then, decrease the microammeter sensitivity to zero by turning the shunt potentiometer. Finally, switch off the sector wheel motor and power supplies, remove the sun director, and replace the cover over the inlet window.

(22) Carefully bisect the traces made on the smoked plate with the stylus and read mean values of the traces to the nearest 0.1 degree. Record the dial readings, mean observing times, and other pertinent information.

6.2.4.2 AD-DSGQP* Observations

          Observations of this type are made only when the sun is very high in the sky, i.e., 1.015 < 1.15. The observing procedure is the same as that for the regular AD-DSGQP measurement except that the lens is removed from the sun director. Removal of the lens allows a larger patch of sunlight to illuminate the ground quartz plate. (See also the footnote on page 33, Section 6.4.2.1).

6.2.4.3 CD-DSGQP Observations

          These observations are made in the same way that AD-DSGQP measurements are made, except that the Q1 and Q2 lever stops are set for C and D wavelengths.

6.2.4.4 AD-ZB Observations

          AD-ZB measurements are made on a 7-degree cone of light emanating from the zenith sky. The observing procedure is the same as for AD-DSGQP measurements except that the sun director and ground quartz plate are not used. (DO NOT USE THE GROUND QUARTZ PLATE). Care must be taken during the measurement to ensure that the cone of light entering the spectrophotometer is unobstructed by centering the instrument accurately beneath the observatory hatch if observations are made from within a building.

6.2.4.5 AD-ZC Observations

          AD-ZC and AD-ZB observations are made in exactly the same way. For accuracy, it is necessary to make the A and D measurements on a very similar cloud; hence, a close watch must be maintained on the zenith sky during observations. If measurements must be made on variable or patchy cloud, two or three observations of the type ADADA should be made in succession, to constitute a single ozone determination.

          AD-ZC measurements can also be made when light rain or snow is falling if the spectrophotometer is housed in an observing shelter. A fan mounted beside the observatory hatch can be used to blow raindrops or snowflakes away from the spectrophotometer inlet window during observations.

6.2.4.6 CD-ZB and CD-ZC Observations

          These observations are made in the same way that AD-ZB and AD-ZC measurements are made. except that the Q1 and Q2 stops are set for C and D wavelengths. The observation are made in the mu-range 1.8 to 5.8, and are useful primarily at high latitude stations at times of the year when the sun is low in the sky.

6.2.4.7 CC'-ZB and CC'-ZC Observations

          These observations are made in the same way that AD-ZB and AD-ZC measurements are made, except that the Q levers are set for C wavelengths and maintained in that position. The sequence of observations is CC'CC'C. Observations on C wavelengths are made with the wavelength selector rod in the SHORT position whereas observations on C' wavelengths are made with this rod in the LONG position. The S4 shutter rod should be set all the way into the spectrophotometer and maintained in that position throughout the observations. As in the case of AD-ZC measurements, care should be taken to see that the C and C' readings are made on closely similar cloud. If measurements must be made on variable or patchy cloud, two or three observations should be made in succession to constitute a single ozone observation.

6.2.4.8 AD-DSFI and CD-DSFI Observations

          AD-DSFI and CD-DSFI measurements are similar to AD-DSGQP and CD-DSGQP measurements, respectively, except that the ground quartz plate is not employed and a focused image of the sun is cast on slit S1 by the sun director. A mark is made on the side of the sun director showing the correct position of the lens at which a sharp image of the sun is thrown on S1 at wavelengths of about 3200 A.U. If such a mark has not been made, it can be made as follows. Using a green filter, find the lens position on the sun-director when a sharp image of the sun is formed on S1. The position for a sharp focus at 3200 A.U. will be 15 mm below that for green light.

          Since it is not possible to see slit S1 easily through the viewing hole at the bottom of the sun director, a small glass viewing prism is fitted under the main quartz prism. Also, small marks are inscribed on the jaws of the slit which permit the positioning of the sun's image approximately centrally along the length of the slit. It will be found that if the sun's image is moved along the length of the slit, the dial readings will change, but it will generally be found that there is some region near the center when a small movement of the image along the slit makes little difference in dial readings. Tests should be made to find this position, and the sun's image should always be set at this point during observations.

          It is necessary to set the sun's image very carefully on the slit and allow for the movement of the sun so that the slit will bisect the sun's image halfway through the observation. A preliminary test must be made to see how much the sun's image moves in, say, 20 seconds. The observation should not be continued for more than 30 seconds without resetting the sun's image on the slit, unless a second observer is present to keep the sun's image continuously in the center of the slit.

          There is a great difference in the intensity of light at the C and D wavelengths when the sun is low. After the observation on C wavelengths has been made, the instrument sensitivity must be greatly reduced before changing to D wavelengths. Often the sensitivity at the D wavelengths is too great even with the lowest voltage on the photomultiplier, and the voltage output of the photomultiplier power supply must then be reduced.

