1. Introduction
The danger caused by large-scale, global and regional pollution started to be recognised in the 1960s. Such pollution might end up resulting in irreversible changes in both terrestrial and ocean ecosystems and global climate change. The research and monitoring efforts required to detect the changes in the atmosphere at global and regional scales must be based on broad-ranging international cooperation. It was, first of all, international institutions (World Meteorological Organization, United Nations Economic Commission for Europe ECE, United Nations Environment Programme) that initiated, in the 1960s and 1970s, the first international monitoring programmes [
1].
To support the above-mentioned programmes, Czech Hydrometeorological Institute (CHMI) established the National Atmospheric Observatory Košetice (NAOK), specialized in monitoring and research of air quality at the background scale of Czech Republic.
After the political changes in 1989, the air quality control and protection became one of the most important political priorities in the Czech Republic. Immense funds were invested in emission reductions (mainly from large power plants) in the Czech Republic during the 1990s, resulting in a marked improvement in the air quality, the levels of which in some regions had previously ranked among the worst in the world. Nevertheless, the growing industry and traffic after 2000 have caused the air quality in the Czech Republic to begin to deteriorate again. Irresponsible conduct of individuals who use low-quality fuels or even municipal waste in household heating systems, emitting hazardous chemicals to the air, is a contributing factor that cannot be neglected. Fine dust is the most serious problem at the moment. The Ministry of the Environment developed a National Emission Reduction Programme of the Czech Republic in 2007, and it has been approved by the government. The document comprises several key measures to contribute to an improvement in the current state of the environment and environmental and health protection.
The objective of the study is to detect the long-term trends of air quality at the background scale of the Czech Republic. Thirty-year data series is sufficient for detection of long-term trends of air quality. The study is based on the data generated within the National Air Pollution Monitoring Network, stored in Air Quality Information System and annually published e.g., [
2]. Generally, the development of air quality in the last three decades was affected by various circumstances: the essential political changes in Central and Eastern Europe in the end of the 1980s brought a substantial decrease in emissions in the Czech Republic and more widely in the Central European region thanks to international conventions and also economic and political development. The meteorological conditions for long-range transport in Europe were changed as well, and the question of global climate change assumed importance. Measurement techniques showed significant improvement, as did our knowledge concerning the behaviour of air pollutants in the atmosphere. All these aspects influenced the long-term trends at the background scale of the Czech Republic very significantly.
3. Results
Political and economic changes after the fall of the iron curtain brought a general drop in industrial production and later significant changes in the structure of the industry. These changes were reflected in the reduction of sulphur emissions in the Czech Republic by almost 90% in the period of 1990–2000 (
Figure 2). The results of long-term monitoring show that the emission decrease was reflected in reduced pollution levels. Sulphur dioxide concentrations in the atmosphere declined nearly by the same order of magnitude as the emissions (
Figure 3c). The steep drop of SO
2 concentrations was more pronounced in the 1990s. The frequency of episodes with extremely high concentrations decreased rapidly (
Figure 3e). In the new millennium, the mean annual concentrations dropped below 5 μg·m
−3, but a slightly decreasing trend was found also in the period 2001–2019 (
Figure 3f). In the EMEP domain, SO
2 emission reductions started in the 1980s–1990s; therefore, changes in concentrations will have occurred earlier than 1990. However, concentrations have continued to decrease continuously during the period under review. The timing of concentration decreases varies between countries according to national implementation of emission reduction strategies, but on average, the decrease was larger in the early 1990s and levelled off since then [
5].
In the period of 1994–2012, no trend of nitrogen oxides concentration was found, in spite of the fact that the nitrogen emissions declined by 54% during the period under review (
Figure 3b). In the period of 1990–2012, the emissions dropped by 72% and in the EU by 55% [
6]. Mean annual concentrations varied around 10 μg·m
−3. These results were in very good correspondence with the trends at the background level in the neighbouring countries (Austria, Germany) [
5]. The reasons are uncertain. One of the explanations could be the significant changes in the structure of nitrogen emissions. In the last 7 years, a visibly decreasing tendency of NO
2 concentrations was found and the mean annual concentrations dropped continuously to 4 μg·m
−3. The evaluation of the data from the EMEP network shows that for the period of 1990–2001, the fraction of sites where significant negative trends were observed was high (58%), but it slowed down after 2002.
NO concentration at the background scale is quite low, and mean annual concentrations varied around 1 μg·m−3. The long-term trend describes similar patterns for NO2: no trend in the period 1994–2012 and decreasing tendency after 2012.
