Constraint on regional land surface air temperature projections in CMIP6 multi-model ensemble

doi:10.21203/rs.3.rs-2095871/v1

Download PDF

Article

Constraint on regional land surface air temperature projections in CMIP6 multi-model ensemble

https://doi.org/10.21203/rs.3.rs-2095871/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 10 Jul, 2023

Read the published version in npj Climate and Atmospheric Science →

You are reading this latest preprint version

The reliability of the near-land-surface air temperature (LSAT) projections from the state-of-the-art climate-system models that participated in the Coupled Model Intercomparison Project phase six (CMIP6) is debatable, particularly on regional scales. Here we introduce a new method of constructing a constrained multi-model-ensemble (CMME), based on rejecting models that fail to reproduce observed LSAT trends. We use the CMME to constrain future LSAT projections under the Shared Socioeconomic Pathways 5-8.5 (SSP5-8.5) in the 21^st century; this scenario represents the high end of the range of future pathway uncertainty. In comparison with the “raw” (unconstrained) CMIP6 multi-model ensemble (MME) mean, the impact of the observation-based constraint is less than 0.1 ^oC/100years at global scale over the 21^st century. However, the regional results show a wider range of positive and negative adjustments from -1.0^oC/100years to 1 ^oC/100years. The ideal 1pctCO2 experiment suggests that the response of LSAT to carbon dioxide (CO₂) forcing on regional scales is responsible for the MME biases in historical period, indicating the higher reliability of CMME in the 21^st century projections. The advantage of CMME is that it goes beyond the idea of “model democracy” assumed in MME. The unconstrained CMIP6 MME may be overestimating the risks of future warming over North America, Europe, and North Africa, but underestimating the risks over Asia.

Land Surface Air Temperature

Projections over the 21st century

Carbon Dioxide

The CMIP6 program is the most recent effort in coordinating the design and distribution of a large number of the state-of-the-art climate model simulations of the past, present, and future climates (Eyring et al., 2016). It serves as the basis for the sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2022). The observed global warming trend can be generally well captured by the state-of-the-art climate system models participating in CMIP6 (Papalexiou et al., 2020). Hansen and Stone (2016) demonstrated that almost two-thirds of the impacts related to atmospheric and ocean temperature changes can be confidently attributed to anthropogenic forcing. There is a linear relationship between global mean surface air temperature and the equivalent-CO2 concentrations (Wu et al., 2019). And the contribution from equivalent-CO2 to historical global mean surface air temperature is about 70% on multi-decadal or longer timescales. The equivalent-CO2 is calculated by converting amounts of all the greenhouse gases to the equivalent amount of carbon dioxide with the same global warming potential.

Greenhouse gases (GHG), CO2 in particular, are consider to be the major external forcing in the 20th century and will also have a crucial impact in the 21st century projection under the high-emission scenario. A high correlation between global mean near-surface air temperature and the CO₂ concentrations has also been found in future projections (Gao et al., 2021). However, many of the latest CMIP6 models have larger climate sensitivities than the previous CMIP5 generation (e.g., Flynn and Mauritsen, 2020; Meehl et al., 2020; Nijsse et al., 2020). Higher climate sensitivity indicates the ‘hot model’ problem in CMIP6 models (Hausfather et al., 2022), and the larger climate sensitivity range suggests a wider range of warming responses to CO₂-forcing and larger model uncertainty in future warming projections (Zelinka et al., 2020).

