|
Abstract
In atmospheric ozone monitoring careful data quality control is essential for correct analysis and evaluation of long-term ozone trends. In this connection, the first World Meteorological Organization (WMO) summary report on the comparison of total ozone measurements of Dobson and Brewer spectrophotometers (WMO：2003) states that the simultaneous operation of Dobson and Brewer instruments at the same station is highly recommended to improve the reliability of the total ozone measurements of the station, and indicates some seasonal change in the trend of total ozone difference between Brewer and Dobson. Recently, Vanicek (2006) presented the analysis results of the relation between high-quality simultaneous Dobson, Brewer ground and satellite total ozone observations, and suggested the difference between Brewer and Dobson to be attributed to the influence of temperature on ozone absorption coefficients and total sulfur dioxide. The Japan Meteorological Agency (JMA) Brewer MKIII UV network is run concurrently with a long-term Dobson ozone network comprised of stations at Sapporo, Tsukuba, Kagoshima, Naha, and Syowa Antarctica including Minamitorishima with Brewer MKII, which has replaced the old Brewer MKII UV network that operated from January 1990 to December 2001. These overlapped networks continue to store quasi-simultaneous total ozone comparison data as well as SO2. This paper presents the ds O3 observation results and the direct sun (ds) O3 difference between Brewer and Dobson without data correction by Brewer internal lamp tests as long-term and high-quality multiple Dobson-Brewer comparisons in the same network. The observation results from 2002 (2001) to 2007 are summarized below.
1) Monthly mean of daily ds O3 differences, expressed as the ratio of (BR－DB)/DB, ranged from ＋1.4 to －2.3% (maximum amplitude: 3.7%) at Sapporo (43.1N/141.3E), ＋1.5 to －3.1% (4.6%) at Tsukuba (36.1N/140.1E), ＋1.8 to －2.1% (3.9%) at Kagoshima (31.6N/130.5E), ＋1.1 to －1.7% (2.8%) at Naha (26.2N/127.7E), and ＋4.5 to －5.0% (9.5%) at Syowa Antarctica (69.0S/39.6E). The seasonal variation of ds O3 difference was clear at Syowa Antarctica and Tsukuba.
2) Seven-day average of daily ds O3 differences indicated "small seasonal differences" of the maximum from November to December and the minimum in June at Sapporo, "medium seasonal differences" of the maximum in January and the minimum from June to July at Tsukuba, "medium seasonal differences" of the maximum from December to January and the minimum from May to July at Kagoshima, "no seasonal difference" at Naha, and "large seasonal differences" of the maximum from April to September and the minimum from November to December at Syowa Antarctica.
3) The relations of "ds SO2" (daily mean of ds SO2), "AVG air mass" (daily mean of air mass at observation) and "O3 STD" (daily standard deviation of ds O3) with Brewer, versus the ds O3 difference are as follows. The first dependency of ds O3 difference on "ds SO2" was found at all stations, excluding Syowa Antarctica where the level of SO2 is very low. The slope of the linear regression was about －0.003 (from －0.0015 to －0.0048). A 10m atm-cm increase of SO2 thus produces about a 3% increase of Dobson ds O3. The second dependency on "AVG air mass" was not discernible at Kagoshima and Naha but was rather clear at Sapporo (slope: ＋0.00007), Tsukuba (slope: ＋0.00014) and Syowa Antarctica (slope: ＋0.00025). The third dependency on "O3 STD" was observed at Naha and Syowa Antarctica, and the slope was about －0.0014 and －0.0016.
4) The daily ds O3 difference between some Brewer observations (including MKII and MKIII) and Dobson observations in the last five years at Tsukuba also indicated seasonal variations. In contrast, the ds O3 difference between Brewer MKIII and Brewer MKII, expressed as the ratio of (BR MKII－BR MKIII)/(BR MKIII), did not exhibit seasonal variations.
5) The seasonal variation of ds O3 differences between Brewer and Dobson observations was reduced to a large extent by adjusting the Brewer ETC constants for ds O3 observation, but could not be removed completely.
6) The relation of temperature in Brewer instruments with the ds O3 differences was not clear.
7) Dependency of the ds O3 difference on turbidity (τ) with pyrheliometer was found at all stations, including Syowa Antarctica where the level of turbidity is very low. The percentage of the ds O3 difference decreased to －1% as turbidity increased from τ= 3.0 to 6.0, excluding Syowa Antarctica. The seasonal variation of daily turbidity at all stations was the same as the ds O3 difference.
8) SO2 is introduced by volcanic eruption as well as air pollution at all these stations but Syowa Antarctica. It is clear that the ds O3 difference, expressed as the value of BR MKIII－DB (m atm-cm), was closely related to the increased ds SO2 (m atm-cm) during ds O3 measurements. For example, the ds O3 difference of 6m atm-cm at air mass 2.5 on 276 JD 2004 was almost the same value of the ds SO2 at the time. These examples quantitatively verify the results of 3).
|
