Total column ozone measurements by the Dobson spectrophotometer at Belsk (Poland) for the period 1963-2019: homogenization and adjustment to the Brewer spectrophotometer

The total column ozone (TCO3) measurements by the Dobson spectrophotometer #84 have been carried out at Belsk (51°50', 20°47'), Poland, since March 23, 1963. In total, ~115,000 intra-day manual observations have been taken up to December 31, 2019. These observations were made for various combinations of double wavelength pairs in UV range (AD, CD) and the observation category, i.e., direct Sun, zenith blue, and zenith cloudy depending on the weather conditions. The 10 long-term stability of the instrument was supported by frequent (~almost every 4 yr.) intercomparisons with the world standard spectrophotometers. Trend analyses, based on the monthly and yearly averaged TCO3, can be carried out without any additional corrections to the intraday values. To adjust this data to the Brewer spectrophotometer observations also performed at Belsk, a procedure is proposed to account for: less accurate Dobson observations under low solar elevation, presence of clouds, and sensitivity of the ozone absorption on temperature. The adjusted time series shows that the Brewer-Dobson monthly averaged 15 differences are in the range of about ±0.5%. The intra-day TCO3 data base, divided into three periods (1963-1979,1980-1999, and 2000-2019), is freely available at https://doi.pangaea.de/10.1594/PANGAEA.919378 (Rajewska-Więch et al., 2020).

fixed temperature. There were several attempts to recalculate the O3AC (Redondas et.al., 2014) and finally the recalculation proposed by Serduchenko et al. (2014) has been recommended for the ground-based TO3 network. The application of these 40 temperature-dependent absorption coefficients significantly reduced the artificial seasonality in the Dobson-Brewer differences to less than 1% (Redondas et al., 2014). However, the data from the Dobson network has not yet been recalculated with use of the new O3AC, as knowledge of the stratospheric temperature and the vertical ozone distribution over the observing sites is required to obtain the effective temperature.
The ozone issue is still important as the ozone recovery rate expected from the regulations of the MP 1987, and its further 45 amendments, has also been driven by the recent climate changes (e.g., negative trends in the lower stratospheric ozone in the NH midlatitudes, Ball et al., 2018). Moreover, surprising increases of the CFC11 concentration in the troposphere have been found in some regions linked to ODS leakage from its long-term storing reservoirs (Lickley et al., 2020) and a return of ODS use in industry (Montzka et al., 2018;Dhomse et al., 2019). The ozone hole surprises with its variability from an extreme small extent in 2019 (Krzyścin, 2020) to extreme large one in 2020 (https://public.wmo.int/en/media/news/2020-antarctic-ozone-50 hole-large-and-deep?). Therefore, a question has to be contemplated as to whether it is still worth performing ozone observations by the Dobson spectrophotometer, which was designed and put into operation almost 100 years ago.
The Central Geophysical Laboratory of the Institute of Geophysics, Polish Academy of Sciences, at Belsk (51.84°N, 20.78°E), Poland, started monitoring of the atmospheric ozone (TCO3 and the ozone vertical profiles by the Umkehr method) on 23 March 1963. In Europe, there are only two stations with longer time series including Arosa (since 1926, Staehelin et al., 1998) 55 and Hradec Kralove (since 1961(since , Vaniček et al., 2012. The importance of the Belsk's TCO3 time series comes from the Dobson's measurements carried out regardless of the weather conditions excluding only the days with continuous rain or snow fall. Various efforts have been made to support the quality of the Belsk's TCO3 data (e.g., Dziewulska-Łosiowa et al., 1983;Degórska and Rajewska-Więch, 1991) as the less accurate zenith sky TCO3 observations in cloudy conditions were only available for some of the days, especially in the late autumn and winter. This paper presents further steps in the homogenization 60 of the Belsk's TCO3 time series for the period 1963-2019 in perspective of the long-term variability of the atmospheric ozone and possibility of a replacement of the Belsk's Dobson with the state-of-the art Brewer spectrophotometer.

