Projection of future carbon benefits by photovoltaic power potential in China using CMIP6 statistical downscaling data

Solar photovoltaic (PV) systems is an affordable solution that significantly contribute to climate adaptation and resilience, energy security and greenhouse gas abatement with respect to fossil fuel electricity generation. Currently, available studies on the benefits of PV power generation only consider the electricity consumption and do not account for the possible future benefits from carbon trading under the combined impacts of pollution emissions and socio-economic. In this study, the downscaling and bias correction were applied to the Coupled Model Inter-comparison Project Phase 6 (CMIP6) multi-model mean data based on the historical data from the China Meteorological Administration (CMA) stations. The corrected measurements of meteorology were used to explore the PV power potential and the environmental and economic benefits offset by solar power generation under SSP126, SSP245 and SSP585 in China during 2023–2100. We found that the annual mean PV power potential across mainland China ranged from 1 to 37 Wm−2 and demonstrated a decreasing trend in the Northwest China and an increasing trend in the Southeast China. Compared to thermal power generation, electricity from solar energy will counteract the total emissions of annual mean 139.54 × 105 t CO2, 1702 t SO2, 2562 t NO X and 3710 t dust in China in SSP126 scenario. The results of variable importance assessment showed that the West Texas Intermediate crude oil price (47.77%), coal price (41.76%), natural gas price (6.65%) and gross domestic product (2.44%) contribute the most to the carbon emissions allowances (CEAs) price. Against a ‘carbon peak’ background in China, the CEA price will reach 80 CNY/t CO2 by 2030 in China, with the carbon trading value potential ranging from 20 billion to 200 billion CNY of each pixel (10 km × 10 km) by 2030. This study would have important implications for the environmental construction and future investment and construction of PV systems in China.


Introduction
The IPCC Sixth Assessment Report highlights the urgent need to limit climate warming to 1.5 • C due to the increasing greenhouse gas (GHG) emissions (Lynn andPeeva 2021, Forster et al 2023). At the 76th United Nations General Assembly, China strives to achieve a 'double carbon' target, with carbon emissions peaking in 2030 and 'carbon emitted' equal to 'carbon absorbed' in 2060 (Xu 2022). Therefore, China has to speed up its conversion to a green and sustainable development, in order to meet the international carbon reduction goals (Escap 2021, Zhang andChen 2022). PV power generation has almost 'zero emissions' during operation (Jäger-Waldau et al 2020), which is an easy and visible way to reduce emissions (Wolske et al 2017). Currently, China has the world's greatest installed PV capacity (Chen and Zhou 2023). With the China Carbon Emissions Trading Market officially went live for trading on July 16, 2021 (Yu 2022), PV power generation had also been demonstrated to have a major impact in electricity industry with high potentially economic benefits (Liu et al 2020). The Carbon Emissions Trading Market is a market mechanism adopted to drive GHG emissions reductions. While, carbon pricing can be used to quantitatively show the overall strength of policy actions needed to reduce emissions. Therefore, evaluating the PV power potential and the benefits it may bring is important to improve decision-making processes to achieve China's dual carbon goal as soon as possible.
Many studies have quantified the future PV power potential based on the CMIP5 data in China (Yang et al 2018b. However, compared with CMIP5, the CMIP6 has designed new scenario prediction experiments for various combinations of shared socio-economic pathway (SSPs) and representative concentration pathways (RCPs), containing substantial a large number of models and simulation data (Eyring et al 2016, Van Vuuren et al 2017. Currently, several studies have been evaluated China's future PV power potential using CMIP6 data. He et al (2023) found that the annual mean PV power potential in China during 2050-2069 for SSP126, SSP245 and SSP585 scenarios are 44.3, 43.6, and 43.3 TWyear −1 , respectively, using CMIP6 multi-model ensemble data. Niu et al (2023) quantified the PV power potential under 8 SSP-RCP scenarios and found that the PV power potential would decreases obviously in the SSP585 scenario from 2023 (192.71 Wm −2 ) to 2100 (189.96 Wm −2 ). In contrast, the PV power potential would increase (1. 36-5.90 Wm −2 ) in low mitigation pressure scenario. Although latest CMIP6 data has been used in several studies to examine the PV power potential, the resolution (spatial and temporal) of these studies were relatively coarse that applications at finer scales would be limited. Furthermore, previous studies had only considered the PV power potential and lacked an assessment of the carbon reduction emission and economic potential generated by PV power based on CMIP6 data.
Several studies have been conducted to examine the environmental and economic benefits based on the PV power potential in China. For example, Liu et al (2022) assessed the possible contributions of solar power to various green electric power systems in China to achieve carbon neutrality by 2050 using CEC (2019) emission factors. According to their findings, the green electric power system would offset 8.58-8.74 Gt yr −1 of CO 2 , 2.04-2.08 Mt yr −1 of SO 2 and 1.94-1.97 Mt yr −1 of NO X . Yang et al (2018a) evaluated the benefits of PV power generation using the government's target of 400 GW of installed capacity by 2030. They found that total national CO 2 emissions decreased by 4.2% compared to the coal-intensive base case in 2030 China. However, there was limited analysis to date on the economic and environmental benefits based on the PV power potential in China using CMIP6 data. Meanwhile, the combined impacts of pollution emissions and socio-economic on the benefits generated by PV under different CMIP6 scenarios was also need to be deeply investigated.
This study is organized as follows. First, we introduce a Delta downscaling method to correct CMIP6 multi-model mean data (2016) (surface downward shortwave radiation, surface diffuse downward shortwave radiation, ambient temperature and surface air pressure) for a finer resolution (hourly, 0.1 • × 0.1 • ) under SSP126, SSP245 and SSP585 scenarios. Second, the downscaled data is used to analyze the variations (from 2023 to 2100) in PV power potential across mainland China. Third, we discuss the factors affecting the CEA price based on the random forest (RF) algorithm in terms of macroeconomic factors, energy factors, and environmental factors. Finally, we project the CEA price using multiple linear regression (MLR) method against a 'carbon peak' background in China, and explore the environmental benefits and carbon trading value potential. This study would provide suggestions for reducing emissions from China's power sector and improving the carbon trade market systems.