There have been several studies trying to narrow uncertainty on estimates of past and future human-induced warming based on detection and attribution techniques or metrics of climate sensitivity (e.g., Stott and Kettleborough, 2002; Knutti et al., 2017). Tokarska et al. (2020) found a strong correlation between the recent global warming trend and TCR, and the past warming trend therefore can be used to constrain future warming projections in climate models on global scales. But most of them are focus on the projections at global scale (e.g., Ribes et al., 2021). However, it is regional-scale changes that are of vital importance for the impacts of climate change (IPCC, 2022). Under the global warming, climate changes are more extreme regionally and induce severer impacts. For example, the rate of heatwave increases since the mid-twentieth century, but trend magnitudes are not globally uniform. Decadal trends in the percentage change of heatwave trends are biggest over northern South America, the Middle East, the Maritime Continent at 50% decade^-1, but range between 10% and 30% decade^-1 over most of other regions (Perkins-Kirkpatrick and Lewis, 2020). In this study, we try to constrain future warming projections but focus on regional land surface air temperature (LSAT) trend projection under the high-emission SSP5-8.5 scenario (O'neill et al., 2016). We only focus on surface air temperature trend over land since Atlantic multidecadal variability and Pacific decadal variability account for most of variability over the oceans (Chylek et al., 2014; Meehl et al., 2016; Wu et al., 2019).

Historical LSAT trend: observation and CMIP6 simulations

Figure 1a shows the 50-year running mean LSAT trend from 1950 to 2014 in the CRU dataset. Almost the entire global land surface has experienced significant surface warming. The warming trend is remarkable over the mid- and high- latitudes of the Northern Hemispheric continent and northeast Africa, weakest over the Tibetan Plateau and its downstream area, and the Andes Mountains along the western edge of South America. The global warming features can be generally well captured by the ensemble mean of the 35 CMIP6 models (MME, Fig. 1b). The spatial correlation coefficient between MME and GISTEMP is 0.57, higher than the correlations for 31 of the 35 models. However, the LSAT trend in MME is spatially much smoother than in CRU. As shown in Fig. 1d, regions with LSAT trend underestimated by more than 0.5^oC/100years in MME are where the observed warming rate higher than 3^oC/100years, e.g., over Northern Asia and Alaska. And the overestimation in MME is where observed warming rate lower than 1^oC/100years, e.g., over south America and the Continental United States. Amplitude of the underestimation over East Asia (EAS) and the overestimation over the U.S. are about 1^oC/100years, comparable with the warming rate in the observation. That is, there is a big challenge for recent models to reasonably reproduce the LSAT trend in EAS and U.S.. The longitudinal gradient between Europe and northern Asia shown in CRU is also missing in MME.

General features of LSAT trend in CMME (Fig.1c), constrained by the observed LSAT trend, are similar to those in MME. The blank areas over land, such as the Andes Mountains, are where no model can capture the LSAT trend reasonably. The improvement from MME to CMME is evident all over the world, although the global mean LSAT trend is barely changed (1.96^oC/100years in MME versus 1.95^oC/100years in CMME). The spatial correlation coefficient with GISTEMP increases to 0.94 in CMME, 0.37 higher than in MME. As expected, the differences between CMME and MME (Fig.1e) are similar to the MME biases but in opposite sign over 94.4% of the total area. Regions with the largest differences between CMME and MME are the areas where the MME shows the largest biases, i.e., most parts of Asia, the United States, Western South America, and surrounding the Black Sea. We conclude that, although the geographic distribution of recent warming in MME is better than most individual models, the warming amplitude may be largely biased at regional scale. The CMME result further improves the geographic distribution of LSAT trend in MME and its amplitude at regional scale, because by construction the CMME is closer to the observed trends.

The 35 models’ mean capabilities in reproducing regional LSAT trend are further quantified in 44 different landscape regions (Fig. 1f). Model capability is defined as the percentage of area where model results are chosen in CMME. Region definitions follow the IPCC AR6 Working Group 1 (WG1) reference regions over land (Iturbide et al., 2020). Region names and their acronym are introduced in Table 2. Model grids with multi-model mean model capability higher than 50% covers over 77.4% of land area north of 60^oS. Model capability is relatively high over the Central Eurasian continent (WCE, EEU, WSB, ESB, WCA, ECA, TIB, ARP, and SAS), Africa (SAM, WAF, SEAF, and ESAF), and Central Australia (CAU), but relatively less good over the American continents, Northern Europe, East Asia, and Australasia.

