Effective emission reductions of some primary pollutants have brought down aerosol loadings but led to increasing relative importance of secondary pollutants, as was indicated by the rising O3 levels during warm seasons within urban and suburban areas of China, which has received much attention in recent years, especially in the North China Plain (NCP). This has raised serious concerns on its agricultural impacts, which were mainly evaluated based upon O3 model simulations or urban/suburban measurements due to a lack in long-term rural observations. In this study, we present highly valuable continuous O3 observations at a rural NCP site during 2013–2019. Compared to nearby urban/suburban sites, which experienced increased O3 levels, rural observations exhibited decreasing O3 mole fractions. While O3 mole fractions and AOT40 widely increased at urban/suburban NCP sites from 2013 to 2019, O3 observations in the rural NCP site (GC) revealed decreases, especially during summer and autumn with greater rates for AOT40. A reassessment of O3 agricultural impacts in the NCP region was performed using rural observations, resulting in wheat, maize and soybean averaged relative yield losses of 37 ± 14, 8 ± 4 and 30 ± 13% yr–1, respectively. O3 impacts on crop yields and resulting economic losses did not increase as was suggested by previous estimations based on urban/suburban O3 data. Our analyses indicated high overestimations (i.e., average relative differences in estimated crop production loss reaching 53%, 112% and 75%, respectively, for wheat, maize, and soybean). Despite alleviated O3 agricultural impacts, the total economic cost loss in Hebei province still took up 0.89% of the gross domestic production (3.47 × 1012 USD) in Hebei province. Since the China National Environmental Monitoring Center mainly aims at monitoring O3 levels in populated areas, observation sites representative of agricultural regions are lacking across China. The current study highlights the urgent necessity for the establishment of rural O3 observation networks and encourages extensive field experiments on exposure–response relationships of different crops varieties to O3 for more accurate agricultural impact evaluations. Additionally, explorations into the underlying mechanisms behind the reversed O3 temporal variation between rural and urban areas should be conducted for future development of pollution control strategies.

Surface ozone (O3) is an important secondary gaseous air pollutant and has adverse effects on air quality, human health, and ecosystem productivity (Ainsworth, 2017; Cohen et al., 2017; Yue et al., 2017; Xu et al., 2020b). Tropospheric O3 acts as a greenhouse gas with radiative forcing of 0.40 ± 0.20 W m–2 (Intergovernmental Panel on Climate Change, 2021), affecting global climate change. Additionally, O3 can enter leaves through open stomata, react with water in cells, and produce a series of hazardous oxidants, such as H2O2, O2, OH, and HO2, damaging vegetation and threatening crop quality and production (Aunan et al., 2000; Felzer et al., 2005; Biswas et al., 2008; Piikki et al., 2008; Bender and Weigel, 2011; EPA, 2013; Ainsworth, 2017; Harmens et al., 2018). The impact of O3 on crops depends both on ambient O3 levels and crop sensitivities toward O3 exposure. While wheat and soybean were categorized as sensitive crops, rice and maize are only moderately sensitive (Mills et al., 2007). High O3 levels have exerted seriously adverse effects on agriculture around the world, with production losses in developing countries exceeding those in developed countries (Mills et al., 2018b). Global economic losses caused by O3 pollution were estimated to be about \$14–26 billion in the year 2000, among which China and India accounted for a total of 40% (Van Dingenen et al., 2009). China is among the largest crop production countries in the world, with wheat, maize, and soybean productions accounting for 50%, 71%, and 40% of the total Asian yields, respectively (Food and Agriculture Organization of the United Nations, 2022b). However, with the rapid expansion of population and growing demands for food, China has also become the largest crop importer (Dong et al., 2021), making domestic crop yields more crucial than ever for the Chinese economy and food supply.

Meanwhile, significant upward trends of O3 in China have been reported during the past decades (Monks et al., 2015; Li et al., 2019; Wang et al., 2019; Lu et al., 2020; Li et al., 2020a; Xu et al., 2020b; Xu, 2021), especially in the North China Plain (NCP) (Ding et al., 2008; Wang et al., 2012b; Tang et al., 2013; Tai et al., 2014; Zhang et al., 2014; Ma et al., 2016). Since 2013, strengthened emission standards were set up for coal-fired power plants, industries, factories, and vehicles, which successfully brought down ambient concentrations of SO2, CO, and NOx as well as that of particulate concentrations, while Volatile Organic Compounds (VOCs) reductions were not as obvious (Zheng et al., 2018). The nonlinear response of O3 formation to its precursors (NOx and VOCs) might have led to the rapid increases in warm-season O3 under current nonproportional emission reductions (Dang et al., 2021; Lu et al., 2021). Since the NCP produces 35.2% and 69.2% of China’s total maize and winter wheat, respectively (Liu et al., 2010), the increasing O3 levels in this area and its effect on crop yield and domestic crop supply have raised serious concerns. Increasing numbers of studies have paid attention to the agricultural impact of O3 pollution and evaluated the current and future O3 induced crop yield losses using Expose–Response (E-R) functions obtained from field experiments (Aunan et al., 2000; Avnery et al., 2011a; 2011b; Tao et al., 2012; Wang et al., 2012a; Tang et al., 2013; Chuwah et al., 2015; Carter et al., 2017; Li et al., 2018; Feng et al., 2019a; Feng et al., 2019b; Feng et al., 2020; Hu et al., 2020; Dong et al., 2021). However, due to the lack in O3 observations at agricultural sites, most of those studies were either based upon modeled O3 concentrations (Avnery et al., 2011a; 2011b; Tang et al., 2013; Mills et al., 2018b; Wang et al., 2021) or upon surface O3 observations from the China National Environmental Monitoring Center (CNEMC) network (Zhao et al., 2018; Feng et al., 2019a; Feng et al., 2020; Ren et al., 2020). Modeling results of O3 bear certain uncertainties (caused by limited spatial resolutions, incomplete or inaccurate emission inventories, and chemical mechanisms, etc.) (Dong et al., 2021), which can be transmitted onto estimated agricultural impacts. Observation sites of the CNEMC are concentrated in urban and suburban areas (county centers) (Li et al., 2018; Feng et al., 2019b), which are heavily influenced by anthropogenic activities and represent land coverage types highly distinct from those in rural areas, and may thus bring great uncertainties to the evaluation of O3 induced agricultural impacts. Some studies used reanalysis datasets obtained through assimilation of CNEMC data to compensate for the lack in rural O3 observations (Dong et al., 2021); however, it has not been validated whether reanalysis data can better represent rural O3 pollution conditions than observations of nearby suburban/urban cities.

Therefore, to assess agricultural impacts of O3 in the NCP more effectively, O3 observations from representative rural regions are urgently required. In this study, we present 7 years of O3 observations performed at a rural site in the NCP, compare them with data from nearby urban sites of the CNEMC network, reveal differences between urban and rural O3 levels and temporal variations, and discuss discrepancies in estimated agricultural impacts on 3 major crops (winter wheat, maize, and soybean). We also evaluate the uncertainties that have resulted from using nonrural measurements as substitute for unavailable rural observations in the assessment of ozone’s agricultural impacts and emphasize the pressing need for establishing networked rural ozone measurements in the future.

### 2.1. Ozone observation

Continuous surface O3 measurements have been conducted at a rural site located in the Gucheng Town, Dingxing County, Hebei Province, China. The Gucheng site (GC, 39.15°N, 115.73°E, 15 m asl) is an ecological and agricultural meteorology station of the Chinese Academy of Meteorological Sciences and is situated between Beijing (approximately 100 km) and Baoding (BD) (approximately 40 km), the 2 key cities in the Beijing–Tianjin–Hebei agglomeration (Figure 1). Being surrounded by farmlands, which are mostly on a summer corn-winter wheat rotation, this site can well represent the polluted agricultural areas of the NCP (Lin et al., 2009; Xu et al., 2019; Kuang et al., 2020a; Kuang et al., 2020b; Xu et al., 2020a). O3 measurements from 2013 to 2019 were conducted using a commercial instrument from Thermo Environmental Instruments (TE 49C), which was housed in an air-conditioned room. The inlet tube (Teflon, 4.8 mm ID × 8 m length) was 1.8 m above the roof and approximately 8 m above ground level. Multipoint calibrations were made every 1–3 months using an O3 calibrator (TE 49C PS), and the accordingly corrected O3 data from 2013 to 2019 were averaged into hourly mole fractions, which were then processed into monthly, seasonal, and annual averages (each step with a minimum 75% data coverage requirement). Seasonal averages in spring, summer, autumn, and winter correspond to averages over MAM, JJA, SON, and DJF, respectively. Based on annually and seasonally averaged O3 levels, linear regressions were calculated to yield the respective interannual change rates.

Figure 1.

Distribution of population density (obtained from https://hub.worldpop.org/geodata/) and the location of Gucheng (GC, red star), Baoding (BD, orange), Langfang (LF, green), Beijing urban (BJurb, purple) and suburban (BJsub, pink), Shijiazhuang urban (SJZurb, turquoise), suburban (SJZsub, light blue), and rural (SJZrur, light blue star) sites.

Figure 1.

Distribution of population density (obtained from https://hub.worldpop.org/geodata/) and the location of Gucheng (GC, red star), Baoding (BD, orange), Langfang (LF, green), Beijing urban (BJurb, purple) and suburban (BJsub, pink), Shijiazhuang urban (SJZurb, turquoise), suburban (SJZsub, light blue), and rural (SJZrur, light blue star) sites.

Close modal

Surface O3 measurements from surrounding CNEMC network sites in the NCP were obtained from the CNEMC for the time period of 2013–2019, which can be downloaded from the Air Quality Historical Data Platform (2020, see reference). After filtering out sites with discontinuous measurements, 151 sites remained, whose seasonal and annual average O3 mole fractions and variation rates were also calculated with a minimum 75% data coverage requirement. In the end, 128, 73, 73, 124, and 129 sites remained with effective change rates of annual, spring, summer, autumn, and winter average O3 mole fractions, respectively. For further comparison with measurements at GC, the CNEMC sites of the 3 cities surrounding GC were selected and grouped into Beijing Urban, Beijing Suburban (BJsub), Langfang (LF), and BD (Figure 1, Table S4). Shijiazhuang (SJZ) city is located southwest of BD, with a rural (mountain) site (SJZrur, 37.91°N, 114.35°E, approximately 250 m asl) located southwest to its urban area (SJZurb) and another suburban site (SJZsub) further away northeast to SJZurb. Sites in SJZ were also compared to each other to reveal that rural/urban discrepancies in O3 levels and agricultural impacts were a common phenomenon.

Monthly tropospheric O3 vertical columns were obtained from Aura OMI/MLS (Ozone Monitoring Instrument/Microwave Limb Sounder) tropospheric ozone productions (grid resolution of 1° × 1°), determined by subtracting colocated MLS stratospheric column ozone from OMI total column ozone (Ziemke et al., 2006). The comparison between the tropospheric O3 column over GC and observed surface O3 mole fractions are shown in Figure 2, as well as surface O3 mole fractions observed in BD. Good correlations were found between tropospheric O3 column and surface O3 levels at GC (r2 = 0.68). Surface O3 observations captured the seasonal and interannual variations of tropospheric O3 well, proving its regional representativeness. Additionally, warm-season surface O3 in BD strongly deviated from that at GC in 2018–2019, while tropospheric O3 over GC remained well correlated to surface observations, suggesting the urban and rural discrepancy in O3 levels and temporal variations to be a regional phenomenon.

