Emergent constraints on future change projections of mean and extreme temperature and precipitation in the global maize harvesting area

Radley Horton,1,a Daniel Bader,1,a Yochanan Kushnir,2 Christopher Little,3 Reginald Blake,4 and Cynthia Rosenzweig5 1Columbia University Center for Climate Systems Research, New York, NY. 2Ocean and Climate Physics Department, Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY. 3Atmospheric and Environmental Research, Lexington, MA. 4Physics Department, New York City College of Technology, CUNY, Brooklyn, NY. 5Climate Impacts Group, NASA Goddard Institute for Space Studies; Center for Climate Systems Research, Columbia University Earth Institute, New York, NY

We analyze annual extremes of daily maximum and minimum surface air temperature and of daily rainfall in East Asia and the Korean peninsula. This study made intensive use of the simulation data available from the CMIP5 (Coupled Model Intercomparison Project Phase 5) multimodels in historical and future experiments up to the year 2100, employing three different radiative forcings: RCP2.6, RCP4.5, and RCP8.5 (representative concentration pathways). Several reanalysis datasets are used to compare and evaluate the simulated climate extremes in the late twentieth century. We estimate the future changes in precipitation and temperature extremes in East Asia and Korea, and compare them to the global result, for the reference period 1986–2005. The rising rate of future cold extremes over East Asia and Korea is faster than that of warm extremes. This phenomenon appears more distinctly in Korea as a local scale, indicating more sensitivity of the Korean peninsula to global warming. The increase of the 20-year return level of maximum precipitation in the CMIP5 over East Asia by the end of twenty-first century is about 7% in the RCP2.6, 15% in the RCP4.5, and 35% in the RCP8.5 experiments, which exceed the corresponding global values. We also estimate the changes in precipitation extremes across East Asia as a function of the annual mean temperature variation at the same location. The CMIP5 sensitivity in maximum precipitation across East Asia is 5.5%/∘C, which is lower than the global figure (5.8%/∘C). The sensitivity for the Korean peninsula is 7.38%/∘C, indicating the strong impact of global warming to Korea. The results will be important in mitigating the detrimental effects of variations of climatic extremes and in improving the regional strategy for water resource and eco-environmental management, particularly for such areas in East Asia under significant changes in temperature and rainfall extremes.

Over the tropical land surface, accurate estimates of future changes in temperature, precipitation and evapotranspiration are crucial for ecological sustainability, but remain highly uncertain. Here we develop a series of emergent constraints (ECs) by using historical and future outputs from the Coupled Model Inter-comparison Project Phase 6 (CMIP6) Earth System Models under the four basic Shared Socio-economic Pathway scenarios (SSP126, SSP245, SSP370, and SSP585). Results show that the temperature sensitivity to precipitation during 2015–2100, which varies substantially in the original CMIP6 outputs, becomes systematically negative across SSPs after application of the EC, with absolute values between −1.10 °C mm−1 day and −3.52 °C mm−1 day, and with uncertainties reduced by 9.4% to 41.4%. The trend in tropical land-surface evapotranspiration, which was increasing by 0.292 mm yr−1 in the original CMIP6 model outputs, becomes significantly negative (−0.469 mm yr−1) after applying the constraint. Moreover, we find a significant increase of 58.7% in the leaf area index growth rate.

