Projected Regional 1.50°C and 2.00°C Warming Threshold-crossing Time Worldwide Using the CMIP6 Models

MENG Yali,DUAN Keqin,SHANG Wei,SHI Peihong,LI Shuangshuang,CHENG Ying,CHEN Rong,ZHANG Zhaopeng

(School of Geography and Tourism,Shaanxi Normal University,Xi’ an 710119,China)

Abstract: The Paris Agreement aims to limit global warming to well below 2.00°C and pursue efforts to limit the temperature increase to 1.50°C.However,the response of climate change to unbalanced global warming is affected by spatial and temporal sensitivities.To better understand the regional warming response to global warming at 1.50°C and 2.00°C,we detected the 1.50°C and 2.00°C warming threshold-crossing time (WTT) above pre-industrial levels globally using the Coupled Model Intercomparison Project phase 6 (CMIP6)models.Our findings indicate that the 1.50°C or 2.00°C WTT differs substantially worldwide.The warming rate of land would be approximately 1.35-1.46 times that of the ocean between 60°N-60°S in 2015-2100.Consequently,the land would experience a 1.50°C(2.00°C) warming at least 10-20 yr earlier than the time when the global mean near-surface air temperature reaches 1.50°C (2.00°C)WTT.Meanwhile,the Southern Ocean between 0° and 60°S considerably slows down the global 1.50°C and 2.00°C WTT.In 2040-2060,over 98.70% (77.50%),99.70% (89.30%),99.80% (93.40%),and 100.00% (98.00%) of the land will have warmed by over 1.50°C (2.00°C) under SSP (Shared Socioeconomic Pathway) 1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.We conclude that regional 1.50°C (2.00°C) WTT should be fully considered,especially in vulnerable high-latitude and high-altitude regions.

Keywords: CMIP6 (Coupled Model Intercomparison Project phase 6);global warming;1.50°C warming time;2.00°C warming time;regional differences

1 Introduction

In 2015,the Paris Climate Agreement aimed at limiting the warming of the global mean near-surface air temperature (SAT) within 2.00°C or,even better,within 1.50°C above pre-industrial levels (UNFCCC,2015).In 2020,the 26th Conference of the Parties (COP26) emphasized the 1.50°C warming target (Allan et al.,2021).However,the global mean SAT was approximately 1.26°C warmer in 2020,relative to the pre-industrial(1850-1900) levels,leaving a gap of only 0.24°C to remain within the 1.50°C warming target of Intergovernmental Panel on Climate Change (IPCC),2021).The latest Sixth Assessment Report (AR6) from IPCC noted that limiting global warming to 1.50°C is impossible unless immediate measures are adopted to reduce emissions (IPCC,2022).Therefore,with the emergence of the concepts of carbon peak and carbon neutralization,the time when global warming will reach 1.50°C and 2.00°C has been investigated for future climate impacts and adaptations (Kraaijenbrink et al.,2017;Lehner et al.,2017;Saeed et al.,2021).Thus,to evaluate warmingrelated impacts and risks on a regional scale,it is imperative to recognize the regional occurrence time of 1.50°C and 2.00°C warming.

The IPCC AR6 showed that the 1.50°C (2.00°C)warming threshold-crossing time (WTT) of the global mean SAT would be in～2033 (never),～2031 (2053),～2031 (2047),and～2028 (2042) under the Shared Socioeconomic Pathways (SSPs) of SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5 (indicating a radiative forcing stabilizing at～2.6,～4.5,～7.0,and～8.5 W/m2,respectively,in 2100) in the Coupled Model Intercomparison Project phase 6 (CMIP6),respectively (IPCC,2021).However,initial studies showed that the 1.50°C(2.00°C) WTT would occur in 2036 (never),2028(2049),2033 (2056),and 2025 (2039) under the Representative Concentration Pathways (RCPs) of RCP2.6,RCP4.5,RCP6.0,and RCP8.5 (indicating a radiative forcing stabilizing at～2.6,～4.5,～6.0,and～8.5 W/m2,respectively,in 2100) in CMIP5,respectively (Hu et al.,2017).Previous studies have indicated that the CMIP6 models are more reliable for simulating future climate trends by considering both the model improvements and different SSPs (Eyring et al.,2016;Gusain et al.,2020).Therefore,the 1.50°C and 2.00°C WTT are more precise in CMIP6 than in CMIP5.Moreover,the warming trends show large discrepancies worldwide,as the response of climate change to greenhouse gas-induced warming is spatially and temporally sensitive (Piao et al.,2014).Further evidence indicates that regional unbalanced warming has caused more severe regional extreme disasters.For instance,as the global mean SAT warming reached 2.00°C,the drylands could experience warming of up to 3.20-4.00°C and significant drought risk (Huang et al.,2017;Lehner et al.,2017).Southwest Asia is expected to exceed the temperature threshold for human adaptability in the future,severely affecting human habitability (Pal and Eltair,2016).In particular,warming over the Arctic and Tibetan Plateau has recently intensified,causing accelerated cryospheric degradation and ecosystem disturbances (Duan et al.,2017;You et al.,2021;Meng et al.,2022).The Antarctic sea ice extent hit a new record low in 2022 (Wang et al.,2022).Furthermore,climate change will exacerbate global income inequality (Burke et al.,2015;Mitchell et al.,2016).Therefore,to further clarify the differences in regional warming,it is imperative to clarify the regional warming trend and the 1.50°C and 2.00°C WTT above pre-industrial levels (1850-1900).