Ground-based daily observations
Daily surface solar radiation and diffuse solar radiation measurements during 2010-2014 at 17 China Meteorological Administration (CMA) stations across mainland China were used for the performance validation of CMIP6 statistical downscaling data. The spatial distribution of the 17 selected CMA stations is shown in figure 1. Table S1 in    Dutta et al 2022). The SSP585 scenario is based on the SSP5 scenario with the same radiative forcing as RCP8.5 of 8.5 Wm −2 . The rapid and intensive development pattern of human society will produce triple GHG emissions in this century, which is the highest of all SSP-RCP scenarios. The SSP245 scenario simulation is based on the SSP2 scenario with the upper limit of radiative forcing set at 4.5 Wm −2 . Future climate change will be represent under non-extreme land use and aerosol scenario. The SSP126 scenario was developed from the IMAGE3.0 (Doelman et al 2018) integrated assessment model with the SSP1 scenario as the baseline setting. SSP1 scenario describes an environmentally focused world with significant investments in education and health, a rapidly growing economy, and a pathway for a transition to sustainable development. In this study, four climate variables (surface downward shortwave radiation (RSDS), surface diffuse downward shortwave radiation (RSDSDIFF), ambient temperature (TAS) and surface air pressure (PS)) and three typical climate change scenarios (SSP585, SSP245 and SSP126) from 11 CMIP6 climate models were used to project and analyze the PV power potential (more details see table S2).

ERA5 reanalysis data
The fifth generation of European Reanalysis (ERA5) data is the current largest global reanalysis product proposed by the European Centre for Medium-Range Weather Forecasts (ECMWF) (Hersbach et al 2020, Jiang et al 2020. The land component of the fifth generation of European Reanalysis (ERA5-Land) data (2019) provides accurate descriptions of global and regional surface climate characteristics over several decades with higher resolution (hourly, 0.1 • × 0.1 • ), covering the period 1950-present. In this study, total sky direct solar radiation at surface (hourly, 0.25 • × 0.25 • ) provided by ERA5 (2018), and surface solar radiation downwards (hourly, 0.1 • × 0.1 • ), surface pressure and 2 m temperature provided by ERA5-Land were used to downscale and bias correct the CMIP6 data.