Figure 2.

(a) Monthly tropospheric and surface O3 levels at GC and BD during 2013–2019 and (b) correlations between tropospheric O3 columns over GC and surface O3 levels in BD with surface O3 levels in GC. Tropospheric O3 column data are from https://acdext.gsfc.nasa.gov/Data_services/cloud_slice/new_data.html. Surface O3 levels at BD are the monthly averages of the O3 mixing ratio at BD sites in the CNEMC network.

Figure 2.

(a) Monthly tropospheric and surface O3 levels at GC and BD during 2013–2019 and (b) correlations between tropospheric O3 columns over GC and surface O3 levels in BD with surface O3 levels in GC. Tropospheric O3 column data are from https://acdext.gsfc.nasa.gov/Data_services/cloud_slice/new_data.html. Surface O3 levels at BD are the monthly averages of the O3 mixing ratio at BD sites in the CNEMC network.

Close modal

### 2.2. Ozone metric

Since O3 impacts on plants are cumulative, cumulative O3 metrics have been developed for assessing such vegetation impacts (Lefohn et al., 2018), among which AOT40 and W126 were most widely used (Mills et al., 2018a). AOT40 is the European standard metric for the protection of vegetation (Lefohn et al., 2018), which is defined as the sum of all hourly O3 at the top of the canopy that exceed 40 ppb during daylight hours (8:00–19:59 local time) within the crop growing season, and is calculated as:

1

Here, [O3]i is the daytime hourly mean O3 mole fraction, and n is the total number of the hourly O3 samples that exceeded the 40 ppb level during the growing season.

As the impact of O3 exposure varies among distinct crop growth stages (Tang et al., 2013; Carter et al., 2017), the growing season of winter wheat was defined as March 15 to June 15, the 3 months prior to the harvest period according to the crop calendar data from Van Dingenen et al. (2009), while the impacts on soybean and maize are calculated for the entire growing season, from June 16 to September 30. The monthly, seasonal, and annual AOT40 were also calculated to compare changes in AOT40 with changes in monthly, seasonally, and annually averaged O3 mole fractions.

### 2.3. Crop relative yield (RY) and economic loss estimation

E-R functions are typically used to estimate the influence of O3 exposure on crops, which were established through field experiments, such as open top chamber experiments (Feng et al., 2003; Mills et al., 2007; Zhu et al., 2011; Wang et al., 2012a; Peng et al., 2019). In this study, the following set of equations was used to estimate the RY of wheat, maize, and soybean, respectively:

2

winter wheat (Wang et al., 2012a).

3

maize (Mills et al., 2007).

4

soybean (Mills et al., 2007).

Based on the estimated RY, the crop relative yield loss (RYL), crop production loss (CPL), and economic cost loss (ECL) can be calculated as (Avnery et al., 2011a):

$RYL=1−RY.$
5
$CPL=CP×RYLRY.$
6
$ECL=CPL×MPP.$
7

where CP is the amount of crop production, which were obtained from the Statistical Yearbook of the National Bureau of Statistics of China (2019), MPP is the annual purchase price of the crops, which were provided by the Food and Agriculture Organization (FAO) of the United Nations Statistical Database (FAO of the United Nations, 2022a).

### 3.1. Distinct ozone levels and temporal variations between urban and rural NCP

The distribution of annually and seasonally averaged O3 mole fraction and AOT40 (calculated as annual or seasonal cumulative O3 exceedance over 40 ppb) variation rates from 2013 to 2019 in the NCP region are displayed in Figures 3 and S1, respectively, including selected urban/suburban sites of the CNEMC network and the rural GC site. O3 variations observed at the GC station (denoted by triangles in Figure 3) were evidently distinct from those of surrounding urban/suburban sites. A declining rate of -1.0 ± 0.8 ppb yr–1 (P = 0.28) in annual average O3 was found at GC for the period of 2013–2018 (annual average O3 in 2019 was excluded due to insufficient sample numbers after September 2019). At only 5 out of 128 CNEMC sites in the NCP, which were either urban park sites or suburban/rural village sites, decreasing O3 temporal variations were observed during 2013–2019 (ranging between –0.8 and –0.04 ppb yr–1), but none of them were significant (P > 0.05). In contrast, increasing rates were observed at 123 CNEMC sites, reaching on average 2.4 ± 1.1 ppb yr–1 (ranging from 0.5 to 5.6 ppb yr–1, P < 0.05), among which 65 sites displayed increases significant at the 95% confidence level (P < 0.05), with the largest O3 increasing rates found in southern Hebei and northern Shandong Province (Figure 3a). Annual AOT40 revealed trends distinct from those of annual average O3 (Figure S1), displaying overall significant increases in Beijing–Tianjin–Hebei region and in Jinan, Shandong (4.5 ± 2.0 ppm h yr–1 on average, P < 0.05), whereas variation rates in surrounding regions were mostly negative and insignificant (–0.2 ± 0.1 ppm h yr–1 on average, P > 0.05). GC revealed large negative slopes in annual AOT40 (–3.78 ppm h yr–1, P = 0.30), in comparison to the strong increasing AOT40 of its surrounding urban sites.

Figure 3.

Distribution of annual (a) and seasonal (b–e) O3 change rates during 2013–2019 in the NCP. Data are from the CNEMC network and GC; dots represent sites with rates at the significance level of P < 0.05, while crosses represent sites with rates that revealed P values greater than 0.05, and the triangles represent GC.

Figure 3.

Distribution of annual (a) and seasonal (b–e) O3 change rates during 2013–2019 in the NCP. Data are from the CNEMC network and GC; dots represent sites with rates at the significance level of P < 0.05, while crosses represent sites with rates that revealed P values greater than 0.05, and the triangles represent GC.

Close modal

At GC, seasonal average O3 mole fractions exhibited distinct interannual variations during different seasons, revealing strong decreases in summer (–1.7 ppb yr–1, P = 0.22, 2013–2019) and autumn (–2.1 ppb yr–1, P < 0.05, 2013–2018) and slight increases in spring (0.5 ppb yr–1, P = 0.69, 2013–2019) and winter (0.6 ppb yr–1, P < 0.05, 2013–2018), while AOT40 declined in all seasons, mostly strongly in summer (–2.9 ppm h yr–1; Table S1). At urban/suburban sites, however, averaged O3 mole fractions have generally gone upward in all seasons (Table 1), with most prominent increasing rates in summer (avg ± σ: 3.5 ± 1.4 ppb yr–1, max: 7.3 ppb yr–1), with seasonal AOT40 also increasing most rapidly during summer (2.7 ± 1.1 ppm h yr–1 on average), followed by spring and only weakly during autumn in the Beijing–Tianjin–Hebei region (Table S1). The fraction of urban/suburban sites with decreasing O3 mole fractions was negligible in spring and summer and only accounted for 14.4% and 8.5% of the total number of sites in autumn and winter, respectively. Declining O3 rates at urban/suburban sites ranged from –0.4 to –0.1, –1.1 to –0.1, –3.3 to –0.07, and –2.8 to –0.01 ppb yr–1 during spring, summer, autumn, and winter, respectively; however, none of them were significant at the 95% confidence level. Weakly and insignificantly declining AOT40 mainly occurred during spring and summer, with rates ranging from –0.6 to 0.0 ppm h yr–1 and from –0.6 to –0.2 ppm h yr–1, respectively, while no declining AOT40 was found in autumn. It should be noted that reported variation rates of averaged O3 values over a time span of 7 years should not be misinterpreted as trends in the climatological sense since some of these variations may be driven by year-to-year variabilities in meteorological conditions as opposed to anthropogenic impacts. However, it is still evident that O3 in GC varied very differently from the majority of urban sites that revealed rapidly rising O3 mole fractions in warm seasons.

Table 1.

Seasonal change rates of O3 levels at GC, BJ, SJZ, and CNEMC NCP sites during 2013–2019

SitesRates (ppb yr–1)AnnualSpringSummerAutumnWinter
BD GC –1.0 0.5 –1.7 –2.1a 0.6a
Urban 1.3 2.7 2.8 0.4 –0.5
BJ North suburban 0.8 0.8 0.9 0.5 0.3
Urban 0.8 0.8 1.1 0.7 0.3
West suburban 0.4 0.3 0.4 0.4 0.3
SJZ Suburban 0.9 1.3 1.9 0.5 0.1
Urban 1.9 2.3 3.7 1.0 0.4
Rural (mountain site) –0.7 –0.2 –1.1 –0.0 –1.0
NCP P < 0.05 2.4 (0.5–5.6)b 2.4 (0.8–5.4)b 3.5 (1.3–7.3)b 2.1 (0.9–4.1)b 1.4 (0.6–2.8)b
P > 0.05 upward 0.0–2.9b 0.1–3.2b 0.1–4.2b 0.0–2.0b 0.1–1.9b
downward –2.3 to 0.0b –0.4 to –0.1b –1.1 to –0.1b –3.3 to –0.1b –2.8 to 0.0b
NDown/NAllc 5/128 4/73 2/73 18/124 11/129
SitesRates (ppb yr–1)AnnualSpringSummerAutumnWinter
BD GC –1.0 0.5 –1.7 –2.1a 0.6a
Urban 1.3 2.7 2.8 0.4 –0.5
BJ North suburban 0.8 0.8 0.9 0.5 0.3
Urban 0.8 0.8 1.1 0.7 0.3
West suburban 0.4 0.3 0.4 0.4 0.3
SJZ Suburban 0.9 1.3 1.9 0.5 0.1
Urban 1.9 2.3 3.7 1.0 0.4
Rural (mountain site) –0.7 –0.2 –1.1 –0.0 –1.0
NCP P < 0.05 2.4 (0.5–5.6)b 2.4 (0.8–5.4)b 3.5 (1.3–7.3)b 2.1 (0.9–4.1)b 1.4 (0.6–2.8)b
P > 0.05 upward 0.0–2.9b 0.1–3.2b 0.1–4.2b 0.0–2.0b 0.1–1.9b
downward –2.3 to 0.0b –0.4 to –0.1b –1.1 to –0.1b –3.3 to –0.1b –2.8 to 0.0b
NDown/NAllc 5/128 4/73 2/73 18/124 11/129

Bold numbers are rates that passed the 95% significant test.

arepresents change rates calculated for 2013–2018.

bvariation range of the O3 change rates observed at distinct sites in the NCP.

cNDown represents the number of sites with downward O3, while NAll is the total number of sites selected.