Future LSAT projections in the raw and constrained ensembles

Here we examine the projected future LSAT trend in the 21^st century based on CMIP6 MME projections under the high scenario SSP5-8.5 and assess the adjustments in CMME. The LSAT trends in four 20-year periods are examined (Figs.2a-2d): 2021~2040, 2041~2060, 2061~2080, and 2081~2100. The LSAT changes in MME features global warming in all four periods. The spatial patterns of the warming are also similar to those featured in the 20^th century. Globally, the LSAT trend in 2021~2040 is about 4.86^oC/100yrs, and the warming rate accelerates gradually in time: 6.81^oC/100yrs in 2041~2060, 7.57^oC/100years in 2061~2080, and 7.90^oC/100years in 2081~2100.

Consistent with what we observed in the historical period, the global mean LSAT trend projections in CMME are also close to those in MME (Figs. 2e-2h). However, the regional differences are pronounced (Figs. 2i-2l). Left out low model capability over the Andes Mountains, constraint of LSAT projection in all the four projection periods generally resemble the effect of constraint ensemble in historical period (Fig. 1e) over more than 2/3 of land surface: regions that show exaggerate historical warming trend are also the regions where CMME tends to reduce the projected warming rate, and vice versa. According to CMME adjustments, warming projection over the Eurasian continent may be more intense than the raw ensemble (MME), whereas the warming risk over North America may be lower than expected.

Regions with the most considerable adjustments in future projection

The regional LSAT projections in the raw and constrained ensembles are further detailed over the 44 landscape reference regions. As shown in Fig.3a, LSAT increases monotonically over all regions but the warming rates are quite different. The LSAT trend during the last 20 years (2080-2100) are listed in Table 2. There are five regions with increasing rate higher than 10^oC/100years in MME (RAR, NEN, RFE, WSB, and ESB), four of which locate at high latitude of Asia. Most of the least warming regions with LSAT trend lower than 6^oC/100years are in South America and Australasia.

The CMME adjustments (Fig. 3b) vary from -1.01^oC/100year in Central North America (CNA) to 0.49^oC/100years in East Siberia (ESB). Most of the regions show a cooling adjustment in CMME. Accounting for the adjustment larger than 0.2^oC/100years, warming trend are intensified only over 5 regions but suppressed over 17 regions. The amplification of warming is evident in Asia, and suppression of warming is mostly located over American continents, Europe and North Africa. Relative to the regional LSAT trend in MME, the warming adjustment is generally weaker and most significant over ESB (4.7%); the cooling adjustment prevails over NEN (-5.3%), ENA (-8.4%), and CNA (-13.2%) in North America, NWS (-5.4%), SES (-5.6%), and NSA (-6.0%) in South America, and NEU (-10.0%) in Europe. The adjustment may be partly attributed to forcings other than GHG. Previous observational and model-based studies have found a cooling trend over the southeast and central United States, i.e., the United States ‘warming hole’. This is suggested to be attributed to anthropogenic aerosol forcing or internal climate variability (Mascioli et al., 2017) with dominant variation by season, region, and time period.

Generally, as suggested by CMME, the warming challenge may be more intense over Asia and less intense over American continents, Europe and North Africa than suggested by the raw CMIP6 MME projections. The physical mechanism of the CMME adjustment is investigated and discussed in the next subsection.

LSAT-response to CO₂-forcingand its impact on the LSAT trend reproduction

Global warming in the 20^th century is mainly attributable to the increase in well-mixed greenhouse gases (WMGHGs), mainly the CO₂. The 1pctCO2 experiment is a transient climate simulation, in which the CO₂ is the only anthropogenic external forcing. It is an idealized CO₂-forced experiment that resembles the CO₂ forcing during the Industrial Era. Here we compare the LSAT-response to CO₂-forcing in MME and CMME in 1pctCO2 to quantify the effect of our bias-correction procedure in eliminating CO₂-forcing response biases in climate models.