This study examines potential future changes of precipitation in China based on Coupled Model Intercomparison Project Phase 5 (CMIP5) climate model projections for the medium (RCP4.5) and high (RCP8.5) emission scenarios. We first evaluate the biases of climate model output and correct the biases through quantile mapping. After bias correction, we examine the changes in mean precipitation as well as shifts in its frequency distribution. We also evaluate the changes in extreme precipitation based on frequency analysis techniques. Our results show that by the end of the century, mean precipitation is going to increase by 8% (12%) under RCP4.5 (RCP8.5) scenarios, resulted from a combination of an increase in precipitation intensity and a slight decrease in precipitation frequency. Spatially, precipitation is projected to increase more in northern China than southern China, and the increase is the least in the southeast. Seasonally, precipitation is projected to increase more in fall and winter, and less in spring and summer. The precipitation intensity distribution is likely to shift towards more heavy events, with a decrease in the contribution from light events and a significant increase in contribution from heavy events. Extreme precipitation is going to increase at much higher rates than mean precipitation, and the increase is more spatially uniform. Changes in annual and seasonal precipitation are closely linked with temperature change. Total precipitation increases at 2.6% (1.9%) per degree warming under RCP4.5 (RCP8.5), but extreme precipitation has much higher sensitivities ranging 4.5–6.5% per degree warming for events of various return intervals. The percentage increase per degree is generally smaller for RCP8.5 than RCP4.5 scenarios, suggesting a reduced sensitivity at higher temperature. In addition, the precipitation increase seems to be linked with changes in the atmospheric circulations that transport moisture in different regions in China. These changes have significant implications for the management of water resources and water‐related hazards.

Recent studies have shown that the observed global warming trend over recent decades provides efficient constraints not only for future global mean temperature increases (ΔTgm) across Earth system models but also for changes in several climate variables that include significant ΔTgm–related uncertainty. However, ΔTgm–related emergent constraints (ECs) cannot reduce the uncertainty unrelated to ΔTgm. Here, to overcome this limitation, we develop an EC method and apply it to future changes in the annual maximum daily precipitation in order to reduce uncertainty therein. An EC for precipitation sensitivity based on historical extreme precipitation biases is combined with the constrained ΔTgm. This combined EC decreases the variance of the global mean precipitation by 42%, an improvement from only using temperature (resulting in 26% reduction), and the variance of regional precipitation by ≥ 30% in 24% of the globe (whereas ≥ 30% reduction is only seen in 2% of the globe with the temperature-related EC).

Future changes in precipitation due to climate change are of great concern to society. However, questions such as “Which weather systems will cause which changes?” and “Is the relative importance of these weather systems likely to change in the future?” have not been addressed fully yet. Here we present the first global estimates of the relative contributions of different weather systems (i.e., tropical cyclones, extratropical cyclones including fronts, and others) to changes in annual mean and extreme precipitation in the late 21st century using multimodel projections of the Coupled Model Intercomparison Project Phase 5. Although the models present biases in tropical cyclones over southern hemisphere, in particular, the representations of global weather system patterns are comparable to the reanalysis data. Total precipitation from tropical cyclones decreases (increases) in the tropics (subtropics) and that from extratropical cyclones including fronts decreases (increases) on the equatorial (poleward) side of the storm tracks. In addition, the mean intensity and frequency of system‐wise precipitation can change significantly even without considerable changes in annual amounts. We found that the subtropics, particularly in the Pacific and North Atlantic, are the regions where the proportions of precipitation by weather systems in annual mean and extreme precipitation display notable changes, suggesting distinct shifts in climate regimes. These regions have a common feature: they undergo the influence of several distinct weather systems in the present climate. In regions where climate regime shifts are projected, even the weather systems that have a minor contribution to precipitation in the present climate may cause considerable changes in annual and extreme precipitation.

This study examines the projected changes in mean and extreme precipitation over the Mediterranean (MED) and Sahara (SAH) regions based on the multi‐model ensemble mean of the Coupled Model Intercomparison Project Phase 6 (CMIP6) global climate model (GCM) datasets. The study employs robust statistical analyses to investigate future changes during 2015–2100 relative to a baseline period (1995–2014), under two Shared Socio‐economic Pathways (SSP) scenarios: SSP2‐4.5 and SSP5‐8.5. Selected indices from the Expert Team on Climate Change and Detection Indices are used in this study. They include those that represent maximum daily precipitation (RX1day), simple daily precipitation intensity (SDII), heavy precipitation days (R10mm), consecutive dry days (CDD), and consecutive wet days (CWD). Historical and projected daily precipitation is first bias‐adjusted using a quantile mapping approach before employing them to compute mean and extreme precipitation changes. The results demonstrate that the bias adjustment largely reduces the biases in the modelled mean and extreme precipitation over MED and SAH regions. Projections show a reduction in mean precipitation over most parts of the study region by the end of the 21st century. The areas encompassing Morocco and Algeria, and the Mediterranean area will experience the highest drying. The projected pattern agrees with the “wet gets wetter, dry gets drier” paradigm. The number of consecutive dry days and wet‐day intensity are also projected to increase and decrease, respectively. Under SSP5‐8.5, significant changes and the largest decrease in SDII attributed to global warming are projected in both regions. The reduction in mean precipitation, coupled with an increase in dry days, is likely to exacerbate the region's droughts and aridity situation and worsen the water scarcity status. Although there are uncertainties in the CMIP simulations, the findings support earlier studies based on varying datasets. This increases confidence in the output for decision‐making.

