Assessment of Ground Ozone Level under The Physiological Strain Conditions

Urban areas are characterised by the impact of negative environmental factors, such as: stress connected with extreme bio-thermal conditions or the presence of high concentrations of air pollutants. This study aims to evaluate the relationship between the hours of O3 concentrations and the levels of physiological strain (PhS) in Legnica, during the period from December 2013 to November 2014. The hourly concentrations of O3, NO2 and meteorological elements used in the study were obtained from the State Environmental Monitoring station in Legnica (Lower Silesia). The evaluation of the bio-thermal conditions was carried out by means of the physiological strain indicator (PhS). The basic statistics were subject to analysis, the frequency of hourly pollution concentrations and the thermal strain were evaluated, as was the Pearson correlation coefficient and multiple regression between O3 and PhS. A detailed analysis was carried out for the summer months (June-August). The most adverse conditions in terms of pollution with tropospheric ozone and heat strain were noted in July. The strongest relations between O3 and PhS were observed in June. In winter months (December-February) no significant dependencies were noted between the tested variables. These tests will help to contribute to increasing the current knowledge on evaluating the thermal comfort of urbanised areas and the accompanying aero-sanitary conditions.


INTRODUCTION
Urban areas modify the atmospheric environment, for instance through the changes of the city's thermal conditions as a result of progressive development and loss of vegetation [Nikolopoulou and Lykoudis 2006, Boumans et al. 2014, van Hove 2015], and changes to air chemistry through the presence of particulate and gas pollution, which includes tropospheric ozone [Carnero et al. 2010].The tests carried out in Seattle showed an approx.10% increase in A&E visits caused by asthma attacks along with increased levels of O 3 by each 10 ppb [Goodin, Hubbard 2016].Additionally, among the children aged 6-15, O 3 and PM 2.5 particulate lead to a decrease in the lung capacity [Chen et al. 2015].Novack et al. [2016] showed the impact of air pollution on the oxidative stress of a human body.In their tests, they noted an inverse relation between the high con-centration of O 3 and a positive relation between suspended particulate PM 10 and multiplication of blood stem cells.The impact of tropospheric ozone concentration on the respiratory system is confirmed by numerous studies from all over the world: Kinney et al. [1996], Frischer [1999], O'Lenick et al. [2007], Khatri et al. [2001], Tager et al. [2003], Uysal and Schapira [2005].The research also shows an increase in the death rate by 2% due to hypofunction of the cardiovascular system caused by O 3 pollution on each 10µg•m -3 [Khaniabadi, et al. 2017].Tropospheric ozone is a secondary pollutant that arises from a series of photochemical reactions [Clapp and Jenkin 2001, Kleinman 2005, Monks et al. 2015, Tiwari et al. 2015].Moreover, the solar radiation necessary to start the chain reaction, which includes: NO X (NO + NO 2 ), CO, VOC's and NMVOCs, is absorbed by the human body.The absorption depends on the factors such as: the intensity and structure of solar radiation, the Sun's altitude (h), ground albedo, body-to-sun orientation and the insulating power and colour of clothing and skin [Błażejczyk 2004].The highest concentration of The aim of the study was to evaluate the relationship between the hours of O 3 concentrations and the levels of physiological strain (PhS) in Legnica, during the period from December 2013 to November 2014.The data was obtained from the State Environmental Monitoring Station in Legnica, marked by the international code PL0190A, indicating a city type station.The data was prepared for universal time (UTC).The initial data verification was carried out on the basis of the correlation between the O 3 hourly concentration and total radiation (Rad, W•m -3 ), air temperature (Ta, °C), relative air humidity (Rh, %), as well as the wind speed (v, m•s -1 ) analysis which showed occurrence of two separate data sets.The daily distribution of average values of the meteorological elements showed a sudden drop in the total radiation values between 12:00 and 14:00, depending on the particular month analysed within the researched period, probably caused by the shadowing of the station in the afternoon.The basis of the division into two data sets was the result of carrying out a distribution of average hourly values of meteorological elements for individual decades of each month.The data sets obtained in this way were further used in the correlation analysis.