Carbon emissions allowances (CEAs) prices projection
Energy prices (Karstensen and Peters 2018), gross domestic product (GDP)  and total carbon emissions (Green 2021) play an important role in predicting the CEA prices. The West Texas Intermediate (WTI) crude oil has become the worldwide standard for crude oil pricing due to the global military as well as economic power of the United States (Su et al 2020). Eviews (Econometric Views) software is one of the most popular econometric software for the rapid creation, estimation, testing and application of econometric models of the quantitative laws of socioeconomic relations and economic activities (Xie 2023). In this study, quarterly series of China's GDP were converted to daily series using the quadratic-match sum method in the frequency conversion tool in Eviews 10 (Junaedi et al 2022, Wafa 2022. In addition, we also considered the price of solar cells in China. Here, we collected daily realistic data and constructed the relationship between CEA prices and daily China coal price, WTI crude oil price, China natural gas price and China's GDP from August 2021 to August 2022. According to Tsinghua University's projection of China's GDP growth rate, China's GDP will reach 170 trillion CNY by 2030. While, US energy information administration (EIA) projected that WTI crude oil price will reach $92.98 per barrel in 2030. The carbonneutral targets set by multiple countries around the world will play a key role in reducing demand for fossil fuels, and coal prices will trend downward as a result. While, it is estimated that the average price of power coal will remain essentially unchanged during the period 2021-2030. In addition, we estimated the price of natural gas in China in 2030 based on the daily growth rate of natural gas over the historical period. Table S3 shows the information of the data to train the MLR model and to predict the CEA price in 2030 in China.

Potential of photovoltaic (PV) power assessment model
The PV power potential was estimated based on the downscaled CMIP6 hourly data (RSDS, RSDSDIFF, TAS and PS) with a finer resolution (hourly, 0.1 • × 0.1 • ) in China (2023-2100). The method we used to evaluate the PV power potential was shown in the supplementary materials (Note S1). Due to the complexity of solar system tracking and the widespread use of fixed tilt systems, we selected the fixedtilt system. Considering the solar azimuth and altitude, the ground reflectance, RSDSDIFF, RSDS and the surface direct downward shortwave radiation (RSDSDIR) and their conversion factors we derived the resource potential (solar radiation reaching the PV panels). Polysilicon panels are currently the most widely used, with peak power per unit area reaching around 200 Wm −2 under standard condition. Conservatively estimated, we adopt PV panel size of 1.65 m × 0.992 m, panel normal working temperature of 44 • C, temperature coefficient of −0.41% • C −1 , panel power of 265 W, system efficiency of 80% of the PV modules. Temperature correction coefficient (TCF) was used to quantify the impact of temperature on PV power potential. Finally, we evaluated the technical potential of PV power, taking into account the packing factor (PF), TCF and the system efficiency, etc.

Statistical downscaling
It is difficult for CMIP6 data to explicitly represent variations in climate scenarios at a finer scale due to its coarse spatial resolution. In this study, the Delta downscaling method (the change factor method of bias correction) is used to correct the CMIP6 data from 2023 to 2100. The change factor represent the difference between the simulations from CMIP6 and observations from 17 stations in the historical period (2010)(2011)(2012)(2013). The downscaled CMIP6 multi-model ensemble data can be calculated as follows: where P_fut and P_his represent the CMIP6 data for the future and historical period, respectively; P_era represents the multi-year mean value of the ERA5 data in the historical period; E represents the ratio factor of model future to historical period.

Accuracy evaluation indicators
In this study, 15 statistical indexes, including mean bias error (MBD), root mean square difference (RMSD), the mean absolute difference (MAD), the standard deviation of the residual (SD), uncertainty at 95% (U95), t-statistic (TS), correlation coefficient (R), slope of the best-fit line (SBF), Nash-Sutcliffe's efficiency (NSE), Will-Mott's index of agreement (WIA), Legates' coefficient of efficiency (LCE), the Kolmogorov-Smirnov test integral (KSI), the critical limit over-estimation (OVER), the combined performance index (CPI) and the relative root mean square error (RRMSE) are selected to assess the performance of the CMIP6 data (RSDS and RSDSDIFF) after statistical downscaling. In addition, the global performance indicator (GPI) is chosen to comprehensively evaluate the accuracy of the CMIP6 data after statistical downscaling in 2014. The formulae for calculating the different indicators are shown in supplementary materials (Note S2).

Variable importance assessment
Scikit-learn was used in this study to evaluate the importance of variables (GDP, coal price, natural gas price, WTI crude oil price, total carbon emissions, solar cells price and polysilicon price) by averaging the contributions made by each variable in each tree of the RF and then comparing the magnitude of contributions between variables. RF is an ensemble learning algorithm with decision trees as the base learners, and each decision tree is a set of internal nodes and leaves. Here, we used GridSearchCV to tune the mainly hyperparameters of RF such as n_estimators (50), min_samples_leaf (2) and oob_score (True). In this study, the Gini index (GI) is chosen for evaluation, assuming that there are J features W 1 , W 2 , W 3 , . . . , W J , I decision trees and C categories, and the variable importance measures of each feature W J i.e., VIM (Gini) j is to be calculated as follows: where P qc denotes the proportion of category c in tree node q.