A slightly decreasing trend was found in mean annual concentrations of tropospheric ozone in the whole period and also in the first part of the period under review (
Figure 3c). On the contrary, a slightly increasing tendency was found after 2006. It is caused probably by increasing temperature during the last two decades. A warm period displays similar patterns as the whole year. On the contrary, no trend was found in the cold period [
7]. More importantly, the number of episodes exceeding the target value for human health dropped significantly during the period (
Figure 4), and interannual variations can be explained by meteorological conditions. The target value of tropospheric ozone for the protection of human health is exceeded when the eight-hour running mean is higher than 120 μg·m
−3 25 times on average for 3 years. Visibly higher values were recorded in the years with extreme summer temperatures over long periods and well-established heat waves over continental Europe (2003, 2015, 2017).
A statistically significant trend was found for carbon dioxide. Mean annual concentrations decreased continuously during the whole period (
Figure 3a).
Most non-methane volatile organic compounds (VOCs) follow an annual course that reflects their emission levels, i.e., with maximums in winter and minimums in summer. Isoprene is an exception. In general, the reduction of VOCs emissions in the last two decades was reflected in a decrease of concentrations at the regional scale of the Czech Republic [
8]. A statistically significant downward trend was found for almost all of measured VOCs, and only the ethane trend was less significant (
Table 2). The trend of isoprene concentrations is controlled first of all by natural conditions and shows different patterns from other VOCs. We detected a statistically very significant upward trend of isoprene concentration in the period under review. Favourable conditions for isoprene emissions are in hot summer periods. An increasing tendency was visible especially in the last decade. This is in good correlation with increasing mean annual temperature in the current period of changing climate conditions (hot summers, long periods with high temperatures). It follows from the current report on VOC measurements in the context of EMEP [
9] that the VOC concentrations continuously decrease on a regional scale and thus reflect the decreasing trend in emissions. The concentration level at NAO Košetice is comparable with those at the German, Swiss and French stations. The Czech station has long been characterised by lower annual average ethane concentrations. For most VOCs, the concentrations measured in the winter are usually similar to those at German stations, while the values at NAOK are slightly lower in the summer.
The measurement of aerosol particles covers periods of different duration. The longest records are available for sulphate in aerosol and PM
10.
Table 3 shows that the concentrations were changed significantly during the period under review. Changes in concentration levels reflect the development of emission (
Figure 2), which is in line with both national and international environmental measures.
Outcomes of sulphur and PM
10 show the highest level of trend significance (
Figure 5c,d). Sulphur concentration continuously decreased during the whole period. On the other hand, the evaluation of PM
10 data shows that the mean annual concentrations in the period of 2001–2006 reached a similar level as in 1996 (29.8 µg·m
−3) (
Figure 5c). The same patterns were observed across the Czech Republic at different types of stations. After 2001, the drop of emission was slower compared to the previous period. The increase of PM
10 concentrations was probably influenced by meteorological and dispersion conditions [
10]. A higher level of trend significance is observed for PM
2.5 concentrations. After 2005, when a level over 18 µg·m
−3 was recorded, the linear decreasing trend is observed. These outcomes are analogous for PM
10.
An insignificant trend is visible for the sums of ammonium and nitrates (
Figure 5f,g) that are measured from 2002. This is in line with the fact that the emission development is more or less at the same level (
Figure 2). No visible annual variation was found (
Table 3). Concentrations of elemental (EC) and organic carbon (OC) dropped from 2010 (
Table 3), but only for EC is the trend statistically significant (
Figure 5a,b).
4. Summary
Generally, the results show that the fundamental drop in emission of basic air pollutants in the Czech Republic and widely in the Central European region in the period under review was reflected in the significant decrease of air pollution levels at the background scale of the Czech Republic. A statistically very significant drop in mean annual concentrations of sulphur dioxide was detected in the period of 1990–2000. After 2000, the mean annual concentrations dropped below 5 μg·m−3, but a slightly decreasing trend was found also in the period of 2001–2019.
No trend was found by the evaluation of nitrogen dioxide in the atmosphere in the period of 1990–2012, in spite of the fact that the nitrogen emissions declined by half during the period under review. In the last 7 years, a visibly decreasing tendency of NO2 concentrations was registered, and the mean annual concentrations dropped continuously to 4 μg·m−3.
A slightly decreasing trend was found in mean annual concentrations of tropospheric ozone in the whole period and also in the first part of the period under review. On the contrary, a slightly increasing tendency was found after 2006. It is caused probably by increasing temperature during the last two decades. More importantly, the number of episodes with the target value for human health exceedances dropped significantly during the period.
The reduction of VOCs emissions in Central Europe was reflected in a statistically significant decrease of concentrations at the regional level of the Czech Republic. Only isoprene, which is of natural origin, displays an inverse trend.
Sulphur concentration in aerosol continuously decreased during the whole period. The evaluation of PM10 data shows that the mean annual concentrations in the period of 2001–2006 reached a similar level as in 1996. The higher level of trend significance is observed for PM2.5 concentrations, but the general outcomes are analogous as for PM10. Concentrations of EC and OC dropped from 2010, but only for EC is the trend statistically significant.