Although the global mean SAT WTT at 1.50°C and 2.00°C in previous studies have been evaluated (Zhou et al.,2018;Aihaiti et al.,2021;Gao et al.,2021;Zhu et al.,2021;Huang et al.,2022),the regional 1.50°C and 2.00°C WTT remain unclear,especially in the light of results of the latest CMIP6 models.Thus,in this study,we aimed to evaluate the regional 1.50°C and 2.00°C WTT above pre-industrial levels using the latest CMIP6 model outputs to better understand global warming and formulate regional carbon emission policies.Our research provides a basis for further reducing regional warming differences,even the resulting differences in economic development,as well as reducing the occurrence of regional disasters.

2 Materials and Methods

2.1 Data

We selected the monthly SAT data of 24 models released in CMIP6 (https://esgf-node.llnl.gov/projects/cmip6/),which were from the ScenarioMIP.These models are widely used,have good simulation capabilities,and include complete historical and future scenarios (Zhu et al.,2021;Huang et al.,2022;Meng et al.,2022).The historical period was from 1850 to 2014,and the future period was from 2015 to 2100.For the future period,we analyzed four SSP scenarios: SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5 (O’Neill et al.,2016;Yang et al.,2021;Zhang et al.,2021).The four SSP scenarios represent different shared socioeconomic pathways of sustainable development,the middle path,regional competition,and fossil fuel development with different combinations of low,medium,medium-to-high,and high levels of radiative forcing.Basic information on the models is presented in Table 1.To systematize the resolution,all models were interpolated to a 0.5° × 0.5°grid using bilinear interpolation.

Table 1 Basic information of 24 models in the Coupled Model Intercomparison Project phase 6

Monthly temperature data from HadCRUT5,a collaborative product of the Met Office Hadley Centre and the Climatic Research Unit at the University of East Anglia(Morice et al.,2021),were used as observational data to evaluate the temporal simulation skills of the CMIP6 models in the historical period 1979-2014.We also used the monthly mean 2-m temperature data from the ERA5 re-analysis (Hersbach et al.,2020) as reference data to evaluate the model simulation skills for the same period.

2.2 Area weight processing method

To calculate the global mean temperature and avoid area deformation with latitude,24 models with different resolutions were bilinearly interpolated onto a common grid of 0.5° × 0.5°.Subsequently,we used the areaweighted method to examine the global mean temperature in both CMIP6 and ERA5 as follows:

whereWdenotes the area weight of each grid point.Lat1andLat2represent the latitudes of the upper and lower boundaries (north and south) of each grid;Lon1andLon2represent the longitudes of the left and right boundaries,respectively.π is the constant (3.141 592 6),andRis the Earth’s radius (using mean radiusR=6371 km).Subsequently,the SAT of each grid was multiplied by the area weight to obtain the new grid temperature data.

2.3 Model evaluation method

A Taylor diagram was used to evaluate the historical(1979-2014) simulation skills of the 24 CMIP6 models and their multi-model ensemble mean (MME) results(Taylor,2001).The Taylor diagram shows the correlation coefficient,root mean square error (RMSE),and standard deviation between the simulation and observations.All RMSE and standard deviations were normalized.The REF represents the reference observation data HadCRUT5 and ERA5.The correlation coefficient and normalized standard deviation were close to 1 and the normalized RMSE was close to 0,indicating the best model simulation results.In other words,the closer the model is to REF,the better the simulation ability is.