To take a closer inspection at the observed rural/urban O3 discrepancies, changes in annually averaged O3 levels at rural GC and its nearby urban/suburban sites (grouped according to their administrative areas into BD, LF, BJU, and BJS) were depicted in Figure 4a, while the same for annual AOT40 was depicted in Figure S2a. Annually averaged O3 at GC fluctuated greatly during 2013–2018, while those in BD, LF, BJU, and BJS steadily increased by 31.2%, 57.8%, 28.7%, and 18.3% during 2013–2019, respectively. At GC, annual average O3 levels were slightly increasing during 2013–2015, however, fell to evidently lower levels than those observed at BD, as well as those observed at the other surrounding urban/suburban sites during 2016–2018. Annual AOT40 in BD, LF, BJU, and BJS also increased steadily by 264%, 182%, 37%, and 35% during the 2013–2019, respectively, revealing stronger increases in higher O3 mole fractions (that were relevant for plant growth) at BD and LF compared to BJ. Annual AOT40 values for GC displayed similar year-to-year variations as annual averaged O3 mole fractions, however, with larger amplitudes. Similar differences between rural and urban O3 variations were also detected in a few other cities, among which was SJZ city (located to the southwest of BD). Decreasing O3 rates and AOT40 in all seasons were detected at a rural mountain site southwest of urban SJZ (SJZ rural in Tables 1 and S1), with the largest negative variation slopes of O3 mole fractions in summer (–1.1 ppb yr–1, P = 0.60) and winter (–1.0 ppb yr–1, P = 0.11). In contrast, evidently growing O3 mole fraction levels and AOT40 were found at urban SJZ and its northern suburban sites, which were particularly strong at urban sites in spring (2.3 ppb yr–1 and 1.6 ppm h yr–1, P < 0.01) and summer (3.7 ppb yr–1 and 3.1 ppm h yr–1, P < 0.05). As was already mentioned, rural sites are very rare in the CNEMC network; the existing few stations with negative O3 mole fraction variation rates confirmed that the rural/urban discrepancy in O3 changes detected at GC was not an individual case but might exist throughout the NCP or even throughout China. Although other suburban sites in the NCP displayed increasing rates in annual average O3 concentrations, in suburban areas downwind of urban regions, O3 revealed stronger increases than other suburban regions in all 4 seasons. For instance, under the local mountain valley circulation pattern, the northern and northwestern suburbs of Beijing (BJ North) are directly downwind of urban Beijing during the afternoon and even more often so during summertime, when synoptic scale circulations are predominated by southern airflows. Thus, weaker O3 increases have been observed at the western suburban site in comparison with the other suburban sites downwind urban Beijing (Table 1). Seasonal average O3 levels at the western suburban site even slightly decreased in spring after 2015, while those observed downwind urban regions displayed continuous upward changes (0.77 ppb yr–1, P = 0.05) during 2013–2019. AOT40 was increasing during spring to autumn at all suburban sites; however, increasing slopes were larger at suburban sites downwind urban regions during spring and summer. Overall, strongest increasing rates were found in urban areas, which transitioned into weak increases at downwind suburban regions, and further became weaker or turned into decreases at less urban impacted suburban or rural areas.

Figure 4.

The changes and relative differences in annual and seasonal O3 between GC and nearby cities. (a) Distinct levels, change rates; (b) relative differences in annual O3 between GC and nearby cities (BD, BJS, BJU, and LF) and (c) seasonal O3 between GC and BD during 2013–2019; the black error bars in (a) represent the standard deviations of the average O3 mixing ratio for each city; numbers in (a) are change rates (in ppb yr–1) and P values of annual averaged O3 in BD, BJS, BJU, LF, and GC.

Figure 4.

The changes and relative differences in annual and seasonal O3 between GC and nearby cities. (a) Distinct levels, change rates; (b) relative differences in annual O3 between GC and nearby cities (BD, BJS, BJU, and LF) and (c) seasonal O3 between GC and BD during 2013–2019; the black error bars in (a) represent the standard deviations of the average O3 mixing ratio for each city; numbers in (a) are change rates (in ppb yr–1) and P values of annual averaged O3 in BD, BJS, BJU, LF, and GC.

Close modal

Aside from distinct temporal variations, discrepancies also existed in observed O3 mole fraction and AOT40 levels among rural and urban/suburban sites. Annually averaged O3 mole fractions at GC were similar to those found in BD during 2013–2015 (Figure 4b), while annual AOT40 at GC evidently exceeded those at BD (Figure S2b). However, from 2016 onward, both average O3 levels and AOT40 at BD exhibited distinctly higher than those at GC, exceeding the later by a maximum of approximately 60% and 161% in 2018, respectively (Figures 4b and S2b). Spring and wintertime average O3 levels in BD were significantly higher than those at GC in most years (27% and 54% on average, respectively), with maximum exceedance of 64% and 82% in spring 2018 and winter 2017, respectively (Figure 4c). Discrepancies in summer and autumn O3 levels between BD and GC were small during 2013–2015, however, steadily increased afterward in summer and also became pronounced in autumn after 2017. Discrepancies in seasonal AOT40 between BD and GC were even more pronounced, especially during spring and summer, with AOT40 in BD being generally lower than those in GC before 2016 and much higher than GC afterward (maximum exceedance of 186% and 161% during spring and summer 2018). Thus, not only did averaged O3 mole fractions vary in reversed directions between rural GC and its surrounding urban/suburban sites, but there were also large discrepancies in O3 levels observed, particularly in the later half of the observation period. Large rural/urban differences in O3 values were also detected in SJZ, where the rural site revealed remarkably higher O3 levels than the urban/suburban sites in all seasons during 2013–2019 (Figure 5), with rural O3 levels reaching on average 1.5, 1.6, 2.2, and 2.8 times of that at urban SJZ in spring, summer, autumn, and winter, respectively (Figure 5). Suburban SJZ sites revealed higher O3 levels than urban SJZ, especially in spring and summer, however, displayed lower levels than rural SJZ. Seasonal AOT40 in suburban SJZ, however, was also significantly higher than urban ones, especially during spring and summer, when it even exceeded rural values (Figure S3). In contrast to the distribution of O3 levels in SJZ, the distribution of averaged O3 mole fractions in Beijing shows a different pattern. Urban BJ revealed higher O3 mole fractions than suburban regions in spring, summer, and autumn, with BJ North sites downwind the urban area revealing the lowest O3 levels and BJ West displaying O3 mole fractions in between urban BJ and BJ North. In winter, seasonal O3 mole fractions at the BJ West suburban site were slightly higher than BJ North and urban BJ. Evidently, highly distinct O3 levels between rural and urban regions were a common phenomenon on the NCP. However, some rural regions showed higher and others displayed lower O3 mole fractions than its nearby urban sites. Additionally, these rural–urban discrepancies varied with season.

Figure 5.

The differences in seasonal O3 between suburban/rural and urban sites in BJ and SJZ during 2013–2019.

Figure 5.

The differences in seasonal O3 between suburban/rural and urban sites in BJ and SJZ during 2013–2019.

Close modal

Thus, neither the temporal variations of O3 nor its levels in rural regions could have been well represented by its nearby urban/suburban sites or by spatial interpolations of their observations. While urban NCP sites mostly exhibited remarkable increasing O3 levels, these transitioned from weakly increasing into negative variations from suburban to rural areas. Rural/urban discrepancies in O3 mole fractions varied over different time scales and were inconsistent among different locations. Additionally, the presence of urban park stations (such as Haidian Park, located on the northwest of central urban in Beijing, 39.59°N, 116.18°E) with decreasing ozone temporal variations reversed to those of nearby urban sites further confirmed that distinct land coverage types and dry deposition rates might have led to significant urban–rural ozone discrepancies throughout the NCP. Previous studies mostly employed CNEMC network data or spatial interpolations/reanalysis datasets based thereupon to evaluate O3 agricultural impacts in China (Zhao et al., 2018; Feng et al., 2019a; Feng et al., 2020; Ren et al., 2020; Dong et al., 2021), which might have introduced large uncertainties into their estimates. The biases in their estimates could have been either positive or negative, varying in space and time. In the next section, the O3 agricultural impacts and the biases in their estimates will be evaluated in detail.

### 3.2. Impacts on the estimation of ozone induced agricultural losses

Since O3 levels in rural areas were not adequately represented by the current CNEMC network, it is necessary to evaluate how much the agricultural impacts were misinterpreted based on the urban/suburban O3 observations. The ozone metric AOT40 was calculated to account for the cumulative harmful impacts of high O3 mole fractions during the growing season of distinct crops, and losses in RYL were estimated based upon that.

Variations of AOT40 calculated based on O3 mole fractions observed at rural GC and nearby urban/suburban (BD, LF) sites for 3 major crops relevant for the NCP region (wheat, maize, and soybean) are displayed in Figure 6a1 and a2. AOT40wheat displayed large annual fluctuations during 2013–2019 at GC, ranging from 6.9 to 26.0 ppm h, with significantly elevated levels in 2014–2015, which rapidly decreased afterward. AOT40wheat at BD sites increased significantly at a rate of 3.6 ppm h yr–1 (R2 = 0.92, P < 0.01) after 2013, reaching 29.9 ± 1.9 ppm h in 2018 and declined markedly in 2019 back to 20.7 ± 1.4 ppm h. AOT40wheat at LF rose rapidly from 3.3 ± 1.6 ppm h in 2013 to 15.2 ± 0.6 ppm h in 2014, after which it steadily increased at a slower pace (ranging from 14.8 to 19.9 ppm h during 2014–2019). For maize and soybean, which share the same growing period, AOT40 varied within ranges of 11.7–48.4 ppm h, 15.3–40.9 ppm h, and 15.0–34.9 ppm h in GC, BD, and LF, respectively. AOT40maize/soybean at GC decreased after reaching a maximum in 2015, while BD and LF both revealed steady increases at respective rates of 3.3 ± 1.3 and 1.5 ± 0.4 ppm h yr–1. Previous studies that directly used data from the CNEMC network or related reanalysis data have all reported upward trends on O3 agricultural impacts for the BD region (Hu et al., 2020; Dong et al., 2021), which obviously is highly inconsistent with actual variations observed at an actual rural site in this area. The AOT40 at CNEMC network sites with decreasing O3 tendencies (urban park sites or rural village sites, Section 3.1) also changed reversed to surrounding urban/suburban sites. At the village site in SJZ, AOT40 revealed large negative variation slopes of –2.5 (P = 0.09) and –6.1 (P < 0.01) ppm h yr–1 from 2013 to 2018 for wheat and maize/soybean, respectively (Figure 7). However, at nearby urban/suburban sites, AOT40 consistently increased after 2013 at respective rates of 1.8 ± 0.5 (wheat, P < 0.01) and 2.9 ± 1.1 (maize/soybean, P < 0.05) ppm h yr–1.

Figure 6.

The changes of AOT40 and relative yield loss during 2013–2019 at BD, LF, and GC. (a) The changes of AOT40 in the wheat and maize/soybean growing seasons during 2013–2019 at BD, LF, and GC; (b) their differences between GC and BD, LF; (c) the wheat, maize, and soybean relative yield changes during 2013–2019 driven by AOT40 at BD, LF, and GC.

Figure 6.

The changes of AOT40 and relative yield loss during 2013–2019 at BD, LF, and GC. (a) The changes of AOT40 in the wheat and maize/soybean growing seasons during 2013–2019 at BD, LF, and GC; (b) their differences between GC and BD, LF; (c) the wheat, maize, and soybean relative yield changes during 2013–2019 driven by AOT40 at BD, LF, and GC.

Close modal
Figure 7.

The variations of AOT40 during 2013–2019 at rural and urban/suburban areas in SJZ. The variations of AOT40 in wheat (a) and AOT40 (b) in maize/soybean growing seasons during 2013–2019.

Figure 7.

The variations of AOT40 during 2013–2019 at rural and urban/suburban areas in SJZ. The variations of AOT40 in wheat (a) and AOT40 (b) in maize/soybean growing seasons during 2013–2019.

