4.1. CONTINUING PROGRAMS
4.1.1. TOTAL OZONE OBSERVATIONS
Total ozone observations continued throughout 1994 and 1995 at 15 of the 16
stations that comprise the U.S. Dobson spectrophotometer network (Table 4.1).
Of the 16 stations, 5 were operated by CMDL personnel, 5 by NWS, 2 are domestic
cooperative stations, and 4 are foreign cooperative stations. All stations are
either fully or semiautomated. In addition, a Brewer spectrophotometer was operated
on a nearly continuous basis at Boulder.
TABLE 4.1. U.S. Dobson Ozone Spectrophotometer Station Network for 1994-1995
|Station||Period of Record||Instrument No.||Agency|
|Bismarck, North Dakota||Jan. 1, 1963-present||33||NOAA|
|Caribou, Maine||Jan. 1, 1963-present||34||NOAA|
|Wallops Is., Virginia||July 1, 1967-present||38||NOAA; NASA|
|SMO||Dec. 19, 1975-present||42||NOAA|
|Tallahassee, Florida||May 2, 1964-Nov. 30, 1989;
Nov. 1, 1992-present
|58||NOAA; Florida State University|
|Boulder, Colorado||Sept. 1, 1966-present||61||NOAA|
|Fairbanks, Alaska||March 6, 1984-present||63||NOAA; University of Alaska|
|Lauder, New Zealand||Jan. 29, 1987-present||72||NOAA; DSIR|
|MLO||Jan. 2, 1964-present||76||NOAA|
|Nashville, Tennessee||Jan. 2, 1963-present||79||NOAA|
|Perth, Australia||July 30, 1984-present||81||NOAA; Australian Bureau Meteorology|
|SPO||Nov. 17, 1961-present||82||NOAA|
|Haute Provence, France||Sept. 2, 1983-present||85||NOAA; CNRS|
|Huancayo, Peru||Feb. 14, 1964-Dec. 31, 1992||87||NOAA; IGP|
|BRW||June 6, 1986-present||91||NOAA|
|Fresno, California||June 22, 1983-March 13, 1995||94||NOAA|
|Hanford, California||March 15, 1995-present||94||NOAA|
The Peruvian station was still out of operation at the end of 1995, although a new baseline monitoring station is under construction. Operations will likely start in late 1996. In May 1995 the Fresno instrument and shelter were moved 30 miles southwest of Hanford, California. The Bismarck instrument and shelter were moved about 150 m in August 1994.
Provisional daily total ozone amounts applicable to local apparent noon for
the stations listed in Table 4.1 were archived at the World Ozone Data Center
(WODC), 4905 Dufferin Street, Ontario M3H 5T4, Canada, in Ozone Data
for the World. Table 4.2 lists the monthly mean total ozone amounts measured
at the various stations for 1994 and 1995. (Monthly means are computed for stations
where observations were made on at least 10 days each month.).
|TABLE 4.2. Provisional 1994 Monthly Mean Total Ozone Amounts (M-Atm-CM)|
|Bismarck, North Dakota||361||367||372||351||327||322||319||302||286||294||286||300|
|Wallops Is., Virginia||320||343||347||325||355||327||309||302||292||281||255||270|
|Lauder, New Zealand||277||267||262||265||289||315||337||361||370||360||341||302|
|Haute Provence, France||326||373||314||382||352||332||320||305||311||295||273||293|
|Huancayo, Peru||Station closed|
|Bismarck, North Dakota||334||358||341||347||347||321||311||282||286||290||292||297|
|Wallops Is., Virginia||307||344||316||325||316||323||312||297||288||267||299||284|
|Lauder, New Zealand||277||267||282||273||280||302||335||336||361||348||328||287|
|Haute Provence, France||293||319||333||340||340||331||322||329||304||277||283||303|
|Huancayo, Peru||Station closed|
Monthly mean ozone values in square brackets are derived from
observations made on fewer than 10 days per month.