Legnica
The evaluation of pollution with the concentrations of O 3 and NO 2 in particular months of the research period was carried out on the basis of descriptive statistics and an analysis of hourly concentration frequencies of O 3 and NO 2 in the adopted classes of pollution levels, consecutively for ozone: 0-40, 41-80, 81-120, >120 µg•m -3 and for nitrogen dioxide: 0-20, 21-40, 41-60, 61-80, >80 µg•m -3 .The evaluation of the bio-thermal conditions was carried out on the basis of the physiological strain (PhS, dimensionless) (Table 1).
The PhS indicator values were calculated by means of the Bioklima 2.6 software (https://www.igipz.pan.pl/Bioklima-zgik.html).The PhS indicator is expressed as the relation of heat loss from a human body by convection (C) to its loss as a result of perspiration evaporation (E) [Błażejczyk, Matzarakis 2007]: (1) For each month, there is a frequency related to the occurrence of different concentrations of physiological strain.In turn, the occurrence frequencies of the adopted O 3 concentration in particular classes of physiological strain concentra- tion were evaluated only for the chosen months characterised by a large and moderate hot strain.
For the month of the most frequent extreme conditions in the O 3 concentrations above 120 µg•m -3 and hot strain load, an analysis of the daily distribution of O 3 , NO 2 , values PhS concentration and meteorological elements in subsequent decades was conducted.Due to the asymmetric, righthand character of the hourly NO 2 concentration schedule, this data was adapted to a normal schedule.In order to do so, the hourly data was stated as a logarithm, which enabled carrying out further statistical analyses.An analysis of the Pearson correlation coefficient between the O 3 and meteorological elements and PhS indicators was carried out for three groups: including the full set of all-day measurements data, morning measurements, that is from sunrise to 13:00 and afternoon measurements data -from 14:00 to sunset.The correlation was evaluated on the level of p=0.05.
For the months when the correlation analysis was statistically significant, the equations of multiple regression analyses between tropospheric ozone and conditions of physiological strain in the presence of NO 2 were built.

RESULTS
The descriptive statistics showed that from March to September, the maximum hourly O 3 concentration values in particular months exceeded 127 µg•m -3 (Tab.2) .The highest concentration was recorded in July at 159 µg•m -3 .In turn, the highest NO 2 concentration in the summer period fell in August and was 100 µg•m -3 .In the same month, the median value calculated for the NO 2 emission was equal to 15 µg•m -3  .In the month when the highest O 3 concentration was documented, the value of the PhS indicator amounted to -0.01, which indicates the occurrence of "extreme hot strain" conditions.In the remaining summer months, i.e.June and August, the maximum values of the PhS indicator showed the conditions of "great hot strain".In the winter months (December-February) the maximum monthly O 3 concentration did not exceed 70 µg•m -3 , the highest recorded in January -69 µg•m -3  .In the same month, the maximum value of the PhS indicator amounted to 6.06.This confirms the occurrence of "great cold strain" in this month.The maximum concentration of NO 2 was documented in March at 128 µg•m -3 ; moreover, the maximum concentration above 100 µg•m -3 was also recorded in December and February.The value of the third percentile indicates that in July, 25% of all O 3 concentrations were over 98 µg•m -3 ; in comparison, in January, 25% of all concentrations were within the range of 36-64 µg•m -3  .The highest median values above 50 µg•m -3 occurred in July, May, August, June, and April.Thus, the median value for the spring-summer months was higher than the value of the third percentile for winter months.
In July, around 40% of all cases constituted the hourly concentrations above 80 µg•m -3 ; out of that concentration >120 µg•m -3 amounted to approx.11.5% (Fig. 1).March was characterised by the least favourable conditions in terms of the frequencies hourly concentrations of NO 2 ; the concentrations above 80 µg•m -3 constituted 2.8% of recorded cases, in turn, in the scope of 60-80 µg•m -3 -4.2%.In July 2014, the conditions connected with hot strain constituted 30% cases.In June and August, the frequencies were comparable, amounting to approx.11% and 9.7%.In the winter period (December-February), there were almost the same conditions connected with "moderate cold strain", except for January when 18.4% of cases were characterised by "great cold strain".
In the remaining months, the strain did not exceed 1%.In the researched period, there were no extreme thermal strain days.Only in July was there a single case of "extreme hot strain".The occurrence of "moderate hot strain" in September was 1.4%, in May approx.4.5%.The summer period, from June to July, was characterised by the occurrence of thermal conditions from "moderate cold strain" to "extreme hot strain".In July and in June the frequency of O 3 concentrations above 120 µg•m -3 was connected with the conditions of hot strain.Extreme conditions were re-corded only in July, which constituted an isolated case.In June, during the "moderate hot strain" approx.93% cases involved the concentrations above 120 µg•m -3 .In turn, August was characterised by, in comparison to the remaining summer months, a lower frequency of O 3 concentrations above 120 µg•m -3 .The average hourly O 3 exceeded the value of 100 µg•m -3 in the second and third decades of July, from 10:00 to 18:00 and from 11:00 to 17:00 (Fig. 2).In the winter period, there was no significant correlation between the hourly O 3 concentrations and the values of the PhS indicator (Tab.4) A weak correlation (r=0.16,p=0.01) for the data set from 24 h in January constituted an exception.This proven relation has a positive direction, in contrast to the proven relations in the remaining months, which adopted a negative direction (March-November).The reason for the change in the direction of relations between the O 3 concentrations and the PhS values in January could be a lack of correlation between O 3 and Ta for all possible data sets.The strongest, negative relation (r = -0.87,p=0.01) was recorded in June for the afternoon data set, between 14:00 and 19:00, that is sunset.For the data from 24 h, the strongest negative correlations in June and July, were r = -0.78(p=0.01) and r = -0.76(p=0.01),respectively.In the spring period (March-June) and in September, clear differences in the relations between the morning data sets -from sunrise to 13:00 and in the afternoon -from 14:00 to sunset, were observed.In July and August, the strength of the correlation between O 3 and PhS was comparable both for data from 24 h and from morning and afternoon data sets.In June in the afternoon hours, the strongest correlation between O 3 and Ta (r = 0.90, p=0.01) was observed.The relation between the hourly O 3 concentrations and total radiation for the data set of afternoon measurements in May and June did not show statistical relevancy on the level of p=0.05.The model of rolling multiple regression was prepared for the months in which there was a significant correlation between O 3 and PhS (Table 5).When examining the model, apart from the PhS indicator, additional hourly NO 2 values in the form of the common logarithm were taken into account.In the adopted models, significant impact of all changeables explaining the size of O 3 immision was confirmed.The greatest fit of the empirical data to the regression function, amounting to 76% and 80%, was obtained for June, for the conditions in the morning and afternoon; in turn, the smallest fit was observed in April (58%, 43%) and in November (50%, 39%).On the basis of the partial correlation factor, stronger connections between O 3 and bio-thermal conditions, other than NO 2 pollutions were proven in the afternoon hours.In the morning hours, a greater impact of NO 2 on O 3 was recorded, in the prepared model.