Close modal

Not only were the AOT40 variations poorly represented by the urban/suburban measurements, the estimation of the AOT40 level was also highly inaccurate. The relative difference in AOT40 between rural GC and BD/LF sites for wheat and maize/soybean are displayed in Figure 6b1 and b2. Generally, AOT40 in rural areas was underestimated before 2016, while afterward, it was strongly overestimated. Using O3 observations in BD, AOT40wheat was overestimated by 140% in 2018, while AOT40maize/soybean revealed overestimations up to 221% in 2019. Similar to GC, rural SJZ revealed AOT40 higher than its nearby urban/suburban sites before 2016 and lower ones afterward (Figure 7), with the largest discrepancies occurring in 2018 (384% and 411% for wheat and maize/soybean, respectively). This indicates that O3 agricultural impacts in rural NCP areas might have declined during 2013–2018, which is reversed to past results based on CNEMC network observations. Using the Weather Research and Forecasting (WRF)-Community Multiscale Air Quality Modeling System (CMAQ) Model, Liu and Wang (2020) proposed there were downward trends in annual O3 MDA8 within large parts of rural areas after 2013, which were not covered by the CNEMC network. However, due to the lack of continuous rural observations, modeling results have not been validated, and the distinct trends and levels between rural and urban O3 and according impacts on agriculture have not gained sufficient attention. The variational differences and corresponding relative discrepancies in AOT40 between the rural and urban/suburban areas will bring large uncertainties to the evaluation of the agricultural and economic losses brought on by O3 agricultural impacts (Table 1).

The RYLs and their variation during 2013–2019 for the 3 crops are shown in Figure 6c and Table 2, which also exhibit the discrepancies in RYL estimations caused by the discrepancies in AOT40 between rural and urban/suburban areas. It can be discerned that RYLs varied overall within a similar range during the entire period of interest. RYLwheat calculated using O3 data from the rural site (GC) was on average 37 ± 14% yr–1 (in the range of 16%–59% yr–1), while those calculated based on O3 observations in the nearby urban/suburban sites in BD and LF averaged 43% ± 13% yr–1 (24%–68% yr–1) and 35% ± 12 % yr–1 (7%–45% yr–1), respectively. RYLmaize calculated applying O3 measurements during 2013–2019 from GC, BD, and LF was on average 8% ± 4% yr–1 (2%–15% yr–1), 9% ± 3% yr–1 (4%–13% yr–1), and 7% ± 2% yr–1 (3%–11% yr–1), respectively, while RYLsoybean reached average levels of 30% ± 13% yr–1 (12%–46% yr–1), 33% ± 10% yr–1 (16%–45% yr–1), and 27% ± 7% yr–1 (14%–49% yr–1) for soybean, respectively. Using urban/suburban AOT40 at BD sites overestimated the RYL in rural areas by 17%, 9%, and 8% on average for wheat, maize, and soybean, while using AOT40 at LF resulted in underestimations of 6%, 13%, and 11%. The RYL obtained using GC O3 measurements displayed larger variation amplitudes than those calculated using O3 data from LF. Despite similar RYL levels obtained through GC and BD O3 data, highly distinct patterns and temporal variations were detected. RYLwheat in rural GC reached a maximum of 59% in 2015 and decreased rapidly afterward; those calculated based on urban/suburban observations clearly underestimated rural observations during 2014–2015 (Table 2). RYLwheat calculated applying BD data increased rapidly to its maximum (68%) in 2018, while that of LF reached its maximum in 2017 and remained at around 40% since then. Both were overestimating rural RYLwheat from 2016 onward, reaching a maximum relative overestimate of 136 and 48% during 2018, respectively. For maize and soybean, the largest underestimates in RYLs from BD/LF observations occurred during 2015, reaching 51/36% and 55/41%, respectively, while overestimates grew steadily from 2017 onward, reaching a maximum of 422/260% and 294/180 % in 2019. These were caused by the reversed O3 variations and large mole fraction differences between nearby urban/suburban and rural areas during the 2 growth seasons. As such, agricultural impact evaluations using nearby urban O3 measurements, which cannot reflect the true year-to-year variations of rural O3, would lead to large discrepancies to the estimation results (absolute differences in RYLs accumulated over 2013–2019 reaching 122%, 32%, and 105% for wheat, maize, and soybean, respectively, using data in BD), because crop cultivation, growth and yield varies from year-to-year.

Table 2.

Annual RYL of wheat, maize, and soybean driven by AOT40 at BD, LF, and GC

WheatMaizeSoybean
YearGC (%)BD (%)LF (%)GC (%)BD (%)LF (%)GC (%)BD (%)LF (%)
2013 16 24 30 16 15
2014 51 29 35 34 34 26
2015 59 43 34 15 54 30 22
2016 28 42 37 27 21 27
2017 42 50 45 10 13 11 37 45 39
2018 29 68 43 11 16 40 27
2019 32 47 41 11 12 41 32
Avg ± σ 37 ± 15 43 ± 14 35 ± 13 8 ± 4 8 ± 3 7 ± 3 30 ± 14 32 ± 11 27 ± 8
Accumulated absolute difference  122 85  32 28  105 88
∑ |RYLsite-RYLGC|
WheatMaizeSoybean
YearGC (%)BD (%)LF (%)GC (%)BD (%)LF (%)GC (%)BD (%)LF (%)
2013 16 24 30 16 15
2014 51 29 35 34 34 26
2015 59 43 34 15 54 30 22
2016 28 42 37 27 21 27
2017 42 50 45 10 13 11 37 45 39
2018 29 68 43 11 16 40 27
2019 32 47 41 11 12 41 32
Avg ± σ 37 ± 15 43 ± 14 35 ± 13 8 ± 4 8 ± 3 7 ± 3 30 ± 14 32 ± 11 27 ± 8
Accumulated absolute difference  122 85  32 28  105 88
∑ |RYLsite-RYLGC|

RYL = relative yield loss.

The RYL values calculated using O3 from GC were also compared against those reported in the literature in Table 3. Most of the previous studies used data from CNEMC network (Feng et al., 2019a; Feng et al., 2019b; Feng et al., 2019c; Hu et al., 2020; Ren et al., 2020; Zhao et al., 2020), while others tried to fill in the spatial gaps through model assimilations of observational data (Dong et al., 2021) or by using modeling results from mesoscale chemistry-transport models (Tang et al., 2013; Lin et al., 2018). Only very few evaluations were based directly on rural observations (Zhu et al., 2015; Zhang et al., 2017). Evaluation results varied greatly among different O3 data sources. CNEMC observation-based results generally revealed higher RYLs in the NCP region than in the Yangtze River Delta or in whole China, suggesting the severity of O3 induced agricultural impacts in the NCP. Our rural observation-based results revealed an even larger variation range of RYLs, with an average level clearly higher than those previously reported for Hebei Province or the NCP, providing evidence that RYLs were even higher than expected, especially before 2016. Simulation results based on chemistry-transport models also revealed large uncertainties going in different directions. RYLwheat based on O3 simulated using WRF-CMAQ was much higher than those based on CNEMC observations, whereas WRF-Chem evidently underestimated that of Hebei Province. RYLmaize based on O3 measurements from GC also revealed larger variability than those reported for the period of 2013–2018 based on urban/suburban observations. Compared to the RYLsoybean (ranging from 23.4% to 30.2%, depending on the O3 index used and soybean varieties) estimated based on open-top chambers experiments at a rural site in Northeast China during 2013–2014, RYLsoybean in the NCP displayed a higher level and a smaller variation range during 2013–2014 and much larger variability during 2013–2019. Aside from the distinct variability in RYLs, many previous studies reported rising trends in O3 induced RYLs within China and even more pronounced increases in the NCP (Feng et al., 2019b; Feng et al., 2019c; Hu et al., 2020; Ren et al., 2020; Dong et al., 2021), which are inconsistent with our results based on rural observation that indicate decreasing RYLs for all crops especially after 2017.

Table 3.

Summary of estimated wheat, maize, and soybean RYLs due to O3 pollution in China

CropsAreasTimeRYL (%)Data Sources
Wheat China 2015–2018 8.5–14.3 CNEMC network (Zhao et al., 2020
China 2014 21–39 WRF-CMAQ (Lin et al., 2018
China 2015 CNEMC network (Feng et al., 2019a
YRD 2014–2019 14.9 CNEMC network (Ren et al., 2020
NCP 2013–2018 17.9–38.6 CNEMC network reanalysis data (Dong et al., 2021
NCP 2014–2017 18.5–30.8 CNEMC network (Hu et al., 2020
Hebei Province 2015–2016 28–59 Rural observation (this study)
2013–2019 16–59
Hebei Province 2015–2016 22.8–23.4 CNEMC network (Feng et al., 2019b
Hebei 2000 7.3–8.0 WRF-Chem (Tang et al., 2013
2020 5.1–6.0
Northwest-Shandong Plain of China 2012 12.9 Rural observation (Zhu et al., 2015
Maize China 2014 3–6 WRF-CMAQ (Lin et al., 2018
NCP 2014–2017 8.2–13.4 CNEMC network (Feng et al., 2019c
NCP 2013–2018 7.5–11.9 CNEMC network reanalysis data (Dong et al., 2021
Hebei Province 2013–2018 4–15 Rural observation (this study)
2013–2019 2–15
Soybean Northeast China 2013–2014 23.4–30.2 Rural observation (Zhang et al., 2017
Hebei Province 2013–2014 30–34 Rural observation (this work)
2013–2019 12–46
CropsAreasTimeRYL (%)Data Sources
Wheat China 2015–2018 8.5–14.3 CNEMC network (Zhao et al., 2020
China 2014 21–39 WRF-CMAQ (Lin et al., 2018
China 2015 CNEMC network (Feng et al., 2019a
YRD 2014–2019 14.9 CNEMC network (Ren et al., 2020
NCP 2013–2018 17.9–38.6 CNEMC network reanalysis data (Dong et al., 2021
NCP 2014–2017 18.5–30.8 CNEMC network (Hu et al., 2020
Hebei Province 2015–2016 28–59 Rural observation (this study)
2013–2019 16–59
Hebei Province 2015–2016 22.8–23.4 CNEMC network (Feng et al., 2019b
Hebei 2000 7.3–8.0 WRF-Chem (Tang et al., 2013
2020 5.1–6.0
Northwest-Shandong Plain of China 2012 12.9 Rural observation (Zhu et al., 2015
Maize China 2014 3–6 WRF-CMAQ (Lin et al., 2018
NCP 2014–2017 8.2–13.4 CNEMC network (Feng et al., 2019c
NCP 2013–2018 7.5–11.9 CNEMC network reanalysis data (Dong et al., 2021
Hebei Province 2013–2018 4–15 Rural observation (this study)
2013–2019 2–15
Soybean Northeast China 2013–2014 23.4–30.2 Rural observation (Zhang et al., 2017
Hebei Province 2013–2014 30–34 Rural observation (this work)
2013–2019 12–46

RYL = relative yield loss; WRF-CMAQ = Weather Research and Forecasting-Community Multiscale Air Quality Modeling System.

