Determination of BVOCs based on high time-resolved measurements in urban and forest areas in Japan

Biogenic volatile organic compounds (BVOCs) with high photochemical activity and short atmospheric lifetimes are major contributors to tropospheric ozone and other photochemical air pollution. Although several studies have been conducted on BVOC emissions in Japan, no comprehensive observations have been made to determine the actual state of BVOCs in the atmosphere. Therefore, we conducted time-resolved measurements of atmospheric BVOCs in urban and forested areas throughout the year. The concentrations of BVOCs were higher in summer than in the other seasons. Isoprene concentrations were higher during hours with higher temperatures and solar radiation. However, there were also months and times of the year when monoterpenes showed high concentrations, which indicates that the behavior of the BVOC components differed depending on the time of the year. The results of the propylene equivalent concentration indicated that BVOCs considerably contributed to tropospheric ozone production. The year-long observations of BVOCs in this study contribute to our understanding of the actual status of atmospheric BVOC concentrations and components and the uncertainty in the calculation results of chemical transport models.


Introduction
In Japan, the achievement rate of the environmental standard for photochemical oxidants (standard in Japan, 1-h concentration less than 60 ppbv) has remained extremely low since its establishment in 1973.The achievement rate for FY2021 announced by the Ministry of the Environment of Japan (MOEJ, 2023) was 0.2%.The low achievement rate is recognized as a long-term atmospheric environmental problem in Japan, requiring improvements.Hereafter, the term ozone is used for photochemical oxidants because tropospheric ozone accounts for a large proportion (MOEJ, 2010).
High concentrations of ozone have been reported to reduce plant yield (Gu et al., 2018) and adversely affect human health by irritating the respiratory tract and mucous membranes (Kariisa et al., 2015;Sacks et al., 2014).In addition, ozone is the third most radiativeforcing greenhouse gas after carbon dioxide and methane and is a short-lived climate pollutant (SLCP) that can impact the local climate (Intergovernmental Panel on Climate Change, 2021).
The concentrations of volatile organic compounds (VOCs) and nitrogen oxides (NOx), which are ozone precursors, have been decreasing owing to various efforts such as regulations under the Air Pollution Control Act and automobile emission controls in Japan (Ohara et al., 2020).Nevertheless, the annual mean ozone concentrations have either remained consistent or increased slightly (MOEJ, 2017;2022).It is desirable to clarify the causes of this trend to promote ozone control measures.
A key issue to be considered in resolving the trend is the effect of biogenic VOCs (BVOCs) (MOEJ, 2017).
Most BVOCs are released through the leaf stomata, and there are dozens to hundreds of component species, including terpenoids, alkanes, alcohols, carbonyls, esters, ethers, and aldehydes (Okumura, 2021;Tani & Mochizuki, 2021;Yang et al., 2021).Terpenoid is the generic name for a group of natural products with five-carbon compounds as their building blocks including isoprene (C 5 H 8 ), monoterpenes (C 10 H 16 ), sesquiterpenes (C 15 H 24 ), and diterpenes (C 20 H 32 ), which are thought to account for the majority of BVOC emissions.Global BVOC emissions have been reported to be approximately 1000 Tg/y (Guenther et al., 2012).However, anthropogenic VOC (AVOC) emissions from fossil fuel combustion and use (including leakage during production and transportation) have been reported to be approximately 127 Tg/y (Glasius & Goldstein, 2016), thereby confirming high BVOC emissions.It is estimated that isoprene accounts for approximately half of BVOC emissions, with monoterpenes and sesquiterpenes accounting for 15 and 3%, respectively (Guenther et al., 2012).Many BVOC components are highly reactive and undergo oxidation reactions with active species (such as hydroxyl radicals) in the atmosphere (Atkinson & Arey, 2003), thereby contributing to the formation of photochemical air pollutants such as ozone and secondary organic aerosols (SOA).It is estimated that BVOCs account for approximately 70% of the annual VOC emissions (3370 Mg/y) in Japan (Morikawa, 2017), and there is a need to understand the actual situation in the atmospheric environment owing to the large amount of emissions.
In urban areas with limited vegetation, one might consider that BVOCs have little or no impact.Chatani et al. (2022) conducted a simulation analysis of VOC individual component concentrations in the Tokyo metropolitan area based on observation and emission inventory data.They reported that, although there is little vegetation in the Tokyo metropolitan area, it is strongly affected by BVOCs emitted from the neighboring Kanto region.They also pointed out that there are major issues regarding the amount and seasonal variability of BVOC emissions.Furthermore, it has been reported that approximately 14 ppbv of the photochemical ozone generated during summer in the Tokyo metropolitan area is due to BVOCs (Chatani et al., 2015).Bao et al. (2010) conducted a simulation analysis of the Kansai region which includes Osaka, the second-largest city after Tokyo, and reported that ozone formation in urban areas is influenced by BVOCs generated in the suburbs.Nishimura et al. (2015) stated from model calculations that BVOCs contribute 10.3 ppb (15.9%) of the daily mean maximum ozone concentration in Osaka, a conclusion similar to that reached by Bao et al. (2010).In addition, the photochemical ozone formation regime (i.e., VOC-limited or NOx-limited) in the Kanto region varies significantly from the results of BVOC emission estimates (Inoue et al., 2010).Although the proportion of AVOCs was higher in urban areas, the results confirmed that the impact of BVOCs was not negligible.
Numerous studies have been conducted on BVOC emissions and fluxes from individual vegetation species (for example, Bao et al., 2008aBao et al., , 2008bBao et al., , 2010;;Ida et al., 2016;Matsunaga et al., 2013;Mochizuki et al., 2011Mochizuki et al., , 2020;;Okumura et al., 2008a).Although there have been several studies of cases related to the actual status of BVOCs in the atmospheric environment outside Japan (Hakola et al., 2009;Jones et al., 2011;Hakola et al., 2012;Hellén et al., 2018), the recent study in Japan was conducted by Suzuki et al. (2012) in the Tokyo metropolitan area from May to June 2011.This study did not consider comprehensive measurement data throughout the year.Previous studies have shown that the emission of BVOCs substantially varies with season (Lim et al., 2008;Matsunaga et al., 2012;Mochizuki et al., 2011;Okumura et al., 2008a;Son et al., 2015).Furthermore, BVOC emission is affected by environmental factors (temperature, solar radiation, and precipitation); therefore, it is important to conduct observations throughout the year in order to clarify the actual conditions of BVOCs and fluctuations in BVOC concentrations in the environment in response to changes in environmental factors (Lun et al., 2020;Okumura, 2021;Tani & Mochizuki, 2021).
In this study, we aim to accumulate basic knowledge on the actual status of atmospheric BVOCs in the Saitama Prefecture, which is adjacent to the Tokyo Metropolitan area.Considering that most BVOCs have high photochemical activities and short atmospheric lifetimes, timeresolved measurements were conducted to determine their actual status as precisely as possible.This is the first comprehensive, year-long study of BVOCs in the ambient atmosphere using high-time-resolved measurements in Japan.Saitama Prefecture has a climate, with a strong northwest monsoon in winter, dry air with many sunny days, and high temperature and humidity in summer.It is expected that the observation of BVOC concentrations at a location where climatic conditions are different from those of other countries in the world will provide novel insights.