Figures 4a and 4b show the LSAT linear trend from 1850 to 1910 in the 1pctCO2 experiments for MME and CMME. The great similarity between the LSAT-response to CO₂-forcing (Figs. 4a) and the LSAT trend with all forcings (Figs. 1b) with correlation coefficient of 0.91, confirms the dominant role of CO₂ on global warming. Warming trend in CMME generally resemble the MME results, but with different amplitude of regional warming (Fig. 4b). Fig. 4c shows the differences between MME and CMME in LSAT-response to CO₂-forcing in the 1pctCO2 experiment, which allows us to compare the CO₂-response in MME to what the response would be if the historical LSAT biases was smaller. Interestingly, their differences closely resemble the historical trend bias in the MME (Fig. 1d). Over more than 15,000 model grids, the corresponding spatial correlation coefficient is 0.72, significant at 99% confidence level using student's t-test. That is, the historical LSAT trend biases in MME are mainly due to model capability in simulating CO₂-forcing response. Future constraint by CMME is therefore physically relative to its more reasonable CO₂-forcing response, since CO₂ is also the dominant forcing in future emission scenario.

The radiative forcing due to anthropogenic activity has increased continuously during the Industrial Era, with increases in atmospheric CO₂ playing the major role. By observational constraint, our study found that model response to CO₂-forcing is crucial for model performances in reproducing recent LSAT trend, especially at regional scale. Therefore, under a high CO₂ emission scenario, model with better capability in recent LSAT trend reproduction is expected to give a better estimate of future LSAT change.

In this study, we introduced a new method to constrain future LSAT projections based on observed constrained model ensemble (CMME), especially at regional scale. Under the SSP5-8.5 scenario, global warming will accelerate as the time period extends further into the future in the raw CMIP6 ensemble (MME). The CMME is close to MME at the global scale, but the regional warming amplitudes are quite different. CMIP6 models may underestimate future warming over Asia but overestimate it more over the American continents, Europe and North Africa. More specifically, the cooling adjustment over Central and Eastern North America (CNA and ENA) and Northern Europe (NEU) are 10% of the MME projection or more.

Therefore, we should take care when handling regional LSAT projections by CMIP6 models, since the CMIP6 model response to CO₂-forcing may be larger or smaller than expected and vary with region. Moreover, we should take great caution with projections over the regions with larger projection adjustment and lower model capability, i.e., CNA, ENA, and NEU, where factors other than CO₂ may also be important for the LSAT changes

Grided observational date sets

The observational gauge-based grided monthly mean LSAT from the Climatic Research Unit (CRU TS4) at University of East Anglia from 1950 to 2014 (Harris et al., 2020) serves as the observation reference. The dataset was derived from historical weather stations and presented as interpolated variable on a 0.5° grid over land surface. It is highly reliable, especially in mid-latitudes where there is a dense observation network. The CRU data and all the model outputs are interpolated to a regular 1° × 1° grid to facilitate the subsequent operation.

CMIP6 models data sets

Simulations of 35 models from the Coupled Model Intercomparison Project phase 6 (CMIP6) were used. Three experiments are relevant for our study: all-forcing historical experiments from 1950-2014, idealized 1pctCO2 experiments from 1850 to 1910, and Shared Socioeconomic Pathways 5-8.5 (SSP5-8.5) experiments in the 21st century. Table 1 lists the models’ name, institution, and the horizontal resolution of the atmospheric component. For the sake of egality, we use only the first member of each individual model if multiple runs were produced.

In the 1pctCO2 experiment, CO₂ increases at a prescribed rate of 1% per year from 285ppmv in 1850AD until the concentration doubles at model year 70. The 1pctCO2 experiment is an idealised configuration but can be used to identify the climate response to CO₂ increase as the experiment does not include any confounding effects from other climate forcings like aerosols and land-use change. 1850 to 1910 in the 1pctCO2

The SSP5-8.5 scenario represents the high end of the range of future pathways and produces a radiative forcing of 8.5 W m⁻² in 2100.