          At least three independent ADADA or CDCDC observations should be made in succession.

          WARNING! In order not to damage the photomultiplier, DSFI observations should not be made when the sun is high in the sky (mu <2.5).

6.2.4.9 AD-RMFI and CD-RMFI Observations

          These observations are made in a manner similar to AD-DSFI and CD-DSFI measurements except that the one-half full to full moon is used as a light source.

          At low latitude stations when the moon is nearly overhead, a satisfactory image of the moon will not be projected onto slit S1 by the sun-director (a skewed image will result because of the oblique angle at which the rays of moonlight strike the face of the sun-director prism). To view a proper image, a moon-director prism unit is provided with each spectrophotometer for attachment to the sun-director whenever observations on high moon are made.

6.3 Umkehr Observations

          Umkehr observations yield highly useful information on the vertical distribution of ozone in the atmosphere. Two types of Umkehr observations are possible. The first involves a standard Umkehr observing procedure (Mateer and Dütsch, 1964) that has been in use since the late 1950's, while the second is a newly developed short method (Mateer et al., 1979) which has a time saving advantage that allows for an increased frequency of such observations to be made during conditions of clear sky and relatively constant total ozone amount. Every effort should be made by observers to make short Umkehr measurements whenever sky conditions permit. Additionally, coincident standard Umkehr observations should be made whenever possible in order that a large amount of data become available for comparing ozone vertical distributions derived from the two observing methods.

          A standard Umkehr observation consists of a series of C-wavelength measurements made on the clear zenith sky during morning or afternoon. The measurements are commenced a few minutes before sunrise and continued until the sun is at an elevation of not less than about 20 degrees, or commenced in the afternoon when the sun is at an elevation of not less than about 20 degrees and continued until shortly after sunset. A short Umkehr observation, on-the-other-hand, is comprised of clear zenith sky observations on A, C, and D wavelengths during morning or afternoon when the sun is at an elevation of from 0 to 10 degrees. The zenith sky must be free from clouds for a period of from one-half to one hour near sunrise or sunset to make Umkehr observations. This is especially true at low latitude stations where the sun rises or sets rapidly. At other times, it is desirable that the zenith sky be cloudless but permissible that clouds cross it periodically when measurements are not made. Umkehr observations cannot be made at a polar station, or at high latitude stations during summertime when the sun does not sink below the horizon.

          To be able to compute the vertical distribution of ozone, it is necessary to know the total amount of ozone present at the time of observations. Several total ozone measurements must, therefore, be made during the morning or afternoon, particularly if the ozone amount is changing fairly rapidly. The measurements should include an ozone observation at mu ~ 3.0.

6.3.1 Observing Times

          Standard Umkehr measurements involve the determination of values of NC when the sun is at discreet positions in the sky, namely, when Z is 60, 65, 70, 74, 75, 77, 80, 83, 84, 85, 86.5, 88, 89, and 90 degrees. Short Umkehr observations involve determinations of values of NA, NC, and ND when Z is 80, 83, 85, 86.5, 88, and 89 degrees. It is convenient to prepare tables (see sample Table 4) showing the approximate times during the year at a particular station when specified solar zenith angles occur. Note that the times listed in Table 4 are only approximate. Their purpose is to aid the observer in effectively planning his schedule of observations. The approximate times are also useful when hand computations are performed (see Section 4, Appendix F) to obtain the exact times during any one-half day when the sun is at 60, 65, 70 degrees, etc.

          It is generally not sufficient to make only one observation at each specified Umkehr zenith angle. Rather, several measurements should be made in order to achieve high measurement precision.

Table 4

MORNING UMKEHR OBSERVATIONS
(Approximate Times of Observation, E.S.T.)