Ten Dobson ozone spectrophotometers in the CMDL network as well as 29 others
were calibrated during 1994 and 1995. Table 4.3 lists all the instruments calibrated
and the resulting calibration difference expressed as a percent ozone difference.
This percent difference is between ozone calculated from the test and the standard
instrument measurements with the ADDSGQP observation type at a value of 2, and
a total ozone value of 300 Dobson Units (DU), before any repair or calibration
adjustment is made. The table also lists the place of the calibration and the
standard instrument used.
TABLE 4.3. Dobson Ozone Spectrophotometers Calibrated in 1994-1995
|Lisbon, Spain||D013||Aug. 2, 1990||+0.7%||65||Izaña Observatory|
|Oslo, Norway||D056||Aug. 21, 1986||+0.5%||65||Izaña Observatory|
|Potsdam, Germany||D064||Aug. 2, 1990||+0.6%||65||Izaña Observatory|
|Huancayo, Peru||D087||May 15, 1985||+0.9%||65||Izaña Observatory|
|Natal, Brazil||D093||May 20, 1986||+2.9%||65||Izaña Observatory|
|Buenos Aires, Argentina||D097||July 15, 1992||+0.1%||65||Izaña Observatory|
|El Arenosillo, Spain||D120||Aug. 9, 1990||+1.3%||65||Izaña Observatory|
|Ushuaia, Argentina||D131||None||N/A||65||Izaña Observatory|
|Tallahassee, Florida||D058||Sept. 9, 1991||-.3||83||Boulder|
|Boulder, Colorado||D061||Aug. 27, 1992||0.0%||65||Boulder|
|MLO||D076||June 13, 1993||N/A||83||Boulder|
|SPO||D080||May 26, 1988||0.5%||83||Boulder|
|BRW||D091||May 26, 1989||0.0%||83||Boulder|
|Fresno, California||D094||June 26, 1989||1.0%||83||Boulder|
|Mexico D.F Mexico||D098||August 1978||-0.3%||83||Boulder|
|RA VI Spare||15||None||N/A||65||LKO Arosa|
|Vindeln, Sweden||30||May 10, 1990||+0.7%||65||LKO Arosa|
|United Kingdom||32||May 1995||N/A||65||LKO Arosa|
|Uccle, Belgium||40||Aug. 1, 1990||+2.0%||65||LKO Arosa|
|United Kingdom, Standard||41||Aug. 2, 1990||+1.2%||65||LKO Arosa|
|Sestola, Italy||48||Nov. 12, 1980||+2.4%||65||LKO Arosa|
|Bordeaux, France||49||July 10, 1990||+0.3%||65||LKO Arosa|
|Reykjavik, Iceland||50||Aug. 2, 1990||+1.1%||65||LKO Arosa|
|Arosa, Switzerland||62||Aug. 7, 1992||+1.7%||65||LKO Arosa|
|Belsk, Poland||84||Aug. 2, 1990||+0.2%||65||LKO Arosa|
|l'Obs. Haute Provence, France||85*||July 10, 1990||+0.7%||65||LKO Arosa|
|Denmark||92*||Aug. 2, 1990||+1.0%||65||LKO Arosa|
|Arosa, Switzerland||101||Aug. 2, 1990||+2.1%||65||LKO Arosa|
|Hohenpeissenberg, Germany||104||Aug. 2, 1990||+1.6%||65||LKO Arosa|
|Moscow, Russia||107||Aug. 5, 1990||+1.4%||65||LKO Arosa|
|Budapest, Hungary||110||Aug. 2, 1990||+0.5%||65||LKO Arosa|
|Tsukuba, Japan, Standard||116*||June 29, 1992||+0.6%||65||LKO Arosa|
|Bucharest, Romania||121||Aug. 5, 1990||Not consistent||65||LKO Arosa|
|Bismarck, North Dakota||D033||Oct. 1, 1993||+1.5||83||Boulder|
|Caribou, Maine||D034||Sept., 9, 1991||+0.3||83||Boulder|
|Wallops Island, Virginia||D038||Sept. 16, 1991||+1.0||83||Boulde|
|Nashville, Tennessee||D079||Aug. 14, 1991||+0.6||83||Boulder|
|Comodoro Rivadavia, Argentina||D133||New Dobson||N/A||83||Boulder|
|Montevideo, Uruguay||D134||New Dobson||N/A||83||Boulder|
The Mauna Loa Observatory, Hawaii (MLO) instrument D076 failed mechanically in February 1994. This automated instrument was repaired and back in service in May 1994. During the repair the left side mirror was replaced due to a deteriorating surface.