The Constrained ensemble

A simple ensemble mean of all the models may fail to capture the real trend since some heavily-biased simulations may diminish the real signals or lead to unreliable results. The capabilities of individual models in reproducing the LSAT trend may also vary with regions, e.g., models that can reasonably produce the LSAT trend in Asia may fail to simulate the LSAT trend in the U.S. or other areas. Therefore, we build a set of ensemble members, which we call a “constrained multi-model ensemble” (noted hereafter as CMME), by selecting models for each grid box based on observed LSAT trend. A 50-year running mean annual mean LSAT trend is performed from 1950 to 2014 both for historical simulations and CRU. The running mean operation was necessary to minimize the impact of internal variability and short-term variations. Models are included in the CMME if the LSAT trend falls within the 50% to 150% range of the corresponding observed counterpart. The selection of models is applied at each grid box, i.e., a different subset of the models is chosen for each grid box. Conversely, a model only contributes to the statistics of the CMME at the grid points for which it satisfies the observational constraint. The constrained ensemble will be used to calibrate the regional LSAT projection under the SSP5-8.5 scenario in the 21^st century.

Data Availability. All the model data can be freely downloaded from the Earth System Federation Grid (ESGF) nodes (https://esgf-node.llnl.gov/search/cmip6/). The CRU dataset is freely available on https://crudata.uea.ac.uk/cru/data/hrg/

Code Availability

All data processing codes are available if a request is sent to the corresponding authors.

Author contributions

The main ideas were formulated by JZ and TW. JZ wrote the original draft. The results were supervised by TW, LL, and KF. All the authors discussed the results and contributed to the final manuscript.

Competing interests

The authors declare that they have no conflict of interest.

Competing interests

The authors declare no competing interests.