Sampling site
The target sites for this study were the urban CESS air quality monitoring station (36°5′ N, 139°33′ E, 14.4 m a.s.l.) and forested mountainous Higashi-Chichibu (36°0′ N, 139°11′ E, 837.2 m a.s.l.) air quality monitoring station.Both are located in the Saitama Prefecture adjacent to the Tokyo metropolitan area.These two study sites were selected to determine the actual status of atmospheric BVOC concentrations in areas with different characteristics: urban and mountainous areas.The locations of the stations and an aerial photograph of the surrounding environment are shown in Fig. 1.
The CESS station is located at the Center for Environmental Science in Saitama and has many residential areas and rice paddy fields in its vicinity.Approximately 350 m northwest of the station is Saitama Prefectural Road No. 38 (Kazo Konosu Line, 24 h weekday traffic volume: 12,466 vehicles), and approximately 420 m northeast is Saitama Prefectural Road No. 313 (Kitane Shobu Line, 24 h weekday traffic volume: 3820 vehicles), which is a representative monitoring station for the general urban air environment.
The Higashi-Chichibu station is a monitoring station located in the forested area of Mt.Dodaira on the municipal border between Higashi-Chichibu Village and Tokigawa Town.The station was selected as a sampling Fig. 1 Location and characteristics of the sampling sites site in the forest area because there are no anthropogenic emission sources in the vicinity and there is only a small amount of traffic from the cars of mountain climbers and recreationists.To understand the status of tree species around the Higashi-Chichibu station, we obtained statistical data on the area by tree species in Higashi-Chichibu Village and Tokigawa Town.These data were based on a survey conducted in 2016 by the Saitama Prefecture Department of the Agriculture and Forestry Division (data not publicly available).Based on these data, the area by species was 4699 ha for conifers, 1848 ha for broadleaf trees, 42 ha for bamboo forests, and 18 ha for treeless land, with conifers and broadleaf trees accounting for 71% and 28% of the total area, respectively, thereby indicating that conifers were the majority.Regarding the coniferous trees, 61% were Cryptomeria japonica, 34% Chamaecyparis obtusa, and 4% Pinus densiflora.For broadleaf trees, Quercus acutissima accounted for 15% and other unspecified species accounted for 85%.
Table 1 summarizes the annual and seasonal ozone concentrations, wind speed and direction, temperature, and precipitation at the CESS and Higashi-Chichibu stations during the study period.Ozone concentrations, wind speeds and directions, and temperatures were measured using the instruments described in Sect.2.4.4.The Higashi-Chichibu station is at a higher elevation than the CESS station and a mean temperature approximately 5 ºC lower.There is no significant difference in the amount of precipitation between the two stations.
The CESS station is dominated by east-to-southerly winds from April to September and northwesterly winds from October to March.The Higashi-Chichibu station tends to be dominated by northwest or southeast winds throughout the year.