CMDL participated in an international Dobson spectrophotometer calibration at Izaña Observatory (Tenerife, Spain) in June 1994 and at Arosa, Switzerland (LKO) in July 1995 as part of its role as the World Center for Dobson Calibrations. Instruments for Mexico City, Mexico; Comodoro Rivadavia, Argentina; and Montevideo, Uruguay, were calibrated in Boulder during this period. A representative from the Czech Hydrological and Meteorological Institute assisted with the latter two instruments. These were new instruments from Ealing Opto-Electronics.
Reevaluation of more than 400 station-years of total ozone data from 25 CMDL (and its predecessor organizations) Dobson spectrophotometer stations was completed in 1995. Corrections were based on field instrument calibrations made throughout the years with World Primary Standard Dobson Instrument no. 83 [Komhyr et al., 1989]. This instrument's long-term ozone measurement precision has been maintained since the early 1960s at 1% by means of Langley calibration observations conducted periodically at MLO and with standard lamps. Procedures used in reevaluating the data are described in detail in Komhyr .
Seasonal and annual downward trends in ozone during 1979-1995, determined
from the reevaluated data for five U.S. mainland stations (Caribou, Maine; Bismarck,
North Dakota; Boulder, Colorado; Wallops Island, Virginia; and Nashville, Tennessee)
and for MLO and Samoa Observatory, American Samoa (SMO), are shown in Table
4.4. Also included in the table are the ozone trends measured at Fresno, California,
during 1985-1995. The statistical method used in determining the trends (G.
C. Reinsel, University of Wisconsin-Madison) was similar to that employed by
the 1988 Ozone Trends Panel [WMO, 1988] whereby solar cycle and ozone
QBO effects are removed from the data. Note that the downward trend in ozone,
averaged over the five contiguous U.S. stations with the longest records, is
largest at -5.45% per decade for spring (March-May) months and smallest at -1.6%
per decade for autumn (September to November) months. On an annual basis, the
downward ozone trend at these sites averages -3.58% per decade. These trends
exhibit a slight recovery from values determined from 1979-1993 data that encompassed
record low ozone values over the U.S. during the winter of 1992-1993 [Komhyr
et al., 1994]. For the earlier time period, the average five-station downward
ozone trend for spring months (not shown) was -5.79% per decade and -3.8% per
decade on an annual basis. Downward trends in ozone measured nearer the equator
at MLO and SMO are significantly smaller.
TABLE 4.4. Annual and Seasonal Trends January 1979-December 1993
|Station||Latitude||Trend||Std. Error||Trend||Std. Error||Trend||Std. Error||Trend||Std. Error||Trend||Std. Error|
|Bismarck, North Dakota||46.8N||-3.30||0.63||-3.25||1.43||-5.69||1.01||-2.00||0.93||-1.62||0.95|
|Wallops Is., Virginia||37.9N||-3.67||0.68||-4.79||1.33||-4.77||1.28||-2.66||0.78||-2.10||1.18|
|Average over first five stations||-3.58||-4.19||-5.45||-2.54||-1.60|
Umkehr observations made with the Automated Dobson Network instruments continued
in 1994 and 1995 at Boulder; Haute Provence, France; Lauder, New Zealand; MLO;
Perth, Western Australia; and at the University of Alaska's Geophysical Institute.