The Dobson Spectrophotometer
The Dobson spectrophotometer is a double monochromator deigned to measure TCO3 by the technique of the Differential

65
Optical Absorption Spectroscopy (Dobson, 1957). The TCO3 values from Direct Sun (DS) observations are calculated based on the Beer's Law applied to spectral irradiances of solar radiation for selected wavelengths with strong and weak absorption by the ozone layer. These wavelengths are denoted as A (305.5/325.0 nm), C (311.5 /332.4 nm), and D (317.5/339.9 nm). The TCO3 values are determined from linear combination of the logarithm of the ratios between A pair and D pair irradiances that is called the double AD wavelengths pair algorithm to reduce the effects of the aerosols scattering on calculated ozone. It is 70 also possible to use C pair irradiances instead of the A pair, i.e., the so-called double CD pair, to obtain TCO3 if the solar irradiance at the shorter wavelength of A pair is too weak because of a large ozone absorption under high air masses (μ >3) and/or during cloudy conditions. Another option to calculate TCO3 by the Dobson spectrophotometer is to use scattered sunlight from the zenith sky (ZS) at the same wavelengths as in the DS observations. In this case, the same algorithm is implemented but the result needs to be reduced 75 by empirical charts. The charts after Rindert (1973) have been used throughout the whole period of the Belsk's observations for AD and CD double wavelength pairs. The operational data archived in the PANGAEA data base was obtained using the https://doi.org/10.5194/essd-2021-39  Komhyr et al. (1993). 80 Figure 1 shows time series of the number of daily TCO3 means obtained from DS intraday observations with AD and/or CD wavelength pairs in January and July for the 1963-2019 period. Number of the DS daily means per month is larger in June because of the usually less cloudy conditions in this month compared to January and the longer day length during the summer (i.e., greater chance for cloudless conditions). The monthly mean TCO3 for January is mostly based on ZS observations. It is worth mentioning that reduced number of DS observations (less than 15 per month) were possible in July for the period 1985-85 1995. In this period, ZS observations under almost cloudless conditions, i.e., the so-called zenith blue (ZB) observations, replaced DS observations, as the solar disc was not clear due to the large contamination of Belsk's atmosphere by industrial aerosols. At that time, the Belsk's observers were recommended to choose ZB type of the observations in clear-sky conditions.  The TCO3 values by ZS observations could be calculated for almost all-weather conditions (excluding cases with heavy and variable cloudiness) but they are reliable, i.e., close to the DS values, for the air masses up to 2.8 (Fig.2a). However, the empirical reduction curves (proposed by Rindert, 1973), which have been applied for the whole period of the Belsk's observations to convert ZS values to equivalent DS values, did not reflect all of the possible combinations of cloud types and their configuration in the sky. Therefore, the DS-ZS differences can sometimes exceed 5%. To eliminate a drift of the DS-ZS 95 differences, the following transfer function is used for large air masses: TCO3, ZS*= (1 +0.0125(μ -2.8) ) × TCO3, ZS , for mi μ ⸦ [2.8, 4.0] (1) TCO3, ZS*= 1.015 × TCO3, ZS , for μ > 4.0 1960ZS , for μ > 4.0 1965ZS , for μ > 4.0 1970ZS , for μ > 4.0 1975ZS , for μ > 4.0 1980ZS , for μ > 4.0 1985ZS , for μ > 4.0 1990ZS , for μ > 4.0 1995  where TCO3, ZS is the TCO3 value derived from ZS observations, μ is air mass during the Dobson observation, and TCO3, ZS* 100 is pertaining the corrected value for μ >2.8. The smoothed pattern of the DS-ZS differences (Fig.2b) is close to zero after the application of the transfer function. DS&AD observations provide the most accurate TCO3 values (Dobson, 1957) when the air mass is below 2.5. However, these 105 observations are not possible for northern sites like Belsk when the noon air mases above 2.5 appear in the period from 3 rd November up to 9 th February with the maximum air mass equal to 3.75 on 23 th December. Therefore, DS&CD observations are recommended for these periods (Dobson, 1957). Additional intraday TCO3 values, especially after 2005, were calculated using both DS&AD and DS&CD type of the Dobson's observations to find out a relationship between these TCO3 values.

115
The monthly mean TCO3 differences between various subsets of the TCO3 data are calculated from the whole period of the Belsk's observations are presented in Fig. 4. The differences are from the monthly averaging of the daily mean TCO3 values for the following classes: DS&AD, DS&CD, ZB, ZS, and for less reliable ZS observations when the sky was covered by clouds i.e., the so-called zenith cloudy (ZC) observations. For all ZS intraday values when μ≥ 2.8, the correction function (1) was applied. The TCO3 differences relative to the DS&AD values are within the ± 1% range almost throughout the whole year 120 excluding autumn/early winter months (November-December) where the AD observations are not recommended because of low solar elevation at noon. For each selected month, from January to September, the range between the monthly mean  1960 1965 1970 1975 1980 1985 1990 1995  differences is about ~1% for the all considered subsets of the Dobson observations. The maximum range of the difference is 4% in December, i.e., -1% for DS&AD versus DS&CD, and 3% DS versus ZC.  (1986, 1990, and 1995) with the WPS Dobson #83. Next intercalibrations (2000,2005,2009,2014) were organized at the European regional Dobson calibration center, the Meteorological Observatory Hohenpeissenberg, Germany. The local