Accordingly, economic loss induced by O3 agricultural impacts was also misinterpreted, whose uncertainties need to be quantified. The cultivating land area of BD city made up on average 12.1% ± 1.2% of that in Hebei Province during 2013–2019, while its total crop production accounted for 13.7% ± 2.0% of the entire Province’s production amount (National Bureau of Statistics of China, 2019). Since the yield data of specific crops were not made public for BD city, the CPL was calculated based on agricultural production data of the entire Hebei Province (Table 4), assuming that GC observations were representative of rural regions of Hebei Province. The average annual CPL of wheat, maize, and soybean reached 9.9 ± 6.1 × 106 ton (2.7–21.5 × 106 ton), 1.7 ± 0.9 × 106 ton (0.5–3.5 × 106 ton) during 2013–2019 and 9.3 ± 4.4 × 104 ton (4.1–18.1 × 104 ton) during 2013–2018, respectively. Similar to the RYL differences, the CPL differences between GC and BD/LF were also increasing from negative to positive during 2013–2019, with the annual discrepancies for wheat, maize, and soybean ranging from –62% to 425%, from –58% to 477% and from –64% to 246% at BD and from –65% to 84%, from –68% to 321% and from –76% to 87% at LF, respectively. According to the annual purchase prices of these crops (Table 5), the total ECL caused by O3 in GC amounts to a total of 3.1 × 1010 USD (2.6 × 1010, 4.9 × 109 and 4.0 × 108 USD for wheat, maize, and soybean, respectively) during 2013–2019, taking up 0.89% and 0.04% of the gross domestic production in Hebei Province (3.47 × 1012 USD) and China (8.27 × 1013 USD), respectively.

Table 4.

Annual CPL and ECL of wheat, maize, and soybean due to GC O3 pollution

MetricsCrops2013201420152016201720182019
CPL (106 ton) Wheat 2.65 15.27 21.53 5.81 10.94 5.87 6.86
Maize 1.68 1.93 3.47 1.58 2.32 0.74 0.45
Soybean 0.08 0.10 0.18 0.06 0.10 0.04 —
ECL (108 USD) Wheat 9.42 57.56 84.28 22.38 36.46 20.02 27.41
Maize 8.24 8.52 14.98 4.04 8.61 2.96 1.83
Soybean 0.54 0.82 1.47 0.35 0.61 0.25 —
MetricsCrops2013201420152016201720182019
CPL (106 ton) Wheat 2.65 15.27 21.53 5.81 10.94 5.87 6.86
Maize 1.68 1.93 3.47 1.58 2.32 0.74 0.45
Soybean 0.08 0.10 0.18 0.06 0.10 0.04 —
ECL (108 USD) Wheat 9.42 57.56 84.28 22.38 36.46 20.02 27.41
Maize 8.24 8.52 14.98 4.04 8.61 2.96 1.83
Soybean 0.54 0.82 1.47 0.35 0.61 0.25 —
Table 5.

Annual crop productionsa in Hebei Province and purchase pricesb of wheat, maize, and soybean

IndexesCrops2013201420152016201720182019
Production (103 ton) Wheat 14,168.2 14,413.5 14,777.0 14,753.3 14,993.8 14,460.9 14,625.7
Maize 19,227.9 18,988.4 18,977.4 20,312.1 20,354.8 19,411.5 19,866.4
Soybean 180.7 182.2 153.3 160.3 170.6 212.3 —
Price (USD ton–1Wheat 355.1 377.0 391.40 385.1 333.3 340.8 399.9
Maize 489.0 441.9 432.4 264.3 370.8 398.0 408.6
Soybean 677.9 869.7 808.2 581.5 596.5 596.8 594.2
IndexesCrops2013201420152016201720182019
Production (103 ton) Wheat 14,168.2 14,413.5 14,777.0 14,753.3 14,993.8 14,460.9 14,625.7
Maize 19,227.9 18,988.4 18,977.4 20,312.1 20,354.8 19,411.5 19,866.4
Soybean 180.7 182.2 153.3 160.3 170.6 212.3 —
Price (USD ton–1Wheat 355.1 377.0 391.40 385.1 333.3 340.8 399.9
Maize 489.0 441.9 432.4 264.3 370.8 398.0 408.6
Soybean 677.9 869.7 808.2 581.5 596.5 596.8 594.2

aThe crop productions were the total crop productions of the 3 crops in Hebei Province, obtained from the Statistically Yearbook of National Bureau of Statistics of China (2019).

bThe purchase prices were taken from Food and Agricultural Organization (FAO) of the United Nations Statistical Database (FAO of the United Nations, 2022a).

Due to the significant rural–urban discrepancies in O3 levels and variations, large uncertainties were also brought into ozone induced agricultural economic loss estimations when using observations from nearby nonagricultural sites instead of rural O3 measurements. Therefore, networked rural O3 observations are urgently required for more accurate evaluations of O3 agricultural impacts and related economic losses.

In this study, it was manifested that large differences in O3 levels existed between rural and urban/suburban sites in the NCP and that O3 behaved in opposite directions, with decreases in rural areas and increases in urban/suburban regions. Since open accessible data from the CNEMC network do not contain rural observations representative of agricultural regions, previous studies on O3 induced agricultural impacts have based their results on urban/suburban observations, which brought large uncertainties into according estimations and led to biased prognostics for future variations. The AOT40 values, RYLs and according CPLs/ECLs, calculated with rural O3 observation, displayed larger variabilities than those calculated using urban/suburban observations. Based on the limited available rural observations, O3 impacts on all assessed crops have shown decreases after 2016 rather than the continuously increasing trend previously diagnosed based on urban/suburban measurements. O3 in the latitude band of the NCP features annual maximum mole fractions during summertime (Figure 8) (Lin et al., 2008; Lin et al., 2009; Liu et al., 2020), with monthly AOT40 also depicting an annual peak during June (Figure S4). Monthly AOT40 exceeds the threshold level for plant health protection only during May to September, which is why AOT40maize/soybean was higher than AOT40wheat (Figure 6a). However, wheat is the most sensitive crop toward O3 exposure, followed by soybean and maize (Peng et al., 2019), which is why RYLwheat was slightly higher than RYLsoybean and significantly exceeded RYLmaize (Figure 6c). O3 damage to crops growing in summertime has been reduced due to lowered summertime O3 levels, while O3 threat to winter wheat, which grows throughout spring to early summer and is more sensitive toward O3 pollution, still requires more attention because of the observed rising tendencies of springtime O3 (Table 1).

Figure 8.

The seasonal cycle of O3 mixing ratio during 2013–2019 at the GC site. The red error bars represent the standard deviations of the hourly O3 mixing ratio.

Figure 8.

The seasonal cycle of O3 mixing ratio during 2013–2019 at the GC site. The red error bars represent the standard deviations of the hourly O3 mixing ratio.

Close modal

These conclusions provide important implications and guidance for future research on O3 related studies. First of all, regarding future studies on O3 agricultural impacts, it should be noted that a fixed exposure–response relationship was applied in evaluating O3 induced crop yield losses based on the AOT40 metric, which is based on the assumption that sensitivities of crops toward O3 exposure did not change over the years. This assumption may introduce uncertainties into current estimations since crop sensitivity toward O3 exposure varies greatly among different crop varieties (e.g., Asian wheat is more sensitive to O3 pollution than European and American varieties, Pleijel et al., 2019). Additionally, crop varieties develop with time under new agricultural technologies. Latest studies have shown that recently developed high-yield wheat varieties are even more sensitive to O3 due to their high stomatal conductance and low antioxidant capacity (Biswas et al., 2008). Increased sensitivity of new wheat varieties in combination with increasing springtime O3 might result in more crop yield reductions and associated economic losses. Therefore, the exposure–response relationships of different varieties of crops to O3 of different levels need to be further assessed in future field experiments to provide better parameterizations for future estimations of O3 induced agricultural impacts.

Second, the damage of high O3 to vegetation is not limited to reductions in crop yields; the growth of all plants including those in farmlands, orchards, grasslands, and forests are under its impact (Biswas et al., 2008; Piikki et al., 2008; Jiang et al., 2018). Compared to urban/suburban regions, rural areas have much higher vegetation coverage and are therefore even more important in the carbon and nitrogen cycle of the ecosystem (Felzer et al., 2005; Sitch et al., 2007; Li et al., 2020b). Exposure to O3 pollution can decrease the CO2 uptake ability of plants and further aggravate global warming in addition to its own greenhouse effects (Collins et al., 2010; Kinose et al., 2020). Thus, the accurate representation of rural O3 is also crucial for future climate predictions. How current O3 changes in rural areas are influencing the carbon and nitrogen cycle is worth deeper investigations.

Third, since there are significant discrepancies in O3 levels and long-term variations between rural and urban/suburban areas, it is necessary to strengthen the observations of O3 in rural areas, especially during crop growing seasons. Rural O3 observation sites are highly advised to be incorporated into the CNEMC network, which is not only crucial for the assessment of ozone’s agricultural, ecological, environmental, and climate impacts on larger spatial scales but also the foundation for future investigations into what has caused the differences between rural and urban/suburban O3 formation, levels, and temporal variations, which brings us to the last point.

Clean air actions were activated in China after 2013, putting in place various regulations including strengthened emission standards of coal-fired power plants, especially the “Ultralow” emission standard since 2016; strengthened industrial emission standards and shutting down outdated industries and factories not able to meet those new standards; elimination of small coal-fired boilers and addition of SO2 and particulate control technologies to large boilers; replacement of residential coal combustion with electricity and natural gas; implementation of Euro 4 standards to heavy duty gasoline vehicles and diesel vehicles in 2013 and 2014, respectively, and transitioning to Euro 5 standards for light duty gasoline vehicles and diesel vehicles by 2017 (Zheng et al., 2018) (Figure 9). Different O3 variations between rural and urban/suburban areas under the implementation of these emission control measures might have been caused by various influencing factors. Distinct emission patterns exist between rural and urban/suburban areas, thus resulting in distinct response of O3 formation to emission reductions of O3 precursors. O3 formation responds highly nonlinear to its precursor concentrations, depending not only on its precursor concentrations but also on their ratios (especially the NOx/VOCs ratio) and spatiotemporal variations. Thus, both differences in O3 production regimes and distinct changes in NOx/VOCs emission ratios between rural and urban/suburban sites could have led to the observed reversed O3 temporal variations. In North America and Europe, annual average O3 increased, while O3 peaks decreased in both urban and rural areas in the initial emission reduction stage (Paoletti et al., 2014). The increase in annual average O3 was mainly a response to weakened NO-titration under strong NOx reductions, which was also observed at all types of stations in our study. However, the observed decrease in O3 peaks at rural GC was not observed at urban and suburban sites, which suggests that O3 production in these regions might still be VOCs-limited, especially during warm seasons. Thus, it is suggested for future studies to systematically investigate the O3 production regime, especially its rural/urban differences and temporal variations, for which long-term rural VOCs and NOx measurements are highly important. Agricultural NOx might further contribute to rural/urban differences; however, it is still undecided whether these emissions are promoting or prohibiting the formation of O3 (Lu et al., 2021; Wang et al., 2022). Aside from that, aerosol formation mechanisms and aerosol composition may also differ among rural and urban/suburban sites due to different relative humidity conditions as well as precursor and oxidant levels, which may result in different gas–particulate interactions. Last but not least, O3 deposition rates differ much among distinct surface coverage types, which might also have partly contributed to the observed rural–urban differences. To better assess the discrepancies in rural and urban O3 deposition, more O3 flux measurements are encouraged in the future, which in agricultural regions can provide important information on how much O3 is lost to crop and soil. The addition of these aforementioned measurements would provide support for the assessment of emission reduction differences between rural and urban regions and also benefit prognostics on future variations of rural O3.

Figure 9.

Changes in emission control measures, rural and urban/suburban O3, and associated crop relative yield loss.

Figure 9.

Changes in emission control measures, rural and urban/suburban O3, and associated crop relative yield loss.

Close modal

The data used in this study are provided in the supplementary material.