Accuracy assessment of the multi-model mean downscaled data
In this study, the results of 16 accuracy indexes were calculated to compare the performance of each CMIP6 multi-model mean downscaled data (RSDS and RSDSDIFF) in the historical scenario in 2014 (see

Changes on PV power potential
Mann-Kendall (MK) (Gumus et al 2022) is a method to determine monostich trend changes (increase or decrease) in a set of data over a time series. Sen's slope  estimation is a method of calculating a robust trend with nonparametric statistics. In this study, we used the MK algorithm and Sen's slope to examine the approximate trend of PV power potential. Figure 2 illustrates the spatial distribution (figure 2(a)) and the spatiotemporal trends (figures 2(b) and (c)) of annual mean China's PV power potential, with the black dot-covered area in figure 2(b) denoting the significant trend with 95 confidence intervals. It is indicated that the PV power potential across mainland China is at a range of 1-37 Wm −2 by taking the PF and temperature correction coefficient into consideration. Overall, the spatial distribution of PV power potential in China is heterogeneous, influenced by latitude, altitude, topography, duration, atmospheric conditions, etc (Qin et al 2018, Wu et al 2022. Higher mean values of PV power potential mainly located in the QTP (>27 Wm −2 ) and YPS (>21 Wm −2 ) due to the weak radiation attenuation there. In contrast, the RSDS in NEC is much weakened by the atmosphere due to high latitude and low altitude, so that the PV power potential is the lowest there (<6 Wm −2 ). In addition, the enclosed basin topography leads to a large accumulation of pollutants and water vapor in the air, which has a strong scattering and absorbing effect on the solar radiation reaching the surface, resulting in low PV power potential in SBS (<15 Wm −2 ). As illustrated in figures 2(b) and (c), Southeast China exhibits an increasing trend with higher Sen's slope and MK values, whereas Northwest China exhibits declining trends. In SSP126 scenario, significant increasing trend are found in Wuhan (Sen's slope = 0.018 Wm −2 yr −1 , MK = 2.92) and Anhui (Sen's slope = 0.017 Wm −2 yr −1 , MK = 2.54) province, while significant decreasing trend is found in Xinjiang province with the mean Sen's slope and MK values of −0.006 Wm −2 yr −1 and −0.96. The PV power potential has substantial increasing trend (mean Sen's slope and MK values >0.018 Wm −2 yr −1 and >3.6) in the Southeast China under the SSP245 scenario.
Against a 'carbon peak' background in China, we analyzed variations in PV power potential (see figure 3) over time slices of the future (2023-2030, 2030-2060 and 2060-2100). It is indicated that the annual mean PV power potential in China is at a range of 16.8-18.2 Wm −2 (161.28 TW-174.72 TW and 14.13 × 10 5 TWH-15.31 × 10 5 TWH), 16.0-17.4 Wm −2 (153.6 TW-167.04 TW and 13.46 × 10 5 TWH-14.63 × 10 5 TWH) and 16.4-17.2 Wm −2 (157.44 TW-165.12 TW and 13.79 × 10 5 TWH-14.46 × 10 5 TWH) in 2023-2100 under SSP126, SSP245 and SSP585 scenario, respectively. The SSP126 scenario describes an environmentally conscious world with increased renewable energy usage on top of carbon extraction and consumption, in which the trends of socio-economic growth and climate policies will lead an overall uptrend (+0.002 Wm −2 yr −1 ) of PV power potential, with a marked uptrend (+0.037 Wm −2 yr −1 ) in 2023-2030. In SSP245 scenario, China is committed to achieving the sustainable development goals, but progress is slow. Compared to the SSP126 scenario, the PV power potential fluctuates more moderately, showing a trend of decreasing (−0.028 Wm −2 yr −1 ) in 2023-2030, then increasing (+0.004 Wm −2 yr −1 ) in 2030-2060, and finally decreasing (−0.002 Wm −2 yr −1 ) in 2060-2100. The SSP585 scenario simulates the development of abundant fossil fuel resources to drive economic and social development. High carbon emissions from rapid economic development will lead to higher temperatures (Wild et al 2015), which will lead to a reduction in the overall potential for PV power.