Changes in precipitation and temperature, especially in the Himalayan region, will have repercussions for socio-economic conditions in the future. Thus, this study aimed to understand the climatic trend and changes in one of the Himalayan River basins, i.e., Gandaki River Basin (GRB), Nepal. In particular, we analysed the historical (1985–2014) and projected (2015–2100) precipitation and temperature trend and their extremes using observation and 13 bias-corrected Coupled Model Intercomparison Project phase 6 (CMIP6) datasets. Additionally, the relationship between extreme precipitation/temperature indices and ocean-atmospheric circulation patterns were also analysed. The results showed an increasing trend of precipitation amount and temperature at annual and seasonal scales with the highest upward trend for precipitation in monsoon season and temperature in winter season. Among nine precipitation indices analysed, the wet extremes are projected to increase in all Shared Socioeconomic Pathways (SSP) scenarios; with the highest increment of high-intensity related extremes (R10 mm and R20 mm). In contrast, dry spells will decline in the distant-future (2075–2100) as compared to near (2015–2044) and mid-future (2045–2074). Further, increment in temperature trend resulted in a decrease in cold related temperature extremes and an increase in warm related extremes. Furthermore, it was observed that the changes in precipitation and temperature extremes over GRB were influenced by large-scale ocean-atmospheric circulation patterns. The Atlantic Multidecadal Oscillation (AMO), Sea Surface Temperature (SST) and Southern Oscillation Index (SOI) were found to have a major role in driving precipitation extremes while AMO, SST and Pacific Decadal Oscillation (PDO) have strong influence on temperature extremes. The results of this study will be useful for better understanding the implications of historical and future changes in precipitation and temperature and their extremes over the GRB.

Quantifying future precipitation changes in central Asia (CA), one of the largest semiarid‐to‐arid regions in the world, is increasingly attracting attention. Extreme precipitation changes in CA have been studied based on global climate models (GCMs) in the Coupled Model Intercomparison Project (CMIP). However, quantitative information regarding projected changes in extreme precipitation in CA under the 1.5–4°C global warming scenario is lacking; the potential advantages of the latest‐generation GCMs (i.e., CMIP6) in characterizing precipitation in CA have not been investigated. Thus, in this study, we evaluated the models' overall performance in reproducing the spatiotemporal pattern of four extreme precipitation indices and further performed an observationally constrained projection of changes in extreme precipitation indices for 1.5–4°C global warming based on a weighted multimodel ensemble. Regarding homologous models, CMIP6 models exhibited limited improvement relative to their earlier versions in CMIP5. We project that extreme precipitation indices in CA will increase approximately linearly as global warming increases, except for the consecutive dry days (CDD) index. The changes in precipitation intensity and accumulation exhibit robust consistency between models, whereas the signal of CDD changes is masked by the noise produced by intermodel uncertainties. The changes in the average annual accumulated and maximum 1‐day precipitation relative to the reference period (1981–2010) in CA are 12.0 and 14.2% at 3°C global warming (similar to late‐century warming projected based on current mitigation policies), respectively. Moreover, we demonstrate the advantage of the weighted scheme over the traditional unweighted scheme for multimodel ensemble projection.