DISCUSSION
The research carried out between December 2013 and November 2014 showed a significant relationship between the level of tropospheric ozone and unfavourable bio-thermal conditions connected with hot strain.These relations are   [2014] showed a significant connection between the urban heat island in Gdańsk (Poland) and the level of researched pollution concentrations.

CONCLUSIONS
July 2014 was characterised by the least favourable bio-thermal conditions, determined on the basis of the physiological strain indicator (PhS), which were accompanied by an increased risk of occurrence pertaining to the high concentrations of tropospheric ozone above 120 µg•m -3 in comparison to the remaining months.These conditions mainly occurred during a significant and moderate hot strain in the afternoon hours, especially in the second decade of the month.Low relative air humidity, high air temperature and weak wind speed had an impact on the deterioration of the bio-thermal conditions and the increase of the O 3 concentration in the summer.Despite the least favourable levels of the O 3 concentrations and PhS values in July, the strongest relations were confirmed in June.The tests carried out may contribute to expanding the current knowledge in the area of thermal comfort evaluation of urban areas and the accompanying aero-sanitary conditions.
O 3 occurs at noon, showing an inverse meridian course from NO 2 , NO and NO x [Escudero et al. 2004, Kalbarczyk et al. 2015, Rozbicka and Rozbicki 2016, Zheng et al. 2017].In the same hours, the least favourable bio-thermal conditions are observed that impact the human body [Park et al. 2014, Kalbarczyk et al. 2015].High O 3 concentration is often accompanied by a high air temperature [Vandentorren et al. 2004, Pascal et al. 2012].Błażejczyk and McGregor [2007], on the basis of 3 bio-thermal indicators: STI, PST, PhS, showed that in London it is possible to explain 20-29% deaths by the certain values of these indicators occuring three days earlier.In Poland, according to Błażejczyka and others [2017], one may expect an increase in mortality on the days characterized by strong and very high heat concentration, with reference to the days when heat stress is observed, by 12% and 47% respectively.
is located in the west of Poland, in the Lower Silesia region (51°12′36″N, 16°09′42″E, hs = 113-168 m. a.s.l.).It is the eighth largest (56 km 2 ) and densely populated (1802 people•km -2 ) city in the region and the third largest in terms of population (100886 people) after Wrocław and Wałbrzych [Central Statistical Office 2015].Legnica belongs to the central bioclimatic region of weak stimulusity, characterised by an average air temperature at 12:00 (UTC) of 21.3°C in the summer and 1.9°C in the winter [Błażejczyk and Matzarakis 2007].This study uses the data of hourly pollution concentrations of O 3, NO 2 and the meteorological data during the period from December 2013 to November 2014.
NO 2 does not show a high changeability of average daily concentration in particular decades.The highest concentration in the second decade was recorded at night.The daily distribution of average O 3 concentrations shows a reverse distribution in reference to the PhS values.The lower values of PhS indicator show the occurrence of the hot strain in the second decade, where the least favourable PhS conditions are recorded, which overlaps with the high level of O 3

Fig. 1 .
Fig. 1.Occurrence frequencies: the hourly concentrations of gas pollution O 3 and NO 2 (a -b), thermal strain in individual months of the research period (c), and hourly O 3 concentrations in individual classes of physiological strain in June, July, and August (d -f), for the period from December 2013 to November 2014

Table 2 .
Descriptive statistics for individual months of the research period: Dec. 2013 -Nov.2014

Table 3 .
Maximum O 3 concentration at particular hours of July 2014 in reference to the accompanying meteorological and bio-thermal conditions Kleinman 2005, Monks et.al. 2015, Tiwari et al. 2015].At the turn of the second and third decade

Table 5 .
Equation of the rolling multiple regression between O 3 and NO 2, and PhS for individual months.inJuly, a significant increase of average hourly O 3 concentrations and the loss of average PhS values at night was observed, which could indicate the shaping of the urban heat island effect at that time.Czarnecka and Nidzgorska-Lencewicz