Sampling period
The sampling period was from December 2021 to November 2022.Sampling for each month took place over a period of 2-4 days and was conducted on nonrainy days after checking the weather forecast issued by the Japan Meteorological Agency (JMA).It should be noted that although this study presents monthly data as a valuable case study of atmospheric BVOCs, they are not necessarily representative of each month.The details of the sampling dates and weather conditions are presented in Table S1.
Based on the definition used by the JMA, the four seasons were assigned as follows: March to May for spring, June to August for summer, September to November for autumn, and December to February for winter.

BVOC components to be observed
Based on a review of previous studies on BVOC emission (Bao et al., 2008a;Chatani et al., 2015;Matsunaga et al., 2011;Miyama et al., 2018Miyama et al., , 2020;;Mochizuki et al., 2011Mochizuki et al., , 2020;;Okumura, 2021;Okumura et al., 2013;Tani, 2010;Tani & Mochizuki, 2021;Tani et al., 2002), 11 components were selected for investigation, including BVOCs that have high photochemical activity and have never been observed in Saitama Prefecture.Table 2 displays the component names, their corresponding CAS number, the selling agency from which standard reagents were purchased, purities of the standard reagents used, and the reaction rate constant values (k OH ) with the OH radical at room temperature (298 K) for the BVOCs analyzed (Atkinson & Arey, 2003;Chameides et al., 1992).k OH is used in the analysis in Sect.4.4.
Table 1 Ozone concentration, temperature, wind speed, wind direction, and precipitation statistics at the CESS and Higashi-Chichibu stations during each season and throughout the study period Ichikawa et al. Asian Journal of Atmospheric Environment (2023) 17:10

Sampling and measurement devices 2.4.1 In-house built sampling device
A sampling device built in-house, shown in Fig. 2, was used to sample BVOCs with a high time resolution.The atmospheric sample flowing into the inlet was divided into four channels using a PTFE branch pipe, with an adsorption tube connected to each channel and a 2-way solenoid valve connected to the downstream side.The opening and closing of the solenoid valve were controlled using a 4-channel timer, allowing samples to be collected at different times of the day.The flow rate was 70-100 mL/min, which was adjusted based on the predicted periods of relatively high and low emissions owing to high and low temperatures.In addition, all the tubes were made of PTFE.Two of these sampling devices were installed at each site to measure the diurnal variation in BVOCs in the atmosphere with a 3-h resolution; 0:00-3:00, 3:00-6:00, 6:00-9:00, 9:00-12:00, 12:00-15:00, 15:00-18:00, 18:00-21:00, and 21:00-24:00 in Japan.