Chylek, P., Klett, J. D., Lesins, G., Dubey, M. K., and Hengartner, N.: The Atlantic Multidecadal Oscillation as a dominant factor of oceanic influence on climate, Geophysical Research Letters, 41, 1689–1697, 10.1002/2014gl059274, 2014.
Eyring, V., Bony, S., Meehl, G. A., Senior, C. A., Stevens, B., Stouffer, R. J., and Taylor, K. E.: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization, Geosci. Model Dev., 9, 1937–1958, 10.5194/gmd-9-1937-2016, 2016.
Flynn, C. M. and Mauritsen, T.: On the climate sensitivity and historical warming evolution in recent coupled model ensembles, Atmos. Chem. Phys., 20, 7829–7842, 10.5194/acp-20-7829-2020, 2020.
Gao, F., Wu, T., Zhang, J., Hu, A., and Meehl, G. A.: Shortened Duration of Global Warming Slowdowns with Elevated Greenhouse Gas Emissions, Journal of Meteorological Research, 35, 225–237, 10.1007/s13351-021-0134-y, 2021.
Hansen, G. and Stone, D.: Assessing the observed impact of anthropogenic climate change, Nature Climate Change, 6, 532-+, 10.1038/nclimate2896, 2016.
Harris, I., Osborn, T. J., Jones, P., and Lister, D.: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset, Scientific Data, 7, 10.1038/s41597-020-0453-3, 2020.
Hausfather, Z., Marvel, K., Schmidt, G. A., Nielsen-Gammon, J. W., and Zelinka, M.: Climate simulations: recognize the 'hot model' problem, Nature, 605, 26–29, 2022.
IPCC, 2022: Climate Change 2022: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, Cambridge, 2022).
Iturbide, M., Gutierrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofino, A. S., Di Luca, A., Henrique Faria, S., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martinez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S.: An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth System Science Data, 12, 2959–2970, 10.5194/essd-12-2959-2020, 2020.
Knutti, R., Rugenstein, M. A. A., and Hegerl, G. C.: Beyond equilibrium climate sensitivity, Nat. Geosci., 10, 727-+, 10.1038/ngeo3017, 2017.
Mascioli, N. R., Previdi, M., Fiore, A. M., and Ting, M.: Timing and seasonality of the United States ‘warming hole’, Environmental Research Letters, 12, 034008, 10.1088/1748–9326/aa5ef4, 2017.
Meehl, G. A., Hu, A. X., Santer, B. D., and Xie, S. P.: Contribution of the Interdecadal Pacific Oscillation to twentieth-century global surface temperature trends, Nature Climate Change, 6, 1005–1008, 10.1038/nclimate3107, 2016.
Meehl, G. A., Senior, C. A., Eyring, V., Flato, G., Lamarque, J.-F., Stouffer, R. J., Taylor, K. E., and Schlund, M.: Context for interpreting equilibrium climate sensitivity and transient climate response from the CMIP6 Earth system models, Science Advances, 6, 10.1126/sciadv.aba1981, 2020.
Nijsse, F. J. M. M., Cox, P. M., and Williamson, M. S.: Emergent constraints on transient climate response (TCR) and equilibrium climate sensitivity (ECS) from historical warming in CMIP5 and CMIP6 models, Earth Syst. Dynam., 11, 737–750, 10.5194/esd-11-737-2020, 2020.
O'Neill, B. C., Tebaldi, C., van Vuuren, D. P., Eyring, V., Friedlingstein, P., Hurtt, G., Knutti, R., Kriegler, E., Lamarque, J. F., Lowe, J., Meehl, G. A., Moss, R., Riahi, K., and Sanderson, B. M.: The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6, Geosci. Model Dev., 9, 3461–3482, 10.5194/gmd-9-3461-2016, 2016.
Papalexiou, S. M., Rajulapati, C. R., Clark, M. P., and Lehner, F.: Robustness of CMIP6 Historical Global Mean Temperature Simulations: Trends, Long-Term Persistence, Autocorrelation, and Distributional Shape, Earth's Future, 8, e2020EF001667, https://doi.org/10.1029/2020EF001667, 2020.
Perkins-Kirkpatrick, S. E. and Lewis, S. C.: Increasing trends in regional heatwaves, Nature Communications, 11, 10.1038/s41467-020-16970-7, 2020.
Ribes, A., Qasmi, S., and Gillett, N. P.: Making climate projections conditional on historical observations, Science Advances, 7, eabc0671, doi:10.1126/sciadv.abc0671, 2021.
Stott, P. A. and Kettleborough, J. A.: Origins and estimates of uncertainty in predictions of twenty-first century temperature rise, Nature, 416, 723–726, 10.1038/416723a, 2002.
Tokarska, K. B., Stolpe, M. B., Sippel, S., Fischer, E. M., Smith, C. J., Lehner, F., and Knutti, R.: Past warming trend constrains future warming in CMIP6 models, Science Advances, 6, 10.1126/sciadv.aaz9549, 2020.
Wu, T., Hu, A., Gao, F., Zhang, J., and Meehl, G.: New insights into natural variability and anthropogenic forcing of global/regional climate evolution, npj Climate and Atmospheric Science, 2, 10.1038/s41612-019-0075-7, 2019.
Zelinka, M. D., Myers, T. A., McCoy, D. T., Po-Chedley, S., Caldwell, P. M., Ceppi, P., Klein, S. A., and Taylor, K. E.: Causes of Higher Climate Sensitivity in CMIP6 Models, Geophysical Research Letters, 47, 10.1029/2019gl085782, 2020.

Table 1. Information of 35 CMIP6 CGCM models.