                                                  Station:  Sterling, Virginia

Date Z 90° 89° 88° 86.5° 85° 83° 80° 77° 74° 70° 65° 60° LAN


Jan. 1 0734  0740  0746  0755  0804  0817  0837  0858  0921  0954  1050        1214
     6 0734  0740  0746  0755  0804  0817  0837  0857  0920  0953  1045        1216
    11 0733  0739  0745  0754  0803  0816  0835  0856  0917  0949  1039        1218
    16 0732  0738  0743  0752  0801  0814  0833  0853  0914  0945  1032  1202  1220
    21 0729  0735  0741  0750  0759  0811  0829  0849  0909  0939  1023  1131  1221
    26 0726  0732  0737  0746  0755  0807  0825  0844  0904  0933  1014  1111  1223
    31 0722  0728  0733  0742  0750  0802  0820  0838  0858  0925  1005  1055  1223
Feb. 5 0717  0723  0728  0737  0745  0757  0814  0832  0851  0918  0955  1040  1224
    10 0712  0717  0723  0731  0739  0751  0808  0825  0844  0909  0945  1026  1224
    15 0706  0711  0717  0725  0733  0744  0801  0818  0836  0901  0934  1013  1224
    20 0700  0705  0710  0718  0726  0737  0754  0811  0828  0852  0924  1000  1224
    25 0653  0658  0703  0711  0719  0730  0746  0803  0820  0843  0914  0948  1223
Mar. 2 0646  0651  0656  0704  0712  0722  0738  0755  0811  0834  0904  0936  1222
     7 0638  0643  0648  0656  0704  0714  0730  0746  0802  0825  0854  0925  1221
    12 0630  0635  0640  0648  0656  0706  0722  0738  0754  0816  0844  0914  1220
    17 0622  0627  0633  0640  0648  0658  0714  0730  0746  0807  0835  0903  1218
    22 0614  0620  0625  0632  0640  0650  0706  0721  0737  0758  0826  0854  1217
    27 0607  0612  0617  0624  0632  0642  0658  0713  0729  0750  0317  0844  1215
Apr. 1 0559  0604  0609  0617  0624  0635  0650  0706  0721  0742  0808  0835  1214
     6 0551  0556  0601  0609  0617  0627  0643  0658  0713  0734  0800  0827  1212
    11 0543  0549  0554  0602  0609  0620  0635  0651  0706  0727  0753  0819  1211
    16 0536  0541  0547  0554  0602  0613  0628  0644  0659  0720  0746  0812  1210
    21 0529  0534  0540  0548  0556  0606  0622  0637  0653  0713  0739  0805  1208
    26 0523  0528  0533  0541  0549  0600  0615  0631  0647  0707  0733  0759  1208
May  1 0516  0522  0527  0535  0543  0554  0610  0625  0641  0702  0727  0753  1207
     6 0511  0516  0522  0530  0538  0549  0605  0620  0636  0657  0722  0748  1206
    11 0505  0511  0517  0525  0533  0544  0600  0616  0632  0652  0718  0744  1206
    16 0501  0507  0512  0520  0529  0540  0556  0612  0628  0649  0715  0740  1206
    21 0457  0503  0508  0517  0525  0536  0553  0609  0625  0646  0712  0737  1206
    26 0454  0500  0505  0514  0522  0533  0550  0606  0622  0643  0709  0735  1207
    31 0451  0457  0503  0512  0520  0531  0548  0604  0620  0642  0708  0734  1207
June 5 0450  0456  0501  0510  0519  0530  0547  0603  0619  0640  0707  0733  1208
    10 0449  0455  0501  0509  0518  0529  0546  0602  0619  0640  0706  0732  1209
    15 0449  0455  0500  0509  0518  0529  0546  0602  0619  0640  0707  0733  1210
    20 0449  0455  0501  0510  0518  0530  0547  0603  0619  0641  0707  0733  1211
    25 0450  0456  0502  0511  0520  0531  0548  0604  0621  0642  0709  0735  1212
    30 0452  0458  0504  0513  0522  0533  0550  0606  0622  0544  0710  0736  1213
July 5 0455  0501  0507  0515  0524  0535  0552  0608  0625  0646  0712  0738  1214
    10 0458  0504  0510  0518  0527  0538  0555  0611  0627  0548  0715  0741  1215
    15 0501  0507  0513  0522  0530  0541  0558  0614  0630  0651  0717  0743  1216
    20 0505  0511  0517  0525  0534  0545  0601  0617  0633  0654  0720  0746  1216
    25 0509  0515  0521  0529  0537  0548  0605  0621  0637  0658  0723  0749  1216
    30 0514  0519  0525  0533  0541  0552  0608  0624  0640  0701  0727  0753  1216
Aug. 4 0518  0523  0529  0537  0545  0556  0612  0628  0644  0704  0730  0756  1216

6.3.2 Recording Observational Data

          The dial readings and times at which the various measurements are made should be entered on a form similar to that shown in Figure 5. The times entered should be correct to within ±5 seconds. Under Notes indicate the condition of the zenith sky at the time of each observation.

 RECORD OF UMKEHR OBSERVATIONS
 STN.  INST. NO.  DATE
GMT RA Notes GMT RC Notes GMT RD Notes
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 
                 

Figure 5. Sample form for recording observational Umkehr data

6.3.3 Observing Procedures

          A short Umkehr observation is made on A, C, and D wavelengths. Observations should be made continuously when the sun is rising or setting rapidly near sunrise or sunset (i.e., when Z is greater than about 78 degrees). One procedure is to obtain a 20 second record of an observation on A wavelengths, beginning the recording 10 second before the minute. Quickly determine the RA value and record it together with the time of observation, adjust instrument controls for C wavelength observations, and commence an observation on C wavelengths beginning 10 s before the next minute. Proceed similarly to obtain a D wavelength observation beginning 10 s before the third minute. Repeat the sequence, beginning 10 s before the fourth minute, etc.