Adsorption tube
Canisters generally used for sampling atmospheric VOCs have a limited number of samples because of the physical factors of container size and weight, as well as the time required for pretreatment before and after collection.
Highly time-resolved measurements with short sampling intervals are desirable for understanding the actual status of BVOCs, which fluctuate over short periods of time;  however, the canister sampling method is not suitable for short-time sampling.Therefore, in this study, compact, lightweight, and easy-to-handle adsorption tubes were connected to the sampler built in-house shown in Fig. 2 to perform highly time-resolved measurements.
Inert-coated stainless-steel adsorption tubes Air Toxics (Camsco, USA) with an outer diameter of 1/4″ and length of 3.5″ were used to collect the BVOCs.According to Camsco's product catalog, carbographs and carbosieves are used as fillers in Air Toxics, and the VOC collection range was C2/C3-C12/C14.
Conditioning of adsorption tubes was performed with a temperature program of 40 ºC (hold 1 min) → 10 ºC/ min → 350 ºC (120 min) in a thermostatic chamber modified from gas chromatography oven while high-purity nitrogen was streaming at a flow rate of 50 mL/min.Immediately after conditioning, the adsorption tubes were tightly closed at both ends with polytetrafluoroethylene (PTFE) analytical caps, placed in stainless-steel containers with activated carbon, and stored in a desiccator under vacuum until sampling.
After sample collection, the adsorption tubes were tightly capped at both ends with PTFE analytical caps, sealed in a desiccator with activated carbon to prevent contamination in an air-conditioned, controlled room, and stored until further analysis.The analysis was performed within a 2-3 weeks of sampling.
A travel blank was used to confirm the effects of contamination during sample transport and storage.The travel blank was analyzed using the same method as that for the actual sample and its value was subtracted from the analysis value of the actual sample.

Thermal desorption-gas chromatography/mass spectrometry (TD-GC/MS) system
A fixed amount of standard solution of each BVOC components dissolved in methanol (LCMS grade, FUJI-FILM Wako Pure Chemical Co., Japan) was added to a clean inert stainless vacuum canister (GL-Scan, GL Sciences Inc., Japan) and heated at 60 °C for 2 h to completely vaporize the components.After returning to room temperature, the vacuum bottle was pressurized with VOC-free high-purity air (ZERO-N, Sumitomo Seika Chemicals Co., Ltd., Japan) to create a standard gas.Toluene-d 8 (NMR grade, 99.5% purity, FUJIFILM Wako Pure Chemical Co., Japan) was used as the internal standard gas and prepared in the same manner as described above.
BVOCs were analyzed using a gas chromatographymass spectrometer (GC/MS; GC/MS-QP2010plus, Shimadzu Co., Japan) connected to an automated thermal desorption system (TurboMatrix 650 ATD, PerkinElmer Japan Co., Ltd., Japan).These devices are used for annual manufacturer inspections to ensure that they are in good condition and free from defects.A certain amount of internal standard gas was spiked into the adsorption tube prior to analysis.
Fragment patterns obtained in scan mode for a single standard reagent for each BVOC component were matched with the NIST library before checking the retention times and selecting the target and qualitative ions.The collected samples were analyzed in SIM mode.The analytical conditions are listed in Table 3, and the chromatograms are shown in Fig. S1.Verification of the chromatograms showed that the waveform processing and quantification performed well, with no interference from foreign substances.Calibration curves for each component were prepared using the internal standard method and good linearity was obtained.
The limit of lower detection (3σ) was calculated from the standard deviation (σ) obtained from repeated measurements of the standard gas (n = 5-7).Measurements below the limit of detection were calculated as half of the limit of detection.The target and qualitative ions selected for the GC/MS analysis of each component and their limit of detection are shown in Table 4.

Measurement of other environmental factors
Other environmental factors such as oxidant concentration, temperature, wind speed, wind direction, solar radiation, and radiation budget were also measured at the sampling sites.The measurement principles, manufacturers, and models of the instruments are listed in Table 5.However, the solar radiation and radiation budget were measured only at the CESS station.
Atmospheric stability is an indicator of vertical mixing and diffusion of the atmosphere.The meteorological instruments installed at CESS station are designed to determine the degree of atmospheric stability based on the Pasquill stability classification from wind speed, solar radiation, and radiation budget.Atmospheric stability is classified into A-G categories, with "extremely unstable" (A), "unstable" (B), "slightly unstable" (C), "neutral stability" (D), "slightly stable" (E), "stable" (F), and "extremely stable" (G).A stable condition means that the air temperature lapse rate is less than the dry adiabatic lapse rate, and that the air is hot (cooler above the ground), which suppresses convection and reduces the diffusion of pollutants.Neutral is the condition in which the air temperature reduction rate from the ground to the sky is approximately equal to the dry adiabatic reduction rate.Unstable is a condition in which the air temperature lapse rate is greater than the dry adiabatic lapse rate and the air temperature above the ground is cooler (bitter above the ground), which tends to create convection and leads to vertical convection.Because atmospheric stability is only determined at CESS station in the Saitama Prefecture, the values from this station were used as representatives.
The equipment installed at the air quality monitoring stations was checked for abnormalities and simple inspections were performed approximately once every 10 days by a contractor with whom the Saitama Prefectural Environmental Department Air Environment Division had concluded an inspection and maintenance contract.Calibration of the ozone measuring instruments was performed at least once per year.In addition, recorder output calibration of the thermometers, anemometers, and solar radiation meters was performed at least four times a year.