Model	Institute	AGCM resolution
ACCESS-CM2	CSIRO-ARCCSS CSIRO and Austr. Res. Council Centre of Excellence for Climate System Science, Australia	140 km, 85L, Top 85 km
ACCESS-ESM1-5	CSIRO Commonwealth Scientific and Industrial Research Organization, Australia	140 km, 38L, Top 39 km
AWI-CM-1-1-MR	AWI Alfred Wegener Institute,Germany	80 km, 95L,Top 80 km
BCC-CSM2-MR	BCC Beijing Climate Center, China	100 km, 46L, Top 45 km
CAMS-CSM1-0	CAMS Chinese Academy of Meteorological Sciences, China	100 km, 31L, Top 31.2 km
CanESM5	CCCMa Canadian Centre for Climate Modelling and Analysis, Canada	250 km, 49L, Top 48 km
CAS-ESM2-0	CAS Chinese Academy of Sciences, China	140km, 30L, Top 42km
CESM2-WACCM	NCAR National Center for Atmospheric Research, U.S.A.	100 km, 70L, Top 80 km
CIESM	THU Department of Earth System Science, China	100 km, 30L, Top 42 km
CMCC-CM2-SR5	CMCC Centro Euro-Mediterraneo sui Cambiamenti Climatici, Italy	100 km, 30L, Top 43 km
CMCC-ESM2	CMCC	100 km, 30L, Top 43 km
EC-Earth3-CC	EC-Earth consortium, Europe	80 km, 91L, Top 80 km
EC-Earth3-Veg	EC-Earth	80 km, 91L, Top 80 km
FGOALS-f3-L	CAS	90 km, 32 L, Top 42 km
FGOALS-g3	CAS	190 km, 26 L, Top 42 km
FIO-ESM-2-0	FIO-QNLM First Institute of Oceanography , and Pilot National Laboratory for Marine Science and Technology, China	100km, 26L, Top 43 km
GFDL-CM4	NOAA-GFDL National Oceanic and Atmospheric Administration, Geophysical Fluid Dynamics Laboratory, U.S.A.	100 km, 33L, Top 48 km
GFDL-ESM4	NOAA-GFDL	100 km, 49L, Top 80 km
HadGEM3-GC31-LL	MOHC Met Office Hadley Centre, U.K.	140 km, 85L, Top 85 km
HadGEM3-GC31-MM	MOHC	60 km, 85L, Top 85 km
IITM-ESM	CCCR-IITM Centre for Climate Change Research, Indian Institute of Tropical Meteorology, India	170 km, 64L, Top 61 km
INM-CM4-8	INM Institute for Numerical Mathematics, Russia	150 km, 21L, Top 31 km
INM-CM5-0	INM	150 km, 73L, Top 61 km
IPSL-CM6A-LR	IPSL Institut Pierre-Simon Laplace, France	160 km, 79L, Top 80 km
KACE-1-0-G	NIMS-KMA National Institute of Meteorological Sciences, Korea Meteorological Administration, Korea	140 km, 85L, Top 85 km
KIOST-ESM	KIOST Korea Institute of Ocean Science & Technology, Korea	190 km, 32L, Top 43 km
MIROC6	MIROC consortium JAMSTEC, AORI, NIES, RCCS, Japan	120 km, 81L, Top 80 km
MIROC-ES2L	MIROC	250 km, 40L, Top 40 km
MPI-ESM1-2-HR	MPI-M Max Planck Institute for Meteorology, Germany	80 km, 95L, Top 80 km
MPI-ESM1-2-LR	MPI-M	170 km, 47L, Top 80km
MRI-ESM2-0	MRI Meteorological Research Institute, Japan	100 km, 80L, Top 80 km
NESM3	NUIST Nanjing University of Information Science and Technology, China	170 km, 47L, Top 48 km
NorESM2-LM	NCC NorESM Climate Modelling Consortium, Norway	190 km, 32L, Top 40 km
NorESM2-MM	NCC	100km, 32L, Top 40 km
UKESM1-0-LL	MOHC	140 km, 85L, Top 85 km

Table 2. Model Capability in reproducing the LSAT trend from 1950 to 2014 in 44 IPCC AR6 WGI reference regions (units: %), and LSAT trend in 2080-2100 under SSP585 in MME and CMME adjustment (CMME-MME,^oC/100year). Regions with adjustment larger than 5% are bolded.