Environmental benefit assessment
In 2021, 6000 KW and above thermal power plants in China will consume 301.5 g of coal as standard for power supply, and emit 0.022 g of dust, 0.101 g of SO 2 , 0.152 g of NO X , and 828 g of CO 2 per unit of thermal power generation, respectively (CEC 2022). Based on the above, we evaluated CO 2 and pollutant emission reductions in China using a method that multiplies emission factors with PV power potential in different geographic regions for three scenarios. Figure 4 illustrates the total annual mean environmental benefit potential of PV power during 2023-2100 across mainland China in SSP126, SSP245 and SSP585 scenarios. It is indicated that the QTP with its highest levels of PV power potential will become the region with the highest potential for environmental benefits in China, followed by NWC and IM. Electricity from solar energy in the QTP will offset annual mean 543.37 × 10 5 t CO 2 , 6680 t SO 2 , 9975 t NO X and 1444 t dust in SSP126 scenario, 518.10 × 10 5 t CO 2 , 6320 t SO 2 , 9511 t NO X and 1377 t dust in SSP245 scenario and 524.38 × 10 5 t CO 2 , 6397 t SO 2 , 9626 t NO X and 1393 t dust in SSP585 scenario, respectively. Meanwhile, lower potential of the environmental benefits are found in NEC, SBS and MYP mainly because of the lower PV power potential there. Additionally, annual mean carbon reduction potential of the SSP126 scenario (139.54 × 10 5 t) is higher than both of the SSP245 (133.96 × 10 5 t) and SSP585 (134.66 × 10 5 t) scenarios in China due to low vulnerability, mitigation pressure and radiation forcing.
Moreover, we compared the total annual mean carbon emission reductions and environmental benefits of each province in China in SSP126, SSP245 and SSP585 scenarios during 2023-2100. As shown in figure 5, total annual mean reductions of CO 2 , SO 2 , NO X and dust of each province range from 0.66 to 331.71 × 10 5 t, 8.08 to 4046.24 t, 12.15 to 6089.39 t and 1.76 to 881.36 t, respectively. Obviously, there is significant heterogeneity in the spatial distribution of carbon emission reductions and environmental benefits of each province in China. Higher environmental benefits and carbon emission reductions are mainly found in Xinjiang and Tibet in western China, which will produce the benefits of >236.5 × 10 5 t CO 2 , >2885 t SO 2 , >4340 t NO X and >628 t dust, respectively. By contrast, benefits in the province in southeast and northeast China are generally low, especially in Shanghai and Beijing, with the benefits of <1.83 × 10 5 t CO 2 , <22.27 t SO 2 , <33.52 t NO X and <4.85 t dust, respectively. However, there is a mismatch between the electricity demand and PV power potential in most provinces across China. It is still a challenge to exploit the high carbon emission reduction as well as environmental benefit potential and the realization of power transmission between provinces.

Carbon trading value potential
The RF algorithm and the GI was used to assess the contribution of each input variable to the CEA price from August 2021 to August 2022. The importance percentage results of 7 input variables of the machine learning model are shown in figure 6. The highest contribution of CEA price is WTI crude oil price with 47.77%, followed by the coal price with 41.76%. The contribution of natural gas price in energy prices is relatively low, at 6.65%. Although China is vigorously  developing renewable energy industries, the demand for traditional energy sources such as coal and oil remains high and usage dominates. Fluctuating prices in energy market affect energy demand and carbon allowances demand, which ultimately has a shock on carbon market supply and leads to subsequent fluctuations in CEA prices. In addition, GDP also affects the CEA price in some degree, with a contribution of 2.44%. Total carbon emissions and solar cells price are weakly correlated with CEA price with a total contribution of 1.39%. Finally, WTI crude oil price, coal price, natural gas price and GDP are selected to train the MLR model and to predict the CEA price in 2030 over China.
In this study, we selected the train_test_split function to divide the sample data into a training set and testing set in the ratio of 7:3. The performance of the MLR model in predicting carbon price on the training and testing sets was shown in the table 1.
(4)  Here, we focus on the year of 2030. The China Carbon Emissions Trading Market officially goes online in 2021. Considering the limited amount of CEA price data available for training in this study, the short-term prediction results are more trustworthy. As a result, we will concentrate on forecasting the CEA price in 2030, which is more relevant to the predicted timing of carbon-peak. Results indicate that China's CEA price will reach 80 CNY/t CO 2 in 2030.
Next, we examined the economic potential (carbon trading value potential) of each pixel (10 km × 10 km) across China's provinces in 2030 using a method that multiplies carbon emission reduction with CEA price (figure 7). The results show that the carbon trading value potential across mainland China is ranging from 20 billion to 200 billion CNY per pixel. The Tibet province has the highest annual mean carbon trading value potential (>175 billion CNY) because of the highest PV power potential there, followed by Qinghai (>167 billion CNY) and Xinjiang (>150 billion CNY). In addition, lower values of the carbon trading value potential are found in three Northeastern provinces (Liaoning, Heilongjiang and Jilin), which are not exceeding 60 billion CNY.