          For solar zenith angles less than about 78 degrees, observations on A and D wavelengths should be discontinued, but continued using C wavelengths in order to obtain standard Umkehr data. As soon as the sun's rate of ascent or descent has diminished, the frequency of C-wavelength observations may be reduced to once per 2 minutes and later to once per 5 to 10 minutes. (Alternatively, measurements may then be made at one minute intervals about the Z times of interest only, starting 2 or 3 minutes before a particular occurs and ending 5 or 6 minutes later).

6.4 Special Observations

6.4.1 Observations Made to Check the Spectrophotometer Calibration at A, C, and D Wavelengths

          The measurements consist of the usual AD-DSGQP and CD-DSGQP observations made during a one-half day when meteorological conditions are stable (i.e., when the total ozone amount is not changing) and the sky is relatively free from dust, smoke, and haze particles. Since a clear sky and stable meteorological conditions may occur relatively rarely, it is recommended that the special observations be made as often as possible.

          Observations must be made twice per one-half day, at a time when the sun is fairly low in the sky (2.5 < 3.2) and also when the sun's elevation is high (1.15 < mu < 1.5). Each set of observations should consist of three CDCDC-DSGQP measurements followed immediately by three ADADA-DSGQP measurements, if made before noon. In the afternoon observations on AD wavelengths should be made first, followed by observations on CD wavelengths. Particular care must be taken to ensure that measurements are made as accurately as possible.

          Record the data in the same manner that routine total ozone observational data are recorded. The method of analyzing the data to check on the calibration state of the spectrophotometer is described in Appendix D.

6.4.2 Observations Made to Correct Empirical Charts

          Each ozone observatory should be provided with zenith sky Charts AD, CD, and C; a Chart of Cloud Corrections for Chart C; and a Table of Cloud Corrections for Chart AD. The charts and the table are used in deducing total ozone amounts from measurements made on the clear and cloudy zenith sky.

          The charts and table supplied to a newly set up station will require some modification to suit the locality where the observations are made, since the vertical distribution of ozone as well as the ground albedo may be different at the new station from the ozone distribution and albedo at the station where the charts and table were drawn up. Furthermore, it may be necessary to affix ordinate or abscissa values to the charts. At an already established station, improved corrections to the charts may be required. All original computations of ozone determined from zenith sky measurements should therefore be regarded as provisional. The observer should make every effort to accumulate a sufficiently large body of direct sun and zenith sky comparison data which will enable him to modify his charts and table appropriately.

          Since AD-DSGQP observations are fundamental, the types of quasi-simultaneous comparisonmeasurements to be made are the following:

AD-DSGQP vs. AD-ZB
AD-DSGQP vs. CD-ZB
AD-DSGQP vs. CC'-ZB
AD-DSGQP vs. AD-ZC
AD-DSGQP vs. CD-ZC
   AD-DSGQP vs. CC'-ZC .

Note that the first three sets of the above observations can usually be made within a few minutes of each other. Direct sun and zenith cloud measurements, on the other hand, may often have to be 1 or 2 hours apart in time. Each comparison should consist of at least two independent observations made on direct sun followed by at least two independent measurements made on the clear or cloudy sky. Observations should be made over as wide an X and mu range as possible so that sufficient data become available for correction of the charts and table.

6.4.3 Determination of Focused Image Corrections

          It is generally found that ozone values deduced from measurements made on direct sun using a ground quartz plate differ from quasi-simultaneous values determined from focused image observations. In the case of single pair wavelengths, the differences may be very large and allowances must be made for them. The differences for double pair wavelengths, e.g., the CD wavelengths, are usually small and may even be negligible.

6.4.4 Multiplying Factors Used in Reducing CD-DSGQP Values to the AD-DSGQP Level

          The ozone absorption coefficients incorporated into equations (3) to (7) given in Section 7.1 are the 1968 IAMAP (International Association of Meteorology and Atmospheric Physics) coefficients adopted for use with Dobson spectrophotometers at the recommendation of the International Ozone Commission (IOC). When quasi-simultaneous ozone measurements are made on direct sun or moon using AD and CD wavelengths, it may be found that each type of measurement gives a slightly different result. Such differences are due to uncertainties in the ozone absorption coefficients employed in reducing the data, as well as instrumental factors and observing conditions (e.g., atmospheric aerosol loading). For- all ozone values to be comparable, it is necessary to determine multiplying factors to ozone values deduced from measurements on CD wavelengths by which results may be reduced to the AD-DSGQP level. The multiplying factors XAD/XCD are derived from a large number of quasi-simultaneous observations covering a broad range of mu. Generally, it is found that CD-DSGQP values do not differ from AD-DSGQP values by more than 2%.


Return to Table of Contents

Forward to Section 7. Reduction of Ozone Data