Significant difference test for seasonal mean concentrations
Assuming that the atmospheric observation data for BVOCs do not follow a Gaussian distribution, a nonparametric method, Steel-Dwass multiple comparison test, was used to check for significant differences in seasonal mean concentrations.

Calculation of propylene equivalent concentration
The propylene equivalent (PE) concentration is a method developed by Chameides et al. (1992) for accessing ozone productivity using reactive VOC species i (VOC i ) in the atmosphere.It is calculated using the following equation: where

Seasonal concentrations of BVOCs
Figure 3 shows the seasonal mean concentrations (ppbv) of BVOCs observed at the CESS and Higashi-Chichibu stations and the composition ratios of each component relative to the values in a pie chart.The order of Table 3 Analytical conditions for thermal desorption-GC/MS the seasonal mean concentrations is summer (0.16 ppbv) > winter (0.15 ppbv) > autumn (0.14 ppbv) > spring (0.13 ppbv) at the CESS station and summer (0.39 ppbv) > spring (0.17 ppbv) > autumn (0.12 ppbv) > winter (0.11 ppbv) at the Higashi-Chichibu station, with the highest concentrations in summer at both stations.At the CESS station, summer was significantly different from the other seasons (p < 0.05) but there were Table 4 Target and qualitative ions, and limit of detection for each component

Table 5 Instruments observing environmental factors at the CESS and Higashi-Chichibu stations
*There are no pyranometer and radiation dosimeter installed at the Highashi-Chichibu station.no significant differences between winter, spring, and autumn.At the Higashi-Chichibu station, there was a significant difference between fall and winter with a p value < 0.05 and for comparison between other seasons with a p value < 0.01.

Monthly concentrations of BVOCs
The observed mean BVOC concentration (ppbv), mean temperature (°C), and mean ozone concentration (ppbv) at 3-h intervals and daily for each month from December 2021 to November 2022 are shown in Figs.S2 and  S3.There were no characteristic differences in the mean concentrations of BVOCs at any time of the month during the period of lower temperatures (September-April), nor were there any significant differences between the CESS and Higashi-Chichibu stations.The monthly atmospheric BVOC concentrations showed an increasing trend from May to August, especially at the Higashi-Chichibu station; therefore, in Fig. 4, the data of these months are shown.In addition, Fig. 5 was prepared to determine the variation in the 3-h interval composition rates (%) and daily mean composition rates (%) for these months.

Calculated propylene equivalent concentration
Figure 6 illustrates the calculated [PE i ] values based on BVOC concentrations observed at the CESS and Higashi-Chichibu stations between May and August.In addition, Fig. 7 illustrates the variation in the 3-h interval composition rates (%) and daily mean composition rates (%) of each BVOC component relative to the total PE concentration for these months.