Umkehr processing is set to resume early in 1996. Processing will begin with
MLO, followed by Lauder and the other stations in a collaborative effort with
the University of Alabama, Huntsville. Since the reprocessing of the total ozone
from these stations will have been completed, the updated calibration tables
and ozone values will be incorporated into the Umkehr processing. Under conditions
of high stratospheric aerosol loading, which was the case following the eruption
of Mt. Pinatubo, reliable ozone profiles can be obtained from the Umkehr technique
only by correcting for aerosols. Such conditions prevailed through 1992 at most
of the sites. The effort with University of Alabama, Huntsville, will include
the application of proper aerosol corrections to the profile data.
4.1.3. SURFACE AND TROPOSPHERIC OZONE
At least 20-year records of observation are now available for each of the four CMDL baseline sites. Records at Bermuda, Barbados, and Niwot Ridge are at least 5 years in length. At Westman Islands, Iceland, observations began in 1992. For several years, data were being obtained from Mace Head, Ireland, in a cooperative program as part of the Atmosphere/Ocean Chemistry Experiment (AEROCE). Data continues to be received from Mace Head but CMDL is no longer actively involved in that measurement program. The aging complement of surface ozone monitors, some of which are 20 years old, has experienced a number of breakdowns. Significant blocks of data were lost at Barbados and SMO during 1994 and 1995.
The extent to which tropospheric ozone may have changed since preindustrial times and over the past 20 years is of significant interest. Surface ozone measurements using modern instruments were made only during the past 25 years. Some quantitative measurements using wet chemical techniques were made in Europe in the 1950s [Staehelin et al., 1994], and one set of measurements dates from the turn of the century [Volz and Kley, 1988]. These measurements show that over Europe ozone in the lower troposphere at least doubled from the beginning of the measurements to the early 1980s. Many of the more recent measurements (since 1970) show that at least over Europe, and probably over other areas in the midlatitudes of the northern hemisphere, tropospheric ozone continued to increase through the 1970s and early 1980s [Oltmans et al., 1995]. An analysis of most of the recent data sets from surface stations (some located above the boundary layer) suggests that over at least the past decade there has been a significant slowing in tropospheric ozone growth at midlatitudes of the northern hemisphere [Oltmans et al., 1995]. Most other regions show no evidence for tropospheric ozone increases over the past 20 years and in some cases, such as South Pole Observatory, Antarctica (SPO), significant decreases are evident.
The four CMDL baseline observatory surface ozone data records are among the
longest available. The annual averages and long-term trends at each location
are shown in Figure 4.1. The numerical trend displayed in the figure is a linear
regression of the monthly mean observations. At Barrow Observatory, Barrow,
Alaska (BRW) there was a significant upward trend prior to 1990, primarily due
to summer increases. Smaller annual averages over the last 5 years have driven
the trend downward to show an overall small but insignificant increase. The
lower amounts in the 1990s are consistent with the results seen in the ozonesonde
record at the Canadian stations [Tarasick et al., 1995]. At MLO the overall
record beginning in 1974 shows a small but significant increase. At the two
southern hemisphere sites there are long-term decreases. At SMO this decrease
is not significant but at SPO a large and significant decline is evident. This
is most apparent after 1986. This decline at SPO was discussed in the 1993 Summary
Report [Peterson and Rosson, 1994].
Fig. 4.1. Annual average surface ozone mixing ratios in parts per billion
(ppbv) for BRW, MLO, SMO, and SPO. The solid line is a linear trend fit to the
monthly anomalies. The trend and 95% confidence levels in percent per year are
The monthly ozone means for each of the four CMDL baseline sites for the period
of observation are given in Table 4.5. For MLO the means are for the hours 0000-0800 LST,
which falls within the time of downslope flow at the observatory.
TABLE 4.5. Monthly Mean Surface Ozone Mixing Ratios (ppbv)
TABLE 4.5. Monthly Mean Surface Ozone Mixing Ratios (ppbv) - Continued
Monthly means are computed from daily (24-hr) averages.