The supplemental files for this article can be found as follows:

Figure S1. Distribution of annual (a) and seasonal (b–d) AOT40 change rates during 2013–2019 in the NCP.

Figure S2. The changes and relative differences in annual and seasonal AOT40 between GC and nearby cities.

Figure S3. The differences in seasonal AOT40 between suburban/rural and urban sites in BJ and SJZ during 2013–2019.

Figure S4. The average monthly change of AOT40 during 2013–2019 at the GC site.

Table S1. Annual and seasonal change rates of AOT40 levels at GC, BD, BJ, SJZ, and CNEMC NCP sites during 2013–2019.

Table S2. Monthly average (± standard deviation) of O3 mixing ratios at GC during 2013–2019 (Units: ppb).

Table S3. Annual AOT40 for wheat, maize, and soybean during the growing seasons at GC from 2013 to 2019.

Table S4. The locations and types of the observation sites at GC, BD, LF, BJ, and SJZ.

This work is supported by the National Natural Science Foundation of China (41875159, 42075112, 41775127, and 41330422) and CAMS projects (2020KJ003, 2020Z002, and 2017Z011).

The authors have no competing interests to declare.

Contributed to conception and design: WX, XX.

Contributed to acquisition of data: WX, GZ, WL, HZ, SR.

Contributed to analysis and interpretation of data: XZ.

Drafted and/or revised the article: XZ, WX, XX.

Approved the submitted version for publication: XZ, WX, XX, GZ, WL, HZ, SR, JC.

Contributed equally to this work: XZ, WX.