Seasonal variations and characteristics of BVOCs by region
In this section, we review and compare seasonal variabilities to understand the general characteristics of the respective BVOC components observed at the CESS and Higashi-Chichibu stations.Basic studies and models related to BVOC emissions from leaves have presented the relationship between leaf surface temperature and BVOC emissions (Bao et al., 2008b;Guenther et al., 1993Guenther et al., , 2012;;Ida et al., 2016;Lim et al., 2008;Matsunaga et al., 2012Matsunaga et al., , 2013;;Mochizuki et al., 2011Mochizuki et al., , 2020;;Okumura et al., 2008a;Son et al., 2015).Previous studies confirmed that higher leaf temperatures increase the enzyme activity and transcription levels necessary to promote terpenoid synthesis (Yang et al., 2021), which is related to the emissions of BVOCs.However, in this study, only the air temperature at the air quality monitoring station was observed; leaf surface temperatures are expected to be dependent on the air temperature.This means that temperature can have a major effect on BVOC emission.Furthermore, a study in Finland showed that the atmospheric concentration of BVOCs in the boreal forest area increased around May-June and peaked from June-August, which is similar to our results (Hakola et al., 2009).This is due to increased emissions of BVOCs from vegetation during the summer months.Thus, higher temperatures may be one of the reasons why the seasonal mean concentrations tended to be higher in summer.
Although the seasonal mean temperatures at the Higashi-Chichibu station was about 5 °C lower than that of the CESS station, Higashi-Chichibu station is in a forested area and thus had higher BVOC concentrations Focusing on summer observations, isoprene accounted for the highest percentage of total BVOC concentrations: 31.0%at the CESS station and 45.6% at the Higashi-Chichibu station.The formation and emission of BVOCs depend on biotic and environmental factors, although the characteristics differ among BVOC components.Biotic factors include tree species, tree age, physiology, and animal feeding, whereas environmental factors include leaf surface temperature, solar radiation, soil moisture, carbon dioxide concentration, ozone concentration, and soil nitrogen content (Miyama et al., 2018;Mochizuki et al., 2017;Okumura, 2021;Tani & Mochizuki, 2021;Yang et al., 2021).Particularly, isoprene is thought to be a component released to protect leaves from heat and oxidative stress (Behnke et al., 2007;Sharkey et al., 2008).Therefore, it is assumed that the isoprene concentration was relatively high during this period.Note that isoprene is used as a raw material in the man-made manufacture of synthetic rubber (for example, polyisoprene rubber).However, there are no industrial sources of isoprene in the vicinity of the study site, so it is unlikely that the area has been affected by isoprene originating from anthropogenic sources.
The CESS station had a slightly higher seasonal mean concentration of BVOCs in summer than in the other seasons (p < 0.05).Typical vegetation in the vicinity includes rice paddies (Oryza sativa).It has been reported that Oryza sativa emits less BVOCs than trees, but emits small amounts of α-pinene, β-pinene, myrcene, p-cymene, and limonene (Bao et al., 2008b).However, the results of this study did not show a trend of significantly higher Regarding monoterpenes, limonene, p-cymene, and α-pinene showed a characteristic trend in summer at the Higashi-Chichibu station.Many conifers release monoterpenes and sesquiterpenes, and their release has been found to be dependent on leaf surface temperature (Bao et al., 2008b;Matsunaga et al., 2011;Mochizuki et al., 2011;Okumura et al., 2008aOkumura et al., , 2013;;Son et al., 2015).In addition, some conifers exhibit light-dependent emission of BVOCs (Bao et al., 2008b).As explained in Sect.2.1, the major coniferous tree species in the vicinity of Higashi-Chichibu station are C. japonica, C. obtusa, and P. densiflora.Based on previous reports, the monoterpenes emitted by these tree species are summarized below.
In the spring season at the Higashi-Chichibu station, the composition of each component in the concentration of BVOCs was different from the summer season: 12.5% for isoprene, 14.0% for p-cymene, 13.8% for β-pinene, 9.1% for α-pinene, and 8.8% for α-terpinene.Thus, it is interesting to note that the BVOC composition of the atmosphere was found to vary with the season.Many components were below the detection limit during periods of low seasonal BVOCs concentrations.It is noteworthy that the sensitivity to β-pinene under the present analytical conditions of TD-GC/MS was not as good as that of other components, and the percentage of β-pinene appeared to be high even if it was below the detection limit.

Monthly variations and characteristics of BVOCs by region
The general trend from May to August was that the compositional percentage of isoprene concentration rate is higher during the day when the temperature is higher.
Several studies have shown that isoprene is a temperature-and light-dependent component (Bao et al., 2008b;Guenther et al., 1993Guenther et al., , 2012;;Mochizuki et al., 2020;Okumura et al., 2008b); hence, its concentration is relatively low during the night and early morning and high during the day.Regarding the total BVOC concentrations at the Higashi-Chichibu station, a higher trend was often observed from 6:00-12:00 than from 12:00-18:00.One possible reason for this trend is that the ozone concentration increased from 12:00-18:00, resulting in photochemical reactions (oxidation reactions) of BVOCs, which led to concentrations lower than those in the morning.Another possible effect is that atmospheric stability tends to be unstable during the day (atmospheric stability and mixed layer altitude are highly correlated).Hakola et al. (2012) observed higher concentrations of atmospheric monoterpenes in forested areas during the night and early morning than during the day.They estimated that altitude difference is one of the factors contributing to the variation in the concentrations during the day based on mixed layer altitudes at different times of Apart from these general trends, the characteristics of each month at the Higashi-Chichibu station have been described here.An interesting feature is that the atmospheric concentrations of monoterpenes were higher in May and July than in June and August.Among the monoterpenes, high composition ratios of limonene, α-pinene, and p-cymene were often observed.As for these components, they are BVOC components emitted from the main tree species in the vicinity of the Higashi-Chichibu station, as described in Sect.4.1.
On the contrary, the concentrations of monoterpenes were relatively low in June and August, with isoprene being the major component.Approximately 71% of the surrounding trees are conifers; C. japonica, C. obtusa, and P. densiflora are the main tree species in this area.Moreover, they are reported to be lower isoprene emitters than monoterpenes (Bao et al., 2008a).However, in a study conducted in Japan, Miyama et al. (2020) reported high isoprene emission from these tree species.In addition, Ida et al. (2016) analyzed BVOCs released using the enclosure method with C. japonica in a plant chamber with controlled temperature and light levels and found that the concentrations of BVOCs increased with increasing temperature, with isoprene emission 9.2 times greater than α-pinene emission at 35 °C.These studies also inferred that high concentrations of isoprene were observed at the Higashi-Chichibu station.
Thus, interesting differences in atmospheric BVOC concentrations and composition rates were observed during different months.The large differences in the concentration and composition of atmospheric BVOC components among the months indicate that the emission behavior of vegetation varies temporally.Some studies have reported that the BVOC emission patterns of trees differ depending on the time of year and location (Okumura et al., 2013), and this pattern may be reflected at the study site.Perhaps, a monthly inventory of BVOCs, which is required as input data for the chemical transport model, might also be needed for accurate calculations.This interesting result could be a focus for future studies.Further research is required to elucidate these mechanisms in detail.
The highly time-resolved measurement of atmospheric BVOC observations allowed us to examine the situation regarding monthly BVOC concentration variation and composition in Japan for the first time, to the best of our knowledge.