Table 4.6 summarizes the 1994-1995 CMDL ozonesonde project involvement. This
includes supplying receiving stations and all ozonesonde supplies, training
where needed, personnel launching ozonesondes at several of the sites, and final
TABLE 4.6. Summary of 1994-1995 Ozonesonde Projects
|Boulder||52||Full year||52||Full year||NOAA long term|
|MLO||52||Full year||62||Full Year||NOAA long term + MLO3|
|SPO||69||Full year||69||Full Year||NOAA long term|
|McMurdo||65||Feb. 3-Dec. 25||6||Jan. 1-Feb. 12||NSF and NOAA|
|Tahiti||-||24||July 31-Dec. 29||PEM-Tropics|
|SMO||-||16||Aug. 1-Dec. 14||PEM-Tropics|
|Azores||30||May 5-June 3||42||June 2-July 27||AEROCE|
|Bermuda||10||Jan 21-May 31`||55||April 17-July 27||AEROCE|
|Maryland||-||12||April 13-May 16||AEROCE|
|Rhode Island||-||7||April 18-May 15||AEROCE|
|Newfoundland||-||20||April 12-Aug. 3||AEROCE|
|Nashville||-||14||June 27-July 21||Southern Oxidant Study|
|Indian Ocean||-||21||Feb. 12-April 14||NSF R/V Malcom Baldrige|
|Pacific Ocean||-||17||Oct. 17-Dec. 11||ACE R/V Discoverer|
PEM-Tropics - Pacific Exploratory Mission in the Tropics (a global
The CMDL long-term stations at Boulder, Colorado; Hilo, Hawaii; and SPO, continued
operating at one launch per week in 1994 and 1995, with SPO increasing to three
per week during the ozone-hole period. The SPO minimum total ozone, measured
by ozonesondes, reached 102 and 98 DU in 1994 and 1995, respectively. The minimum
profiles and the predepletion profiles are shown in Figure 4.2. Severe depletion
was observed in the 14-20 km region (nearly 100%) but did not extend down
to the 10-14 km region as it did in 1993 when a record low of 91 DU was
measured [Hofmann et al., 1994]. This extended ozone-depletion layer
in the lower stratosphere, observed in 1992 and 1993, was due to the effects
of the Mt. Pinatubo volcanic aerosol layer [Hofmann and Oltmans, 1993].
By 1994, the Mt. Pinatubo volcanic layer had decayed to background levels over
McMurdo Station, Antarctica [Deshler et al., 1996].
Fig. 4.2. Vertical profiles of ozone partial pressure in millipascals (mPa)
at SPO during the ozone hole of 1994 and 1995. The lighter line represents the
predepletion profile while the thicker line is the profile observed at the total
NOAA was also involved in regular ozonesonde and water vapor measurements at McMurdo Station, Antarctica, from February to August 1994 during a winterover project designed to study the development of polar stratospheric clouds using balloonborne instruments. The University of Denver, University of Wyoming, and NOAA conducted balloon flights to measure ozone, water vapor, nitric acid, and particle concentration profiles during the austral summer, fall, and winter of 1994 prior to and during the development of polar stratospheric clouds and ozone depletion. An early sign of ozone depletion was observed in the June 1994 profile in the 12-20 km layer [Vömel et al., 1995a] (section 4.2.3., this report). CMDL continued measuring ozone profiles on a weekly basis at McMurdo from November 1994 to February 1995. This was done in order to complete the first full year of ozonesonde profiles from McMurdo (February 1994-February 1995). The University of Wyoming launched ozonesondes during the ozone hole period from August to November 1994.
Weekly ozonesondes began at Tahiti and SMO in July 1995 as part of the Global Tropospheric Experiment Pacific Exploratory Mission in the Tropics (PEM-Tropics).
The intensive, short-term ozonesonde projects were all part of the AEROCE II and AEROCE III and the North Atlantic Regional Experiment (NARE). Ozone profiles were measured on a nearly daily basis from several sites (Table 4.6) in the spring and summer of 1994 and 1995 to investigate the sources (anthropogenic and natural) of high ozone layers in the troposphere over the north Atlantic Ocean region.