135
European sub-standard, DI #64, was used as the reference instrument during inter-comparisons at the center. After the intercomparisons, the new R/N tables were obtained and used immediately, i.e., after the instrument arrival to the observing site, as input to TCO3 calculation algorithm. The retrieval software has not been changed since its first operational use in 1981 140 (Degórska et al., 1978).
The first homogenization of the Belsk's TCO3 data for the period 1963-1981 by Dziewulska-Łosiowa et al., (1983) was based on the Potsdam's R/N tables, extraterrestrial constants determined at Belsk in 1974 during the first DI intercomparison, the O3AC by Vigroux (1967), and zenith sky irradiances (for AD and CD observations) were reduced to the DS equivalent values by means of the empirical charts after Rindert (1973). The next homogenization was in 1991 by Degórska and Rajewska-145 Więch (1991). All data collected before 1992 was re-evaluated on the reading-by-reading basis using the O3ACs (at fixed temperature at -46.3 °C) after Bass and Paur (1985) that were recommended for processing Dobson TCO3 measurements since 1 January 1992.

150
The optical wedge calibrations taken during the whole period of the ozone observations at Belsk only slightly influenced the TCO3 values. Figure 5 shows the comparison of TCO3 calculated for the first 2-year period (1963)(1964) and last 2-year period (1980)(1981)  1963 -1964 1980-1981 1963-1964 and 1980-1981, respectively. The maximum difference between TCO3 by the 1979 Potsdam R/N tables and presently used the 2014 Hohenpeissenberg R/N tables is below 1 %. The small differences, less than 1% between TCO3 values are also found after comparison of DS&AD TCO3 in August 160 2020 derived using various R/N tables that were operationally used in the 1990s and 2020s (Fig.6). Therefore, it can be concluded from these comparisons of the DS&AD measurements that the Belsk's Dobson showed a stable performance over the whole period of the measurements. A linear approximation of TCO3 values in periods between the 165 calibration campaigns was not necessary to account for changes of R/N values and instrument's aging.

Adjustment to the Brewer spectrophotometer
The complementary monitoring of TCO3 and vertical profiles by the Brewer spectrophotometer #64 (BS64) Mark II (with single monochromator) has been started at Belsk in 1991. Like other Brewers, the BS64 is a fully automated, self-testing, PCcontrolled instrument designed for continuous long-term observations in all weather conditions. The quality of BS64 170 measurements has been supported by regular (yearly or at least every two years) comparisons with the international travelling reference instrument provided by the International Ozone Services Inc. (https://www.io3.ca/index.php). The BS64 instrument's constants re-defined after each comparison were immediately included to the operational software.
The Brewer TCO3 is derived from the weighted linear combination of the solar irradiances at five wavelengths (306.3, 310.0, 313.5, 316.8, and 320.0 nm) to eliminate noise caused by SO2 absorption in the UV range (Kerr et al., 1988   of the monitoring policy to less frequent intraday Dobson measurements carried out only near the noon during perfect weather conditions. The well-known seasonality of the Brewer-Dobson differences, i.e., underestimation of the Dobson TCO3 in the cold period of the year, could be observed after the smoothing by the locally weighted scatterplot smoothing (LOWESS, Cleveland, 1979) applied to time series of Brewer-Dobson differences.
where Teff is the effective (weighted by the ozone vertical profile) temperature, TCO3,OLD is the TCO3 value at the fixed 200 temperature, and TCO3, NEW is the temperature corrected value. Prior application formula (2) Koukouli et al. (2016) found that the ECMWF effective temperature was in good agreement with 205 that from the ozone sounding in northern hemisphere midlatitudes. provides that the seasonal difference is of ~4.5% as calculated from the monthly mean difference value of ~3.5% and -1% in January and July (or August for ZS data), respectively. After application formula (2), the difference is smaller, i.e., ~ 2.5% (DS) and 3% (ZS).
It is worth mentioning that the monthly Brewer-Dobson differences for the temperature corrected data are out of ± 1 % range only in January and December (DS subset) and in addition in November for the ZS subset. To the authors knowledge, the 215 Brewer-Dobson differences for less accurate ZS subset have not yet been discussed by other authors. It seems that a stray light effect, which may be different in the Dobson and Brewer instruments (Belsk's BS64 is a single monochromator unit), is responsible for larger differences between these spectrophotometers in periods with low solar elevation at noon.