2.4 Uncertainty assessment

Based on the global SAT anomaly time series of each model under different scenarios,the proportion of models that reach the SAT warming threshold of 1.50°C and 2.00°C each year was obtained by a probability distribution function to describe the uncertainty of model projections.

The Signal-to-noise ratio (SNR) was used to analyze the spatial uncertainty of model projections (Zhou and Yu,2006;Wu et al.,2019),which is defined as:

where σ indicates the SNR,xiis the warming result of a single model at the time when the MME is estimated to reach the global warming threshold.is the warming result predicted by the MME,andnis the number of models.This index reflects the relationship by signal and noise between the MME set and the standard deviation between models and then reflects the reliability of future predictions.σ >1 indicates that the signal is greater than the noise,and thus the future prediction results are more reliable.A greater σ indicates a higher credibility.

2.5 Definition of 1.50°C and 2.00°C WTT

The time series of the global mean near-surface air temperature change from 24 CMIP6 models and their MME relative to 1850-1900 were first smoothed by the nineyear running mean to remove interannual variability under the four SSP scenarios.Then the 1.50°C and 2.00°C WTT were defined as the first year when the nine-year running mean temperature results exceed pre-industrial(1850-1900) levels by 1.50°C or 2.00°C.For example,2019 represents the nine-year running mean from 2015 to 2023.

3 Results

3.1 Model evaluation results

Based on the ERA5 and HadCRUT5,we evaluated the monthly SAT simulation abilities of 24 CMIP6 models and their MME results (Fig.1).The temporal correlation coefficients between CMIP6 and HadCRUT5 exceeded 0.6,which is similar to those between CMIP6 and ERA5.The spatial correlation coefficients between the CMIP6 models and ERA5 exceeded 0.95.Differences in the standardized standard deviation and RMSE were observed among all models.We also considered the simulation bias of the annual mean SAT from 1979 to 2014 among the models.Based on the overall model rankings of all indicators,the MME showed better agreement with HadCRUT5 and ERA5 than any single model during the overlap period.Thus,the MME is the best candidate to project the 1.50°C and 2.00°C WTT(Cai et al.,2021;Huang et al.,2022).

A good friend of mine was going away on a long trip during the fall. Miriam thought she had given herself plenty of time to do all the things that are required when one goes out of town.

Fig.1 Taylor diagrams for annual mean near-surface air temperature over the globe between Coupled Model Intercomparison Project phase 6 models and (a) HadCRUT5,(b) ERA5: temporal scale,and (c) ERA5: spatial scale for temporal 1979-2014.REF,referenced observation data;MME,multi-model ensemble;RMSE,root mean square error

3.2 Global 1.50°C and 2.00°C WTT

As shown in Fig.2,the global mean SAT showed an increasing trend from 1850 to 2100 under all 24 models and different scenarios,except for a slightly decreasing trend in the late 21st century under SSP1-2.6.The warming rates were low before 2000 and gradually increased thereafter.The interdecadal warming trend increased with an increase in the emission scenarios from 2015 to 2100.The discrepancies in the global mean SAT warming trend were projected by the 24 CMIP6 models,leading to global 1.50°C and 2.00°C WTT above pre-industrial levels in a wide range (Table 2).Globally,under the SSP1-2.6,the 1.50°C and 2.00°C WTT range from 2019 to ‘never’ among 24 models is projected,as half of them would not reach the 2.00°C warming threshold until 2100 (Fig.a2).However,under the SSP2-4.5,SSP3-7.0,and SSP5-8.5,the 1.50 and 2.00°C WTT would appear in all the models and would occur earlier in the higher emission scenarios (Figs.b2-d2).

Table 2 Years (nine-year running mean) in which the global mean near-surface air temperature reaches 1.50°C and 2.00°C warming thresholds above pre-industrial (1850-1900) levels based on Coupled Model Intercomparison Project phase 6 models

We found that the global 1.50°C WTT of MME is relatively consistent in 2027,2027,2028,and 2025 under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.Therefore,the 2.00°C WTT would occur in 2068,2046,2044,and 2040 in the four scenarios,respectively.Additionally,the 1.50°C and 2.00°C WTT would be considerably advanced in high-emission scenarios.The period from 1.50°C to 2.00°C would be shortened from 41 yr in SSP1-2.6 to 15 yr in SSP5-8.5.Furthermore,as shown in Figs.a2-d2,the nine-year running mean temperature results of MME indicate that global warming will reach 1.99°C,2.89°C,4.09°C,and 4.91°C above pre-industrial levels by the end of this century under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.