The 21 ozonesondes flown from the R/V Malcom Baldrige cruise in the
Indian Ocean began near South Africa at 30°S, 30°E and ended near
Sri Lanka at 7°N, 73°E. This was a preliminary study for the Indian
Ocean Experiment (INDOEX) planned for January 1998 to study the chemical and
radiative composition of the atmosphere over the Indian Ocean particularly in
the region south of the Indian subcontinent. The Aerosol Characterization Experiment
(ACE) R/V Discoverer cruise in the Pacific Ocean extended from 31°N,
214°E to 45°S, 145°E. This set of measurements provided the first
ozone profiles in a long cross section through the mid-Pacific.
4.1.5. ATMOSPHERIC WATER VAPOR
Monthly water vapor profile measurements continued at Boulder. As was noted
earlier [Oltmans and Hofmann, 1995; Ferguson and Rosson, 1992],
water vapor in the stratosphere over Boulder has increased significantly. The
updated trend information is summarized in Table 4.7. As was reported in the
past, the largest trends are seen in the lowest part of the stratosphere over
Boulder (16-20 km). This change of about 0.8% yr-2 is somewhat
less than reported earlier. This is primarily because the seasonal minimum which
occurs in winter and early spring was somewhat lower than in recent years (Figure 4.3).
This may be associated with enhanced transport from the tropics during early
1995. Lower stratospheric ozone amounts were also less than normal, indicative
of tropical transport. Above 20 km the increase is about 0.5% yr-2,
which is consistent with the expected increase resulting from increasing CH4
concentrations in the atmosphere.
TABLE 4.7. 1981-1995 Water Vapor Mixing Ratios Over Boulder, Colorado
|Level||Mean||Standard Deviation||Number of||Trend*||95% Confidence|
|(km)||(ppmv)||(ppmv)||Observations||(% yr-1)||Interval (% yr-1)|
*The trends are computed for deseasonalized values.
The 95% confidence interval is based on students t-distribution.
Significant at 95% confidence level.
Fig. 4.3. Layer average water vapor mixing in parts per million (ppmv) at
Boulder, Colorado, for the 16-18 km and 22-24 km layers. The dashed
line is the overall mean and the solid line is the linear trend fitted to the
monthly anomalies. The trend and 95% confidence levels in percent per year are
4.1.6. ATMOSPHERIC TRANSPORT
CMDL supports various research efforts to verify sources and sinks of trace gases and aerosols. The CMDL isentropic transport model [Harris and Kahl, 1994] calculates trajectories at requested elevations, including those in the stratosphere. Trajectories may then be compared to data collected at the surface or data collected at elevation (sonde, aircraft, and lidar data). Variations in concentration may be linked to transport where applicable.
Trajectories were used to describe seasonal flow patterns to MLO [Harris and Kahl, 1990], SPO [Harris, 1992], and BRW [Harris and Kahl, 1994]. A similar study of SMO flow patterns is underway. Highlights of this study appear as a special project at the end of this section (4.2.4). These four studies summarize many years of trajectories for each observatory in order to understand the average flow patterns, their meteorological causes, and the range of yearly and seasonal variations. Knowledge of transport characteristics has led to a better understanding of seasonal patterns in MLO methane data [Harris et al., 1992] and SMO carbon dioxide data [Halter et al., 1988], among other constituents.
Information about the transport model (methodology, description of plots,
and formats of data files) is available on the Internet at http://www.cmdl.noaa.gov/traj.
This home page also serves as the distribution site for "real-time"
trajectories. These are trajectories calculated from data downloaded twice a
day as it becomes available from National Centers for Environmental Prediction
(NCEP) (previously National Meteorological Center). For various CMDL baseline
and regional observing sites, trajectories are thus provided within a day of
when measurements are actually made. This capability will be expanded in the
future to include any site on the globe. The trajectory home page also includes
pointers to several archives of trajectory data.