225
The stray light within the instrument causes underestimation of the Dobson TCO3 of about few percent when slant TCO3, ~μ×TCO3 (in DU), is over 800-900 (Evans et al., 2009). This observation also concerns the single monochromator Brewer spectrophotometer (e.g., Karppinen et al., 2015). Here, it is decided to construct a simple correction function (i.e., linear function of μ) that will move all-monthly means Brewer-Dobson differences for DS and ZS subset (

Results and Discussion
The following subjects were considered to construct the Brewer adjusted record of the Belsk's Dobson measurements for the whole period of the observations : the focus during the inter-calibration Dobson campaigns was on DS&AD type 245 of observations, DS&AD observations have been well calibrated during the whole period of observations (within 1%), the inter-comparisons were carried out in the warm period of the year (usually in summer), there was the seasonal difference between the Dobson and Brewer spectrophotometers depending in some extent on the effective temperature and individual instrument stray light sensitivity. It is worth mentioning that BS64 will ultimately be the primary instrument for the ozone monitoring at Belsk. First step of the recalculation was to correct ZS Dobson's observations taking into account a drift relative to DS&AD values (Figs.2). Application of the correction function (1) removed this shift (Fig.2b). Moreover, Fig.3a and Fig. 3b provided that further correction to other types of the measurements was not necessary to eliminate a long-term drift relative to the qualitycontrolled DS&AD subset. Thus, ZS-DS conversion charts proposed by Rindert (1973) were valid throughout the whole period of the observations suggesting small changes in cloud characteristics in the period 1963-2019.

255
Next step of the Dobson recalculation was to account for the temperature dependence of O3AC using the correction function (2) based on very recent O3AC derived by Serdyuchenko et al. (2014). Application of this formula removed part of the bias between the Brewer and Dobson spectrophotometers both for DS and ZS subset of the observations (Figs.7-8). The agreement within ± 1% range exists in almost all months but not in months with low solar elevation at noon (November-December-January). The last step of the homogenization is to eliminate the difference between the spectrophotometers that has been 260 found in periods of low solar elevation. This is probably related to a technical problem of stray light contamination of TCO3 observations for large air masses.  of about a few percentages are found in the late autumn and early winter months due to: the TCO3 underestimation found for 265 ZS observations, increase of O3AC with decreasing effective temperature, and the stray light effect. For warm part of the year (April-September) and corresponding higher solar elevation, the differences are only in the range of ± 0.5 % range. Figure 9 show that there was an apparent decline in the effective temperature in December for the period 1963-2019 of ~ -4 °C and slight decline in July of ~ -1 °C. This changed only slightly (i.e., 0.4% in December, and ~0.1% in June) TCO3 values at the end of time series comparing to the TCO3 values at the start.  Figure 10a and 10b show the long-term variability of the monthly mean TCO3 for December and July, respectively, derived from the original data (archived in the PANGAEA data base) and the data after application of all correction functions (i.e., the

275
Brewer adjusted Dobson data). The monthly means were calculated averaging the daily means from DS observations. If there were no such observations within the day, the daily means based only on ZS measurements were considered. The decline in the effective temperature ( Fig.9) does not change the variability of the long-term pattern of the smoothed TCO3 time series. A constant upward shift is found in the colder months (e.g., Fig.10a) and a slight downward shift in the warmer months (e.g., Fig.10b). A constant shift (upward) is also seen in the yearly mean values calculated as the average of all monthly means for 280 each year (Fig.10c). Therefore, the trend estimates in DU will be almost the same if the differences in DU relative to the reference TCO3 value (e.g., the TCO3 monthly means before 1980) are considered. formula (1) and (3) are linear function of air mass, so it can be easily calculated by any user. The effective temperature can be taken from the TEMIS data base and the correction for the ozone absorption dependence on temperature given by formula (2) also requires simple calculations. Thus, it will be possible to use the homogenized Dobson data for comparative studies with other TCO3 data sources.
The present analysis shows that the original data (archived in PANGAEA data base) can be used in trend analyses based on 305 the TCO3 monthly and yearly means. Application of all described corrections to the original data provides that the resulting smoothed time series exhibits a constant shift (upward for the cold season and much less downward for the warm season) relative to the original long-term pattern. This finding supports the validity of previous trend analyses of the Belsk's Dobson data (e.g., Krzyścin and Rajewska, 2009;Krzyścin et al., 2013;Krzyścin et al., 2014). The resulting Brewer adjusted Dobson TCO3 values are found in good agreement with the Brewer data. This will allow for constructing the Dobson-Brewer merged 310 time series as the Brewer observations will soon replace the Dobson measurements at Belsk.
Author contributions. JK wrote the paper and did statistical analysis, BR carried out the data processing, and JJ provided the Brewer data.
Competing interests. The authors declare that they have no conflict of interest.