3.3 Spatial distribution of regional 1.50°C and 2.00°C WTT

The warming trends are spatially uniform from 2015 to 2100 under the SSP1-2.6 scenario (Fig.3).Subsequently,a slight downward trend is observed in the North Atlantic region.Under the other three scenarios,SSP2-4.5,SSP3-7.0,and SSP5-8.5,the warming trends have a similar pattern,whereby warming at high latitudes is higher than at low latitudes and faster over land than over the ocean.These differences become more apparent with an increase in the emission scenarios.This implies that future warming trends will be more substantial over high latitudes and land and that regional differences will be more prominent.Generally,the global SAT warming trend becomes faster with an increase in emission scenarios from 0.09°C/10yr to 0.21°C/10yr,0.37°C/10yr,and 0.47°C/10yr under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.

Fig.3 Spatial distribution of the regional near-surface air temperature warming trends between 2015 and 2100 under the Coupled Model Intercomparison Project phase 6 Shared Socioeconomic Pathway scenarios of a.SSP1-2.6;b.SSP2-4.5;c.SSP3-7.0;d.SSP5-8.5.All trends are statistically significant at the 0.01 significance level

Owing to the uneven warming trends across different regions under the four SSP scenarios,regional 1.50°C and 2.00°C WTT exhibit extreme discrepancies worldwide (Fig.4).Generally,the faster the warming rate,the faster the corresponding warming thresholds are reached(Fig.3 and Fig.4),and the 1.50°C and 2.00°C WTT occur earlier from low to high latitudes under all SSPs.In particular,the Arctic warmed by over 1.50°C in the 1980s and 2.00°C in the 2010s because of the Arctic amplification effect (Chen et al.,2019;Cai et al.,2021).Second,the WTT was reached earlier over land than over the ocean.Most parts of Eurasia,North America,the hinterland of Africa,Australia,and South America had warmed over 1.50°C before 2020.In contrast,the 1.50°C WTT in the ocean between the equator and 50°N would occur after 2030.However,most of the Southern Ocean would never exceed 1.50°C under SSP1-2.6.In the case of 2.00°C WTT,most of each continent would reach the target before 2030 in all scenarios.The 2.00°C WTT in the Southern Ocean ranges from ‘never’ in SSP1-2.6 to 2070s in SSP5-8.5.

Fig.4 Spatial distribution of the regional 1.50°C and 2.00°C warming threshold-crossing time under Shared Socioeconomic Pathway scenarios of a,b.SSP1-2.6;c,d.SP2-4.5;e,f.SSP3-7.0;g,h.SSP5-8.5;and the zonal warming trend (2015-2100) of the land and ocean and their corresponding zonal warming threshold-crossing time.The white areas are the regions that would not reach the 1.50°C warming threshold

Approximately 18.00% (6.40%) of the land warmed by over 1.50°C (2.00°C) by 2020.In 2040-2060,over 98.70% (77.50%),99.70% (89.30%),99.80% (93.40%),and 100.00% (98.00%) of the land would warm by over 1.50°C (2.00°C) under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.However,only 6.80%(4.20%) of the ocean warmed by over 1.50°C (2.00°C)in 2020 owing to the large heat capacity of water.In particular,the 1.50°C and 2.00°C WTT of the Southern Ocean between 0°-60°S would occur over 20-70 yr later than the global mean SAT.Therefore,the later warming times in the Southern Ocean dramatically slowed the time required for the global mean SAT to reach 1.50°C and 2.00°C WTT.

4 Discussion

4.1 Uncertainty in the projected warming threshold

The uncertainty in the warming threshold predicted by the model is mainly caused by differences in climate sensitivity,which may be due to uncertain intrinsic climate systems (Zhou et al.,2018).Fig.5 shows the proportion of the CMIP6 models that reached the warmingthreshold for a particular year.Although the proportion of global warming reaching 1.50°C by 2100 under the SSP1-2.6 scenario will not be 100.00%,it is unlikely to be realized from the currently observed warming.Under the other three SSP scenarios,the proportions of models which reach the 1.50°C warming threshold will reach 95.00% around 2050 and will reach 100.00% by the end of the 21st century.Furthermore,the possibility of reaching the warming threshold of 2.00°C by the end of the 21st century will also reach 100.00% under SSP3-7.0 and SSP5-8.5.The global mean SAT projections by the CMIP6 models indicate that it is possible to control global warming to within 2.00°C only under the SSP2-4.5 scenario.This highlights the urgent need for controlling global warming.