Region	Acronym	Full Name	Model Capability	LSAT trend (Adjustment)
North America	NWN	North-Western North America	45.2	9.81 (0.18)
	NEN	North-Eastern North America	35.7	12.11 (-0.64)
	WNA	Western North America	49.8	7.16 (-0.25)
	CNA	Central North America	38.1	7.66 (-1.01)
	ENA	Eastern North America	28.2	7.87 (-0.66)
Central America	NCA	Northern Central America	47.1	6.58 (-0.25)
	SCA	Southern Central America	21.7	5.75 (-0.06)
	CAR	Caribbean	6.7	4.68 (0.09)
South America	NWS	North-Western South America	29.4	5.79 (-0.31)
	NSA	Northern South America	39	6.36 (-0.38)
	NES	North-Eastern South America	46.7	6.11 (0.11)
	SAM	South American Monsoon	24.2	7.31 (-0.06)
	SWS	South-Western South America	10.9	5.85 (0.04)
	SES	South-Eastern South America	32.6	6.03 (-0.34)
	SSA	Southern South America	9.2	4.67 (0.15)
Europe	GIC	Greenland/Iceland	31.7	8.87 (-0.23)
	NEU	Northern Europe	23.2	7.76 (-0.77)
	WCE	Western and Central Europe	62.8	7.83 (-0.18)
	EEU	Eastern Europe	67.7	9.36 (-0.44)
	MED	Mediterranean	34.7	7.55 (-0.20)
Africa	MED	Mediterranean	34.7	7.55 (-0.20)
	SAH	Sahara	77.9	7.94 (-0.02)
	WAF	Western Africa	50.7	6.77 (-0.01)
	CAF	Central Africa	40.2	6.87 (-0.12)
	NEAF	North Eastern Africa	44.4	6.27 (-0.23)
	SEAF	South Eastern Africa	55.4	6.15 (0.01)
	WSAF	West Southern Africa	43.8	7.34 (0.32)
	ESAF	East Southern Africa	52.3	6.78 (0.05)
	MDG	Madagascar	20.9	5.29 (-0.14)
Asia	RAR	Russian Arctic	38.9	12.84 (-0.24)
	WSB	West Siberia	77.8	10.58 (-0.41)
	ESB	East Siberia	72.8	10.33 (0.49)
	RFE	Russian Far East	39.9	10.82 (0.27)
	WCA	West Central Asia	62.8	8.75 (0.04)
	ECA	East Central Asia	70.7	8.62 (0.25)
	TIB	Tibetan Plateau	54.9	8.20 (-0.34)
	EAS	East Asia	25.9	7.40 (0.24)
	ARP	Arabian Peninsula	66.3	8.22 (-0.17)
	SAS	South Asia	44.9	6.99 (0.08)
	SEA	South East Asia	16.6	4.94 (0.07)
Australasia	NAU	Northern Australia	15.7	5.86 (-0.18)
	CAU	Central Australia	52.8	6.93 (0.01)
	EAU	Eastern Australia	49.3	6.08 (0.08)
	SAU	Southern Australia	22.6	6.01 (0.08)
	NZ	New Zealand	1.4	4.98 (-0.71)

(Not answered)

Download PDF

Journal Publication

published 10 Jul, 2023

Read the published version in npj Climate and Atmospheric Science →

Editorial decision: revise
11 Nov, 2022
Review #3 received at journal
26 Oct, 2022
Review #1 received at journal
20 Oct, 2022
Review #2 received at journal
18 Oct, 2022
Reviewer #3 agreed at journal
04 Oct, 2022
Reviewer #2 agreed at journal
30 Sep, 2022
Reviewer #1 agreed at journal
29 Sep, 2022
Reviewers invited by journal
28 Sep, 2022
Editor assigned by journal
27 Sep, 2022
Submission checks completed at journal
26 Sep, 2022
First submitted to journal
23 Sep, 2022

You are reading this latest preprint version

Constraint on regional land surface air temperature projections in CMIP6 multi-model ensemble

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Result

Discussion

Method

Declarations

References

Tables

Additional Declarations

Status:

Journal Publication

Version 1