Relationship between BVOC concentration and temperature
Figure 8 shows a scatter plot of the relationship between isoprene concentrations and temperature (n = 248) and Σmonoterpenes and temperature (n = 248) observed at the CESS and Higashi-Chichibu stations.Note that Σmonoterpenes is the total concentration of monoterpenes measured in this study."The Model of Emissions of Gases and Aerosols from Nature (MEGAN)" (Guenther et al., 2012) is a model used to estimate BVOC emissions from vegetation leaves.In this model, isoprene emission depends on leaf surface temperature (correlates with air temperature) and photosynthetic photon flux density (PPFD; number of photons per unit time and the unit area contained in the wavelength of 400-700 nm required for photosynthesis), indicating that isoprene concentration increases exponentially with increasing leaf surface temperature up to around 40 °C, but decreases thereafter.
No relationship with temperature could be confirmed for isoprene at the CESS station, perhaps because there were few clear sources of isoprene in the vicinity.On the other hand, at the Higashi-Chichibu station, which has many BVOC emission sources, isoprene concentrations increased when temperatures exceeded 10 °C and increased exponentially with further increases in temperature.This is consistent with reports that isoprene is released to protect vegetation from heat stress (Behnke et al., 2007;Sharkey et al., 2008).Although the Higashi-Chichibu station does not experience high temperatures due to its high elevation, the atmospheric concentration of BVOCs may be even higher in areas with many trees where temperatures are higher, and further observational studies are needed in such areas.In a previous study on atmospheric observations; Lee et al. (2006) reported an exponential increase in isoprene concentrations with increasing summer (August) temperatures in suburban London, which is consistent with the results of this study.
In the MEGAN model, monoterpene emissions were also designed to increase exponentially with increasing leaf temperatures (Guenther et al., 2012).The relationship between Σmonoterpenes and temperature could not be characterized at the CESS station.However, at the Higashi-Chichibu station, a trend of increasing concentration with increasing temperature was observed.There was a slight difference in the scatter diagrams of isoprene and monoterpenes, which may be due to the way they are emitted.Isoprene is emitted immediately after formation, whereas monoterpenes accumulate in storage tissues before being emitted (Okumura, 2021;Tani & Mochizuki, 2021;Yang et al., 2021).