Fig.5 Proportion of Coupled Model Intercomparison Project phase 6 model simulations that have temperatures reach the 1.50°C and 2.00°C warming thresholds each year.The vertical black lines represent the reach time for the multi-model ensemble mean to reach the corresponding warming threshold,the red and blue vertical lines refer to 90.00% confidence intervals

Uncertainty in future climate prediction results is related to the formulation of greenhouse gas emission-reduction policies (Zhou and Chen,2015).Fig.6 shows the SNR distribution of global mean SAT change at the warming thresholds of 1.50°C and 2.00°C.The SNR distribution of the global warming projections showed relatively consistent distribution characteristics under the four scenarios,irrespective of the warming threshold used.The SNR was largest between the north and south latitudes of 30° and lower in the 60° south latitude zone,the Indian Peninsula,and the Tibetan Plateau.Moreover,because of the small background variability,the SNR was largest at low latitudes (Jiang et al.,2016).Except for a small region of the sea to the east and south of Greenland (north of the Atlantic Ocean) and south of the Pacific Ocean,the SNR of the rest of the world was >1.This indicates that the projected global SAT changes at 1.50°C and 2.00°C global warming thresholds have good consistency among the models,and the simulation results are relatively reliable.This can be used as the basis for formulating emission-reduction policies for government departments.

Fig.6 Spatial distribution of the signal-to-noise ratio when global warming reaches 1.50°C and 2.00°C under Shared Socioeconomic Pathway (SSP) scenarios of a,b.SSP1-2.6;c,d.SSP2-4.5;e,f.SSP3-7.0;g,h.SSP5-8.5.White areas represent the signal-to-noise ratio(SNR) <1

Additionally,the areas with SNR <1 are consistent with the regions in the world that have the slowest warming trends (even cooling trends under SSP1-2.6)and will reach the global warming threshold of 2.00°C the latest.This indicates substantial differences in the simulation results of the CMIP6 models in these areas.

4.2 Specific analysis of 1.50°C and 2.00°C WTT

To further quantify the geographical pattern of 1.50°C and 2.00°C WTT,we compared the zonal mean warming times with the mean warming trends over the land and ocean,respectively.Zonal WTT was negatively correlated with the warming trend.Generally,the warming trends are faster over high latitudes than at low latitudes,and faster over land than the ocean,corresponding to the 1.50°C or 2.00°C WTT being earlier at high latitudes than at low latitudes,and earlier over land than the ocean (Jiang et al.,2016;Shen et al.,2021).However,the Arctic,with the fastest warming rate,which is three times higher than the global mean,exceeded the 1.50°C(2.00°C) WTT before 2000 (2010).

The warming rate of land is approximately 1.35-1.46 times higher than that of the ocean between 60°N-60°S in 2015-2100.Thus,when the global mean SAT reaches 1.50°C (2.00°C) WTT,the warming over land between 60°N-60°S could be approximately 1.85-1.89°C(2.50-2.53°C) under all the four SSP scenarios.The results show that the land between 60°N-60°S will reach the 1.50°C (2.00°C) WTT in 2025 (2040),2024 (2041),2024 (2038),and 2023 (2035) under the SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,and 2(28),3(5),4(6),2(5) yr earlier than the global mean SAT,respectively.As shown in Fig.7,when the global warming reaches the 1.50°C warming threshold,the warming magnitude of 80.73%,80.49%,81.95%,and 78.83% of the land will rise above or even exceed 1.50°C by far under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.Moreover,the maximum temperature increase in the high latitudes of the northern hemisphere will exceed 5.00°C.Similarly,when global warming reaches the 2.00°C warming threshold,the areas where land warming exceeds 2.00°C under the four scenarios will be 82.57%,81.57%,81.91%,and 81.64%,respectively,and the maximum regional warming magnitude will exceed 7.00°C.Although this warming magnitude was higher than that of previous research results,the spatial distribution of the warming magnitude was relatively consistent (Huang et al.,2022).In conclusion,when emphasizing the global mean SAT at 1.50°C and 2.00°C warming targets,the high-latitude and high-altitude regions should receive more attention as they are the most sensitive and vulnerable to global warming(Wang et al.,2020;Huang et al.,2022;Qi et al.,2022).