Ainsworth
,
EA
.
2017
.
Understanding and improving global crop response to ozone pollution
.
The Plant Journal
90
(
5
):
886
897
. DOI: https://doi.org/10.1111/tpj.13298.
Air Quality Historical Data Platform
.
2020
.
Air quality data [dataset]
.
Available at
https://aqicn.org/data-platform.
Accessed 27 April 2022
.
Aunan
,
K
,
Berntsen
,
T
,
Seip
,
H
.
2000
.
Surface ozone in China and its possible impact on agricultural crop yields
.
Ambio
29
. DOI: https://doi.org/10.1579/0044-7447-29.6.294.
Avnery
,
S
,
Mauzerall
,
DL
,
Liu
,
J
,
Horowitz
,
LW
.
2011
a.
Global crop yield reductions due to surface ozone exposure: 1. Year 2000 crop production losses and economic damage
.
Atmos Environment
45
(
13
):
2284
2296
. DOI: https://doi.org/10.1016/j.atmosenv.2010.11.045.
Avnery
,
S
,
Mauzerall
,
DL
,
Liu
,
J
,
Horowitz
,
LW
.
2011
b.
Global crop yield reductions due to surface ozone exposure: 2. Year 2030 potential crop production losses and economic damage under two scenarios of O3 pollution
.
Atmos Environment
45
(
13
):
2297
2309
. DOI: https://doi.org/10.1016/j.atmosenv.2011.01.002.
Bender
,
J
,
Weigel
,
H-J
.
2011
.
Changes in atmospheric chemistry and crop health: A review
.
Agronomy for Sustainable Development
31
(
1
):
81
. DOI: https://doi.org/10.1051/agro/2010013.
Biswas
,
DK
,
Xu
,
H
,
Li
,
YG
,
Sun
,
JZ
,
Wang
,
XZ
,
Han
,
XG
,
Jiang
,
G
.
2008
.
Genotypic differences in leaf biochemical, physiological and growth responses to ozone in 20 winter wheat cultivars released over the past 60 years
.
Global Change Biology
14
(
1
):
46
59
. DOI: https://doi.org/10.1111/j.1365-2486.2007.01477.x.
Carter
,
CA
,
Cui
,
X
,
Ding
,
A
,
Ghanem
,
D
,
Jiang
,
F
,
Yi
,
F
,
Zhong
,
F
.
2017
.
Stage-specific, nonlinear surface ozone damage to rice production in China
.
Scientific Reports
7
(
1
):
44224
. DOI: https://doi.org/10.1038/srep44224.
Chuwah
,
C
,
van Noije
,
T
,
van Vuuren
,
DP
,
Stehfest
,
E
,
Hazeleger
,
W
.
2015
.
Global impacts of surface ozone changes on crop yields and land use
.
Atmos Environment
106
:
11
23
. DOI: https://doi.org/10.1016/j.atmosenv.2015.01.062.
Cohen
,
AJ
,
Brauer
,
M
,
Burnett
,
R
,
Anderson
,
HR
,
Frostad
,
JJ
,
Estep
,
K
,
Balakrishnan
,
K
,
Brunekreef
,
B
,
Dandona
,
L
,
Dandona
,
R
,
Feigin
,
V
,
Freedman
,
G
,
Hubbell
,
B
,
Jobling
,
A
,
Kan
,
H
,
Knibbs
,
L
,
Liu
,
Y
,
Martin
,
R
,
Morawska
,
L
,
Pope
,
CA
,
Shin
,
H
,
Straif
,
K
,
Shaddick
,
G
,
Thomas
,
M
,
Van Dingenen
,
R
,
van Donkelaar
,
A
,
Vos
,
T
,
Murray
,
CJL
,
Forouzanfar
,
MH
.
2017
.
Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: An analysis of data from the Global Burden of Diseases Study 2015
.
The Lancet
389
(
10082
):
1907
1918
. DOI: https://doi.org/10.1016/S0140-6736(17)30505-6.
Collins
,
WJ
,
Sitch
,
S
,
Boucher
,
O
.
2010
.
How vegetation impacts affect climate metrics for ozone precursors
.
Journal of Geophysical Research: Atmospheres
115
(
D23
). DOI: https://doi.org/10.1029/2010JD014187.
Dang
,
R
,
Liao
,
H
,
Fu
,
Y
.
2021
.
Quantifying the anthropogenic and meteorological influences on summertime surface ozone in China over 2012–2017
.
Science of the Total Environment
754
:
142394
. DOI: https://doi.org/10.1016/j.scitotenv.2020.142394.
Ding
,
AJ
,
Wang
,
T
,
Thouret
,
V
,
Cammas
,
JP
,
Nédélec
,
P
.
2008
.
Tropospheric ozone climatology over Beijing: Analysis of aircraft data from the MOZAIC program
.
Atmospheric Chemistry and Physics
8
(
1
):
1
13
. DOI: https://doi.org/10.5194/acp-8-1-2008.
Dong
,
C
,
Gao
,
R
,
Zhang
,
X
,
Li
,
H
,
Wang
,
W
,
Xue
,
L
.
2021
.
Assessment of O3-induced crop yield losses in northern China during 2013–2018 using high-resolution air quality reanalysis data
.
Atmos Environment
259
:
118527
. DOI: https://doi.org/10.1016/j.atmosenv.2021.118527.
Environmental Protection Agency
.
2013
.
Integrated science assessment (ISA) for ozone and related photochemical oxidants (Final Report, Feb 2013)
.
Washington, DC
:
Environmental Protection Agency
.
Felzer
,
B
,
Reilly
,
J
,
Melillo
,
J
,
Kicklighter
,
D
,
Sarofim
,
M
,
Wang
,
C
,
Prinn
,
R
,
Zhuang
,
Q
.
2005
.
Future effects of ozone on carbon sequestration and climate change policy using a global biogeochemical model
.
Climatic Change
73
(
3
):
345
373
. DOI: https://doi.org/10.1007/s10584-005-6776-4.
Feng
,
Z
,
De Marco
,
A
,
Anav
,
A
,
Gualtieri
,
M
,
Sicard
,
P
,
Tian
,
H
,
Fornasier
,
F
,
Tao
,
F
,
Guo
,
A
,
Paoletti
,
E
.
2019
a.
Economic losses due to ozone impacts on human health, forest productivity and crop yield across China
.
Environment International
131
:
104966
. DOI: https://doi.org/10.1016/j.envint.2019.104966.
Feng
,
Z
,
Hu
,
T
,
Tai
,
APK
,
Calatayud
,
V
.
2020
.
Yield and economic losses in maize caused by ambient ozone in the North China Plain (2014–2017)
.
Science of the Total Environment
722
:
137958
. DOI: https://doi.org/10.1016/j.scitotenv.2020.137958.
Feng
,
Z
,
Kobayashi
,
K
,
Li
,
P
,
Xu
,
Y
,
Tang
,
H
,
Guo
,
A
,
Paoletti
,
E
,
Calatayud
,
V
.
2019
b.
Impacts of current ozone pollution on wheat yield in China as estimated with observed ozone, meteorology and day of flowering
.
Atmos Environment
217
:
116945
. DOI: https://doi.org/10.1016/j.atmosenv.2019.116945.
Feng
,
Z
,
Shang
,
B
,
Gao
,
F
,
Calatayud
,
V
.
2019
c.
Current ambient and elevated ozone effects on poplar: A global meta-analysis and response relationships
.
Science of the Total Environment
654
:
832
840
. DOI: https://doi.org/10.1016/j.scitotenv.2018.11.179.
Feng
,
Z-W
,
Jin
,
M-H
,
Zhang
,
F-Z
,
Huang
,
Y-Z
.
2003
.
Effects of ground-level ozone (O3) pollution on the yields of rice and winter wheat in the Yangtze River delta
.
Journal of Environmental Sciences
15
:
360
362
.
Food and Agriculture Organization of the United Nations
.
2022
a.
Prices_E_All_Data_(Normalized) [dataset]
.
Microsoft Excel workbook (146346 KB)
.
Available at
http://www.fao.org/faostat/en/#data/QCL.
Accessed 27 April 2022
.
Food and Agriculture Organization of the United Nations
.
2022
b.
Production_Crops_E_All_Data_(Normalized) [dataset]
.
Microsoft Excel workbook (256181 KB)
.
Available at
http://www.fao.org/faostat/en/#data/QCL.
Accessed April 27, 2022
.
Harmens
,
H
,
Hayes
,
F
,
Mills
,
G
,
Sharps
,
K
,
Osborne
,
S
,
Pleijel
,
H
.
2018
.
Wheat yield responses to stomatal uptake of ozone: Peak vs rising background ozone conditions
.
Atmos Environment
173
:
1
5
. DOI: https://doi.org/10.1016/j.atmosenv.2017.10.059.
Hu
,
T
,
Liu
,
S
,
Xu
,
Y
,
Feng
,
Z
,
Calatayud
,
V
.
2020
.
Assessment of O3-induced yield and economic losses for wheat in the North China Plain from 2014 to 2017, China
.
Environmental Pollution
258
:
113828
. DOI: https://doi.org/10.1016/j.envpol.2019.113828.
Intergovernmental Panel on Climate Change
.
2021
.
Climate change 2021: The physical science basis
.
Cambridge, UK
:
Cambridge University Press
.
Jiang
,
L
,
Feng
,
Z
,
Dai
,
L
,
Shang
,
B
,
Paoletti
,
E
.
2018
.
Large variability in ambient ozone sensitivity across 19 ethylenediurea-treated Chinese cultivars of soybean is driven by total ascorbate
.
Journal of Environmental Sciences
64
:
10
22
. DOI: https://doi.org/10.1016/j.jes.2017.07.002.
Kinose
,
Y
,
Yamaguchi
,
M
,
Matsumura
,
H
,
Izuta
,
T
.
2020
.
Impact assessment of ozone absorbed through stomata on photosynthetic carbon dioxide uptake by Japanese deciduous forest trees: Implications for ozone mitigation policies
.
Forests
11
(
2
):
137
.
Kuang
,
Y
,
He
,
Y
,
Xu
,
W
,
Yuan
,
B
,
Zhang
,
G
,
Ma
,
Z
,
Wu
,
C
,
Wang
,
C
,
Wang
,
S
,
Zhang
,
S
,
Tao
,
J
,
Ma
,
N
,
Su
,
H
,
Cheng
,
Y
,
Shao
,
M
,
Sun
,
Y
.
2020
a.
Photochemical aqueous-phase reactions induce rapid daytime formation of oxygenated organic aerosol on the North China Plain
.
Environmental Science & Technology
54
(
7
):
3849
3860
. DOI: https://doi.org/10.1021/acs.est.9b06836.
Kuang
,
Y
,
Xu
,
W
,
Lin
,
W
,
Meng
,
Z
,
Zhao
,
H
,
Ren
,
S
,
Zhang
,
G
,
Liang
,
L
,
Xu
,
X
.
2020
b.
Explosive morning growth phenomena of NH3 on the North China Plain: Causes and potential impacts on aerosol formation
.
Environmental Pollution
257
:
113621
. DOI: https://doi.org/10.1016/j.envpol.2019.113621.
Lefohn
,
AS
,
Malley
,
CS
,
Smith
,
L
,
Wells
,
B
,
Hazucha
,
M
,
Simon
,
H
,
Naik
,
V
,
Mills
,
G
,
Schultz
,
MG
,
Paoletti
,
E
,
De Marco
,
A
,
Xu
,
X
,
Zhang
,
L
,
Wang
,
T
,
Neufeld
,
HS
,
Musselman
,
RC
,
Tarasick
,
D
,
Brauer
,
M
,
Feng
,
Z
,
Tang
,
H
,
Kobayashi
,
K
,
Sicard
,
P
,
Solberg
,
S
,
Gerosa
,
G
.
2018
.
Tropospheric ozone assessment report: Global ozone metrics for climate change, human health, and crop/ecosystem research
.
Elementa: Science of the Anthropocene
6
. DOI: https://doi.org/10.1525/elementa.279.
Li
,
K
,
Jacob
,
DJ
,
Liao
,
H
,
Shen
,
L
,
Zhang
,
Q
,
Bates
,
KH
.
2019
.
Anthropogenic drivers of 2013–2017 trends in summer surface ozone in China
.
Proceedings of the National Academy of Sciences of the United States of America
116
(
2
):
422
. DOI: https://doi.org/10.1073/pnas.1812168116.
Li
,
K
,
Jacob
,
DJ
,
Shen
,
L
,
Lu
,
X
,
De Smedt
,
I
,
Liao
,
H
.
2020
a.
Increases in surface ozone pollution in China from 2013 to 2019: Anthropogenic and meteorological influences
.
Atmospheric Chemistry and Physics
20
(
19
):
11423
11433
. DOI: https://doi.org/10.5194/acp-20-11423-2020.
Li
,
P
,
De Marco
,
A
,
Feng
,
Z
,
Anav
,
A
,
Zhou
,
D
,
Paoletti
,
E
.
2018
.
Nationwide ground-level ozone measurements in China suggest serious risks to forests
.
Environmental Pollution
237
:
803
813
. DOI: https://doi.org/10.1016/j.envpol.2017.11.002.
Li
,
P
,
Yin
R
,
Shang
,
B
,
Agathokleous
,
E
,
Zhou
,
H
,
Feng
,
Z
.
2020
b.
Interactive effects of ozone exposure and nitrogen addition on tree root traits and biomass allocation pattern: An experimental case study and a literature meta-analysis
.
Science of the Total Environment
710
:
136379
. DOI: https://doi.org/10.1016/j.scitotenv.2019.136379.
Lin
,
W
,
Xu
,
X
,
Ge
,
B
,
Zhang
,
X
.
2009
.
Characteristics of gaseous pollutants at Gucheng, a rural site southwest of Beijing
.
Journal of Geophysical Research: Atmospheres
114
(
D2
). DOI: https://doi.org/10.1029/2008JD010339.
Lin
,
W
,
Xu
,
X
,
Zhang
,
X
,
Tang
,
J
.
2008
.
Contributions of pollutants from North China Plain to surface ozone at the Shangdianzi GAW Station
.
Atmospheric Chemistry and Physics
8
(
19
):
5889
5898
. DOI: https://doi.org/10.5194/acp-8-5889-2008.
Lin
,
Y
,
Jiang
,
F
,
Zhao
,
J
,
Zhu
,
G
,
He
,
X
,
Ma
,
X
,
Li
,
S
,
Sabel
,
CE
,
Wang
,
H
.
2018
.
Impacts of O3 on premature mortality and crop yield loss across China
.
Atmospheric Environment
194
:
41
47
. DOI: https://doi.org/10.1016/j.atmosenv.2018.09.024.
Liu
,
N
,
Ma
,
J
,
Xu
,
W
,
Wang
,
Y
,
Pozzer
,
A
,
Lelieveld
,
J
.
2020
.
A modeling study of the regional representativeness of surface ozone variation at the WMO/GAW background stations in China
.
Atmospheric Environment
242
:
117672
. DOI: https://doi.org/10.1016/j.atmosenv.2020.117672.
Liu
,
S
,
Mo
,
X
,
Lin
,
Z
,
Xu
,
Y
,
Ji
,
J
,
Wen
,
G
,
Richey
,
JE
.
2010
.
Crop yield responses to climate change in the Huang-Huai-Hai Plain of China
.
Agricultural Water Management
97
(
8
):
1195
1209
. DOI: https://doi.org/10.1016/j.agwat.2010.03.001.
Liu
,
Y
,
Wang
,
T
.
2020
.
Worsening urban ozone pollution in China from 2013 to 2017—Part 1: The complex and varying roles of meteorology
.
Atmospheric Chemistry and Physics
20
(
11
):
6305
6321
. DOI: https://doi.org/10.5194/acp-20-6305-2020.
Lu
,
X
,
Ye
,
X
,
Zhou
,
M
,
Zhao
,
Y
,
Weng
,
H
,
Kong
,
H
,
Li
,
K
,
Gao
,
M
,
Zheng
,
B
,
Lin
,
J
,
Zhou
,
F
,
Zhang
,
Q
,
Wu
,
D
,