Discussion and conclusions
In this study, we evaluate the PV power potential and the carbon benefits produced by solar power generation based on the downscaled and bias-corrected data (hourly, 10 km × 10 km) under SSP126, SSP245 and SSP585 scenarios in China from 2023 to 2100. The annual mean PV power potential across mainland China demonstrates a significantly decreasing trend in the NWC and increasing trend in the Southeast China, with an obvious spatial mismatch between population distribution and power demand, making cross-regional power deployment essential. In addition, changes of PV power potential increase in SSP126 scenario, while decrease in SSP585 scenario with higher carbon emissions and mitigation challenges. We also demonstrate an overall slightly decreasing trend of −0.0004 Wm −2 yr −1 (from 2023 to 2100) in PV power potential across China, which is in agreement with the findings of Ha et al (CMIP6), Lu et al (2022a) (CMIP6) and Niu et al (CMIP6). The findings are different from the conclusions of Crook et al (2011) (CMIP3) and Zou et al (CMIP5) due to the extra accounting for the ensemble impacts of pollution emissions and socio-economic in the CMIP6 data. In addition, considering the suitable land area (993 000 km 2 ) for PV power generation in 2015 (Qiu et al 2022), the total annual PV power potential during 2023-2100 across mainland China is 152.5, 146.3 and 147.09 PWH under SSP126, SSP245 and SSP585 scenarios, respectively. While, Qiu et al found that the total PV power potential was 131.9 PWH in 2015.
In the future, we will explore the land suitability of the PV power potential and uncover where the PV potential lies in China from 2023 to 2100 based on CMIP6 data against the global development trend of 'Photovoltaic + Energy Storage' .
The negative economic externalities caused by the high carbon emission characteristics of the traditional generation industry are gradually expanding due to its diverse and highly dependent downstream industries, and the introduction of the market mechanism of carbon trading market mechanisms is an important way to solve this problem. The results of variable importance assessment show that the WTI crude oil price (47.77%), coal price (41.76%), natural gas price (6.65%) and GDP (2.44%) contribute the most to the CEA price. There is a certain correlation between traditional energy prices and CEA prices, and the correlation between them in China will become closer as the carbon market construction is improved and the financial market becomes more mature. Therefore, it is necessary to emphasize the regulation of traditional energy market and improve the construction of traditional energy market. For example, it is possible to improve the traditional energy security early warning system and the operation mechanism of the energy market to grasp the source of carbon price fluctuations, so as to guarantee the smooth operation of the carbon trading market. In addition, based on their linearity, we found that the CEA price will reach 80 CNY/t CO 2 by 2030 in China, with an economic benefit potential in a range of 20 billion to 200 billion CNY of each pixel. The carbon emission reductions from PV are equivalent to providing additional carbon allowances, however, the current regulations stipulate that the offset ratio is no more than 5% of the carbon emission allowance. Due to the fact that China's economy is still in a period of medium-to high-speed development, and that the development of regions and industries is unbalanced, the total amount of allowances in the national carbon market is not set at a fixed value, but rather is set through a combination of a 'bottom-up' and 'top-down' approach, resulting in a total amount that has a certain degree of flexibility. Based on the total amount of carbon allowances of 50.89 million tons in 2022, the PV power potential can offset a maximum of 2.54 million tons of total carbon emissions, generating an economic benefit of about 200 million CNY across China in 2030.
The sustainability scenario (SSP126) simulates a total change of temperature controlled within 2 • C relative to the 1986-2005 period (You et al 2021). With respect to the electricity from thermal power plants of 6000 KW and above, PV power potential in China will counteract annual mean 139.54 × 10 5 t CO 2 , 1702 t SO 2 , 2562 t NO X and 3710 t dust in SSP126 scenario by the latest emission factors (CEC 2022). Based on the emission factors from CEC (2019) Moreover, long-term projection of the CEA price will be evaluated.