Fig.7 Spatial distribution of the near-surface air temperature warming magnitude over land above pre-industrial (1850-1900) levels when global warming reaches 1.50°C and 2.00°C warming thresholds under the Shared Socioeconomic Pathway (SSP) scenarios of a,b.SSP1-2.6;c,d.SSP2-4.5;e,f.SSP3-7.0;g,h.SSP5-8.5

4.3 The urgency of controlling global warming

By controlling global warming within 1.50°C rather than 2.00°C,the gained 0.50°C buffer could reduce the occurrence of drastically more severe climate impacts and disasters (Su et al.,2018;Zhang et al.,2019).However,as the global mean SAT increases by 1.50°C,some regions,such as the Arctic and Tibetan Plateau,have already suffered from warming over 1.50°C for a long time.The Tibetan Plateau warmed to over 1.50°C in 2017 relative to the pre-industrial levels,resulting in glaciers having retreated heavily (Yao et al.,2019).Reducing snow and ice cover exacerbates global warming through albedo feedback.Therefore,keeping the global mean SAT even at 1.50°C will cause adverse impacts and disasters on the regional climate.Notably,this study predicts that global warming will reach 1.99°C,2.89°C,4.09°C,and 4.91°C above pre-industrial levels by the end of this century under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.Meanwhile,unmitigated warming will undoubtedly result in climate damage and economic losses,and exacerbate regional income inequality (Burke et al.,2015;Wang and Teng,2022).

According to the most stringent warnings in the latest AR6 from the IPCC,humanity will likely not prevent the Earth from at least temporarily warming to 1.50°C above pre-industrial levels (IPCC,2022;Tollefson,2022).This study also indicates that it would be difficult to control the global mean SAT within 1.50°C above pre-industrial levels by 2100,despite the carbon peaking and carbon neutralization policies that have been formulated (Gillett et al.,2021;Tollefson,2021;Ma et al.,2022).Additionally,it would be a major challenge to control the global mean SAT within 2.00°C by the end of the 21st century unless collective actions are taken immediately.From the perspective of fairness and morality,developed countries should help developing countries to ensure that the global average carbon peak occurs as soon as possible.People are responsible for maintaining climate change and its devastating effects(Filippelli et al.,2021).The time for emission reduction actions is now (Kang et al.,2020;IPCC,2022).

5 Conclusions

In this study,we evaluated the simulation abilities of 24 CMIP6 models.Subsequently,we projected the global and regional 1.50°C and 2.00°C WTT relative to the preindustrial levels using the MME.Our results show that the global 1.50°C (2.00°C) WTT will be reached in 2027 (2068),2027 (2046),2028 (2044),and 2025(2040) under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.Furthermore,global warming will reach 1.99°C,2.89°C,4.09°C,and 4.91°C above pre-industrial levels by the end of the 21st century.However,the results also revealed that the regional 1.50°C and 2.00°C WTT exhibit a large discrepancy in the background of consistent global warming under the SSPs.Generally,warming proceeds faster at high latitudes than at low latitudes,and over land than over the ocean,leading to the 1.50 or 2.00°C WTT to occur earlier at high latitudes than at low latitudes and over land than over the ocean.Approximately 18.00% (6.40%) of the land warmed by over 1.50°C (2.00°C) by 2020.In 2040-2060,over 98.70% (77.50%),99.70% (89.30%),99.80% (93.40%),and 100.00% (98.00%) of the land would warm by over 1.50°C (2.00°C) under SSP1-2.6,SSP2-4.5,SSP3-7.0,and SSP5-8.5,respectively.The regional 1.50°C (2.00°C) WTT should be fully considered when discussing the global mean SAT warming at 1.50°C (2.00°C),and vulnerable regions,such as the Arctic and Tibetan Plateau that have already warmed by over 1.50°C as the global mean SAT has warmed by 1.50°C,should receive particular attention.Our study provides a reference for the formulation of regional emission-reduction policies.

Conflict of Interest

The authors have no competing interests to declare that are relevant to the content of this article.

Author Contributions

MENG Yali: conceptualization,methodology,formal analysis and investigation,writing-original draft preparation,writing-review and editing;DUAN Keqin: conceptualization,methodology,resources,writing-review and editing,supervision;SHANG Wei: methodology,writing-review and editing;SHI Peihong: software,writing-review and editing;LI Shuangshuang: writingreview and editing;CHENG Ying: writing-review and editing;CHEN Rong: software,writing-review and editing;ZHANG Zhaopeng: software,investigation.