Propylene equivalent concentration
We first focus on the Higashi-Chichibu station, where characteristic results were obtained.Here, the maximum total PE concentrations for each month were as follows: 1.6 ppbC during 9:00-12:00 in May, 3.1 ppbC during 12:00-15:00 in June, 4.5 ppbC during 6:00-9:00 in July, and 1.3 ppbC during 9:00-12:00 in August (Fig. 6).The PE concentrations of isoprene were relatively higher than those of the other components.In June, isoprene emission increased in the morning (6:00-9:00), and its air concentrations remained high until sunset, with isoprene contributing 77.5% of the total PE concentration during 12:00-15:00.It is interesting to note that isoprene appears to exert greater impact during the daytime when solar radiation and temperatures are higher.While monoterpenes exhibit a certain impact even during the early morning and nighttime when there is no solar radiation.This trend was particularly evident in July, where the contribution of limonene to the total PE concentration during 3:00-9:00 ranged (mean values) from 29.1 to 40.4% (34.8%).These results suggest that the conditions contributing to ozone productivity of BVOC components vary with time of day and time of year.
At the CESS station, the contribution from isoprene was relatively larger than that from other monoterpenes in all months (with a maximum isoprene PE concentration of 0.49 ppbC during 9:00-12:00 in August).The PE contribution of isoprene to the total PE concentration was considerable, at 37.9% during 9:00-12:00 in June, 35.4% during 9:00-12:00 in July, and 36.1% during 9:00-12:00 in August.AVOCs were observed at the Kounosu station in the vicinity of the CESS station.The mean PE concentration of the major AVOCs of benzene, toluene, ethylbenzene, p&m-xylene, and o-xylene, at the Kounosu station from May to August 2022 were 0.0034 ppbC, 0.29 ppbC, 0.060 ppbC, 0.13 ppbC, and 0.034 ppbC, respectively.This analysis suggests that BVOCs have a considerable impact on ozone productivity.

Conclusions
In this study, two study sites (CESS and Higashi-Chichibu stations) with different characteristics in the Saitama Prefecture adjacent to the Tokyo metropolitan area were selected for atmospheric observations of BVOCs, which have high photochemical activity and a short atmospheric lifetime.This is the first study in Japan to conduct year-round observations of atmospheric BVOCs with high time resolution.The findings of this study are summarized below: • On many study days, isoprene accounted for a large proportion of the BVOC components during the daytime, when temperatures and solar radiation were higher.• Both stations showed considerably higher mean seasonal BVOC concentrations in summer than in other seasons.• There were substantial differences in the concentrations and compositions of atmospheric BVOCs by the time of day and month, suggesting that there were significant differences in the emission characteristics of BVOCs from vegetation by month.
• Depending on the time of year, isoprene concentrations were prominent in some cases and monoterpenes in others.• At the Higashi-Chichibu station, located in a forested area, an increase in the concentrations of isoprene and monoterpenes in the atmosphere was observed with increasing temperature.Isoprene tends to increase with temperature, as reported in previous studies.• Analysis by the PE concentration suggests that BVOCs have a substantial impact on ozone productivity.
The results of this study suggest that atmospheric BVOCs are present in Japan and contribute significantly to ozone production, especially during periods of high temperature and solar radiation.
Chemical transport model calculations that attempt to reproduce ozone concentrations using emission inventories of BVOCs and other data are very important for the future development of ozone control measures.Although the accuracy of model calculations has improved, uncertainty in model calculations remains an issue because this leads to discrepancies between the measured data and calculated values.As a solution, we believe that enhanced atmospheric observations of BVOCs, which are known to contribute significantly to ozone production, are of great significance.In addition, we hope that high-precision observational data with highly time-resolved measurements, such as the results of this study, will contribute to improving the uncertainty in model calculations.

Fig. 2
Fig. 2 Diagram of the in-house built sampling device

Fig. 3
Fig. 3 Seasonal mean concentrations and composition of BVOCs at the CESS and Higashi-Chichibu stations

Fig. 5
Fig. 5 Mean composition rates (%) of BVOC components by 3-h intervals and by day for each month of May to August 2022 at the CESS and Higashi-Chichibu stations

Fig. 6
Fig. 6 PE concentrations of BVOC components observed at the CESS and Higashi-Chichibu stations from May to August

Fig. 7
Fig. 7 Mean percent composition (%) of BVOC components relative to the total PE concentrations at 3-h intervals and daily for each month from May to August 2022 at the CESS and Higashi-Chichibu stations

Fig. 8
Fig. 8 Relationship between isoprene and Σmonoterpenes and temperature observed at the CESS and Higashi-Chichibu stations

Table 2
CAS numbers, reagent information used, and the reaction rate constant values (k OH ) with the OH radical at the room temperature (298 K) for targeted BVOC components Atkinson and Arey (2003)ent the PE concentrations of VOC i and VOC i in ppbC, respectively; and k OH (VOC i ) and k OH (C 3 H 6 ) indicate the reaction rate constant of VOC i and propylene with OH radical at room temperature (298 K), respectively.The k OH of the BVOCs analyzed in this study are shown in Table2.In addition, the k OH of propylene (26.3 × 10 12 cm 3 molecule −1 s −1 ) was taken fromAtkinson and Arey (2003).