Zhang
,
L
,
Zhang
,
Y
.
2021
.
The underappreciated role of agricultural soil nitrogen oxide emissions in ozone pollution regulation in North China
.
Nature Communications
12
(
1
):
5021
. DOI: https://doi.org/10.1038/s41467-021-25147-9.
Lu
,
X
,
Zhang
,
L
,
Wang
,
X
,
Gao
,
M
,
Li
,
K
,
Zhang
,
Y
,
Yue
,
X
,
Zhang
,
Y
.
2020
.
Rapid increases in warm-season surface ozone and resulting health impact in China Since 2013
.
Environmental Science & Technology Letters
7
(
4
):
240
247
. DOI: https://doi.org/10.1021/acs.estlett.0c00171.
Ma
,
Z
,
Xu
,
J
,
Quan
,
W
,
Zhang
,
Z
,
Lin
,
W
,
Xu
,
X
.
2016
.
Significant increase of surface ozone at a rural site, north of eastern China
.
Atmospheric Chemistry and Physics
16
(
6
):
3969
3977
. DOI: https://doi.org/10.5194/acp-16-3969-2016.
Mills
,
G
,
Buse
,
A
,
Gimeno
,
B
,
Bermejo
,
V
,
Holland
,
M
,
Emberson
,
L
,
Pleijel
,
H
.
2007
.
A synthesis of AOT40-based response functions and critical levels of ozone for agricultural and horticultural crops
.
Atmospheric Environment
41
(
12
):
2630
2643
. DOI: https://doi.org/10.1016/j.atmosenv.2006.11.016.
Mills
,
G
,
Pleijel
,
H
,
Malley
,
CS
,
Sinha
,
B
,
Cooper
,
OR
,
Schultz
,
MG
,
Neufeld
,
HS
,
Simpson
,
D
,
Sharps
,
K
,
Feng
,
Z
,
Gerosa
,
G
,
Harmens
,
H
,
Kobayashi
,
K
,
Saxena
,
P
,
Paoletti
,
E
,
Sinha
,
V
,
Xu
,
X
.
2018
a.
Tropospheric ozone assessment report: Present-day tropospheric ozone distribution and trends relevant to vegetation
.
Elementa: Science of the Anthropocene
6
. DOI: https://doi.org/10.1525/elementa.302.
Mills
,
G
,
Sharps
,
K
,
Simpson
,
D
,
Pleijel
,
H
,
Broberg
,
M
,
Uddling
,
J
,
Jaramillo
,
F
,
Davies
,
WJ
,
Dentener
,
F
,
Van den Berg
,
M
,
Agrawal
,
M
,
Agrawal
,
SB
,
Ainsworth
,
EA
,
Büker
,
P
,
Emberson
,
L
,
Feng
,
Z
,
Harmens
,
H
,
Hayes
,
F
,
Kobayashi
,
K
,
Paoletti
,
E
,
Van Dingenen
,
R
.
2018
b.
Ozone pollution will compromise efforts to increase global wheat production
.
Global Change Biology
24
(
8
):
3560
3574
. DOI: https://doi.org/10.1111/gcb.14157.
Monks
,
PS
,
Archibald
,
AT
,
Colette
,
A
,
Cooper
,
O
,
Coyle
,
M
,
Derwent
,
R
,
Fowler
,
D
,
Granier
,
C
,
Law
,
KS
,
Mills
,
GE
,
Stevenson
,
DS
,
Tarasova
,
O
,
Thouret
,
V
,
von Schneidemesser
,
E
,
Sommariva
,
R
,
Wild
,
O
,
Williams
,
ML
.
2015
.
Tropospheric ozone and its precursors from the urban to the global scale from air quality to short-lived climate forcer
.
Atmospheric Chemistry and Physics
15
(
15
):
8889
8973
. DOI: https://doi.org/10.5194/acp-15-8889-2015.
National Bureau of Statistics of China
.
2019
.
China statistical yearbook 2019
.
Beijing, China
:
China Statistics Press
.
Paoletti
,
E
,
De Marco
,
A
,
Beddows
,
DCS
,
Harrison
,
RM
,
Manning
,
WJ
.
2014
.
Ozone levels in European and USA cities are increasing more than at rural sites, while peak values are decreasing
.
Environmental Pollution
192
:
295
299
. DOI: https://doi.org/10.1016/j.envpol.2014.04.040.
Peng
,
J
,
Shang
,
B
,
Xu
,
Y
,
Feng
,
Z
,
Pleijel
,
H
,
Calatayud
,
V
.
2019
.
Ozone exposure- and flux-yield response relationships for maize
.
Environmental Pollution
252
:
1
7
. DOI: https://doi.org/10.1016/j.envpol.2019.05.088.
Piikki
,
K
,
De Temmerman
,
L
,
Ojanperä
,
K
,
Danielsson
,
H
,
Pleijel
,
H
.
2008
.
The grain quality of spring wheat (Triticum aestivum L.) in relation to elevated ozone uptake and carbon dioxide exposure
.
European Journal of Agronomy
28
(
3
):
245
254
. DOI: https://doi.org/10.1016/j.eja.2007.07.004.
Pleijel
,
H
,
Broberg
,
MC
,
Uddling
,
J
.
2019
.
Ozone impact on wheat in Europe, Asia and North America – A comparison
.
Science of the Total Environment
664
:
908
914
. DOI: https://doi.org/10.1016/j.scitotenv.2019.02.089.
Ren
,
X
,
Shang
,
B
,
Feng
,
Z
,
Calatayud
,
V
.
2020
.
Yield and economic losses of winter wheat and rice due to ozone in the Yangtze River Delta during 2014–2019
.
Science of the Total Environment
745
:
140847
. DOI: https://doi.org/10.1016/j.scitotenv.2020.140847.
Sitch
,
S
,
Cox
,
PM
,
Collins
,
WJ
,
Huntingford
,
C
.
2007
.
Indirect radiative forcing of climate change through ozone effects on the land-carbon sink
.
Nature
448
(
7155
):
791
794
. DOI: https://doi.org/10.1038/nature06059.
Tai
,
APK
,
Martin
,
MV
,
Heald
,
CL
.
2014
.
Threat to future global food security from climate change and ozone air pollution
.
Nature Climate Change
4
(
9
):
817
821
. DOI: https://doi.org/10.1038/nclimate2317.
Tang
,
H
,
Takigawa
,
M
,
Liu
,
G
,
Zhu
,
J
,
Kobayashi
,
K
.
2013
.
A projection of ozone-induced wheat production loss in China and India for the years 2000 and 2020 with exposure-based and flux-based approaches
.
Global Change Biology
19
(
9
):
2739
2752
. DOI: https://doi.org/10.1111/gcb.12252.
Tao
,
Y
,
Huang
,
W
,
Huang
,
X
,
Zhong
,
L
,
Lu
,
S-E
,
Li
,
Y
,
Dai
,
L
,
Zhang
,
Y
,
Zhu
,
T
.
2012
.
Estimated acute effects of ambient ozone and nitrogen dioxide on mortality in the pearl river delta of Southern China
.
Environmental Health Perspectives
120
(
3
):
393
398
. DOI: https://doi.org/10.1289/ehp.1103715.
Van Dingenen
,
R
,
Dentener
,
FJ
,
Raes
,
F
,
Krol
,
MC
,
Emberson
,
L
,
Cofala
,
J
.
2009
.
The global impact of ozone on agricultural crop yields under current and future air quality legislation
.
Atmospheric Environment
43
(
3
):
604
618
. DOI: https://doi.org/10.1016/j.atmosenv.2008.10.033.
Wang
,
R
,
Bei
,
N
,
Wu
,
J
,
Li
,
X
,
Liu
,
S
,
Yu
,
J
,
Jiang
,
Q
,
Tie
,
X
,
Li
,
G
.
2022
.
Cropland nitrogen dioxide emissions and effects on the ozone pollution in the North China plain
.
Environmental Pollution
294
:
118617
. DOI: https://doi.org/10.1016/j.envpol.2021.118617.
Wang
,
T
,
Dai
,
J
,
Lam
,
KS
,
Nan Poon
,
C
,
Brasseur
,
GP
.
2019
.
Twenty-five years of lower tropospheric ozone observations in tropical East Asia: The influence of emissions and weather patterns
.
Geophysical Research Letters
46
(
20
):
11463
11470
. DOI: https://doi.org/10.1029/2019GL084459.
Wang
,
T
,
Zhang
,
L
,
Zhou
,
S
,
Zhang
,
T
,
Zhai
,
S
,
Yang
,
Z
,
Song
,
H
,
Wang
,
D
.
2021
.
Effects of ground-level ozone pollution on yield and economic losses of winter wheat in Henan, China
.
Atmospheric Environment
:
118654
. DOI: https://doi.org/10.1016/j.atmosenv.2021.118654.
Wang
,
X
,
Zhang
,
Q
,
Zheng
,
F
,
Zheng
,
Q
,
Yao
,
F
,
Chen
,
Z
,
Zhang
,
W
,
Hou
,
P
,
Feng
,
Z
,
Song
,
W
,
Feng
,
Z
,
Lu
,
F
.
2012
a.
Effects of elevated O3 concentration on winter wheat and rice yields in the Yangtze River Delta, China
.
Environmental Pollution
171
:
118
125
. DOI: https://doi.org/10.1016/j.envpol.2012.07.028.
Wang
,
Y
,
Konopka
,
P
,
Liu
,
Y
,
Chen
,
H
,
Müller
,
R
,
Ploeger
,
F
,
Riese
,
M
,
Cai
,
Z
,
,
D
.
2012
b.
Tropospheric ozone trend over Beijing from 2002–2010: Ozonesonde measurements and modeling analysis
.
Atmospheric Chemistry and Physics
12
(
18
):
8389
8399
. DOI: https://doi.org/10.5194/acp-12-8389-2012.
Xu
,
W
,
Kuang
,
Y
,
Liang
,
L
,
He
,
Y
,
Cheng
,
H
,
Bian
,
Y
,
Tao
,
J
,
Zhang
,
G
,
Zhao
,
P
,
Ma
,
N
,
Zhao
,
H
,
Zhou
,
G
,
Su
,
H
,
Cheng
,
Y
,
Xu
,
X
,
Shao
,
M
,
Sun
,
Y
.
2020
a.
Dust-dominated coarse particles as a medium for rapid secondary organic and inorganic aerosol formation in highly polluted air
.
Environmental Science & Technology
54
(
24
):
15710
15721
. DOI: https://doi.org/10.1021/acs.est.0c07243.
Xu
,
W
,
Kuang
,
Y
,
Zhao
,
C
,
Tao
,
J
,
Zhao
,
G
,
Bian
,
Y
,
Yang
,
W
,
Yu
,
Y
,
Shen
,
C
,
Liang
,
L
,
Zhang
,
G
,
Lin
,
W
,
Xu
,
X
.
2019
.
NH3-promoted hydrolysis of NO2 induces explosive growth in HONO
.
Atmospheric Chemistry and Physics
19
(
16
):
10557
-
10570
. DOI: https://doi.org/10.5194/acp-19-10557-2019.
Xu
,
X
.
2021
.
Recent advances in studies of ozone pollution and impacts in China: A short review
.
Current Opinion in Environmental Science & Health
19
:
100225
. DOI: https://doi.org/10.1016/j.coesh.2020.100225.
Xu
,
X
,
Lin
,
W
,
Xu
,
W
,
Jin
,
J
,
Wang
,
Y
,
Zhang
,
G
,
Zhang
,
X
,
Ma
,
Z
,
Dong
,
Y
,
Ma
,
Q
,
Yu
,
D
,
Li
,
Z
,
Wang
,
D
,
Zhao
,
H
.
2020
b.
Long-term changes of regional ozone in China: Implications for human health and ecosystem impacts
.
Elementa: Science of the Anthropocene
8
. DOI: https://doi.org/10.1525/elementa.409.
Yue
,
X
,
Unger
,
N
,
Harper
,
K
,
Xia
,
X
,
Liao
,
H
,
Zhu
,
T
,
Xiao
,
J
,
Feng
,
Z
,
Li
,
J
.
2017
.
Ozone and haze pollution weakens net primary productivity in China
.
Atmospheric Chemistry Physics
17
(
9
):
6073
6089
. DOI: https://doi.org/10.5194/acp-17-6073-2017.
Zhang
,
Q
,
Yuan
,
B
,
Shao
,
M
,
Wang
,
X
,
Lu
,
S
,
Lu
,
K
,
Wang
,
M
,
Chen
,
L
,
Chang
,
C-C
,
Liu
,
SC
.
2014
.
Variations of ground-level O3 and its precursors in Beijing in summertime between 2005 and 2011
.
Atmospheric Chemistry Physics
14
(
12
):
6089
6101
. DOI: https://doi.org/10.5194/acp-14-6089-2014.
Zhang
,
W
,
Feng
,
Z
,
Wang
,
X
,
Liu
,
X
,
Hu
,
E
.
2017
.
Quantification of ozone exposure- and stomatal uptake-yield response relationships for soybean in Northeast China
.
Science of the Total Environment
599–600
:
710
720
. DOI: https://doi.org/10.1016/j.scitotenv.2017.04.231.
Zhao
,
H
,
Zheng
,
Y
,
Wu
,
X
.
2018
.
Assessment of yield and economic losses for wheat and rice due to ground-level O3 exposure in the Yangtze River Delta, China
.
Atmospheric Environment
191
:
241
248
. DOI: https://doi.org/10.1016/j.atmosenv.2018.08.019.
Zhao
,
H
,
Zheng
,
Y
,
Zhang
,
Y
,
Li
,
T
.
2020
.
Evaluating the effects of surface O3 on three main food crops across China during 2015–2018
.
Environmental Pollution
258
:
113794
. DOI: https://doi.org/10.1016/j.envpol.2019.113794.
Zheng
,
B
,
Tong
,
D
,
Li
,
M
,
Liu
,
F
,
Hong
,
C
,
Geng
,
G
,
Li
,
H
,
Li
,
X
,
Peng
,
L
,
Qi
,
J
,
Yan
,
L
,
Zhang
,
Y
,
Zhao
,
H
,
Zheng
,
Y
,
He
,
K
,
Zhang
,
Q
.
2018
.
Trends in China's anthropogenic emissions since 2010 as the consequence of clean air actions
.
Atmospheric Chemistry and Physics
18
(
19
):
14095
14111
. DOI: https://doi.org/10.5194/acp-18-14095-2018.
Zhu
,
X
,
Feng
,
Z
,
Sun
,
T
,
Liu
,
X
,
Tang
,
H
,
Zhu
,
J
,
Guo
,
W
,
Kobayashi
,
K
.
2011
.
Effects of elevated ozone concentration on yield of four Chinese cultivars of winter wheat under fully open-air field conditions
.
Global Change Biology
17
(
8
):
2697
2706
. DOI: https://doi.org/10.1111/j.1365-2486.2011.02400.x.
Zhu
,
Z
,
Sun
,
X
,
Zhao
,
F
,
Meixner
,
FX
.
2015
.
Ozone concentrations, flux and potential effect on yield during wheat growth in the Northwest-Shandong Plain of China
.
Journal of Environmental Sciences
34
:
1
9
. DOI: https://doi.org/10.1016/j.jes.2014.12.022.
Ziemke
,
JR
,
Chandra
,
S
,
Duncan
,
BN
,
Froidevaux
,
L
,
Bhartia
,
PK
,
Levelt
,
PF
,
Waters
,
JW
.
2006
.
Tropospheric ozone determined from Aura OMI and MLS: Evaluation of measurements and comparison with the Global Modeling Initiative's Chemical Transport Model
.
Journal of Geophysical Research: Atmospheres
111
(
D19
). DOI: https://doi.org/10.1029/2006JD007089.

How to cite this article: Zhang, X, Xu, W, Zhang, G, Lin, W, Zhao, H, Ren, S, Zhou, G, Chen, J, Xu, X. 2022. Discrepancies in ozone levels and temporal variations between urban and rural North China Plain: Possible implications for agricultural impact assessment across China. Elementa: Science of the Anthropocene 10(1). DOI: https://doi.org/10.1525/elementa.2022.00019

Domain Editor-in-Chief: Detlev Helmig, Boulder AIR LLC, Boulder, CO, USA

Associate Editor: Samuel J. Oltmans, NOAA/ESRL Global Monitoring Division, CIRES, University of Colorado, Boulder, CO, USA

Knowledge Domain: Atmospheric Science

This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/.