Twenty-year temperature and precipitation extremes and their projected future changes are evaluated in an ensemble of climate models participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5), updating a similar study based on the CMIP3 ensemble. The projected changes are documented for three radiative forcing scenarios. The performance of the CMIP5 models in simulating 20-year temperature and precipitation extremes is comparable to that of the CMIP3 ensemble. The models simulate late 20th century warm extremes reasonably well, compared to estimates from reanalyses. The model discrepancies in simulating cold extremes are generally larger than those for warm extremes. Simulated late 20th century precipitation extremes are plausible in the extratropics but uncertainty in extreme precipitation in the tropics and subtropics remains very large, both in the models and the observationally-constrained datasets. Consistent with CMIP3 results, CMIP5 cold extremes generally warm faster than warm extremes, mainly in regions where snow and sea-ice retreat with global warming. There are tropical and subtropical regions where warming rates of warm extremes exceed those of cold extremes. Relative changes in the intensity of precipitation extremes generally exceed relative changes in annual mean precipitation. The corresponding waiting times for late 20th century extreme precipitation events are reduced almost everywhere, except for a few subtropical regions. The CMIP5 planetary sensitivity in extreme precipitation is about 6 %/°C, with generally lower values over extratropical land.

Climate change influences both average states and extremes in temperatures and precipitation. Southeast Asia, one of the most vulnerable regions worldwide to floods and heatwaves, indicates an escalation of the possibility of severe climate extremes. Extreme precipitation events can bring increasing floods, leading to considerable damage to property and human well-being. However, droughts also occur in Southeast Asia due to changing precipitation patterns in different regions. In big cities, such damages and risks are highly influential, given the large population density and the proximity to the coast or rivers.Global Earth system models predict the average trends and extremes of the climate system, including record-shattering extreme events that have exceeded previous records. I use the recently released Coupled Model Intercomparison Project Phase 6 (CMIP6), a multi-model large ensemble of climate predictions under different scenarios, to investigate changes in climate conditions in ten highly populated Southeast Asian cities. I first evaluated the CMIP6 simulations in the present day (2005-2014) in the ten cities and found good consistency in temperature and precipitation. Then, I examine changes in the mean, minimum, and maximum temperature and precipitation on an annual, monthly, and daily basis, respectively, under the high-emission SSP5-8.5 scenario. Furthermore, in the late century, annual maximum temperatures hit more than 40 °C in Bangkok, Chiang Mai, and Vientiane. Cities in our study are projected to experience 5-6 °C increases in temperature from November to April, indicating significant changes in the seasonal cycle. Precipitation increases significantly from May to October in most large cities in our study, except for Johor Bahru, Malaysia, where some summer precipitation reductions are expected. Yangon, in particular, is projected to increase more than 4 millimeters per day in July, indicating a very high challenge from flooding as a city facing flood risk every year at present.In summary, our results indicate significant changes in the mean and extreme states of temperature and precipitations in Southeast Asia. Based on these, I identified major physical risks of climate change among the ten cities. Decision-makers should build resilience to these risks to avoid significant damage. Yukiko Hirabayashi et al. (2013) discovered that the risks of floods increase due to the degree of warming. Increasing floods might lead to damage to households and other organisms. Moreover, a greater incidence of climate change, such as droughts, in Southeast Asia is caused by decreasing precipitation in regions (Teerachai Amnuaylojaroen and Pavinee Chanvichit (2019)).This investigation particularly utilizes CMIP6 model simulations to predict the future pattern of climate extremes based on historical and present statistics. CMIP6 performs well in reproducing the climatological spatial distribution of temperature and precipitation, with better performance for temperature than precipitation to predict near future and future climate extremes is that it (Yang et al. for China). With large-ensemble data for climate extremes, CMIP6 can predict more precise climate extremes. This model is necessary for research about extreme climate, as it addresses the climate system's natural variability in different regions with different topographic features.

Based on multiresource high-resolution in situ and satellite merged observations along with model simulations from the Coordinated Regional Climate Downscaling Experiment (CORDEX), this paper first investigated historical changes in extreme temperature and precipitation during the period of 1979–2018 in areas along the Sichuan–Tibet Railway, and then projected the future changes in the frequency and intensity of extreme temperature and precipitation under the representative concentration pathways (RCPs) of the RCP 4.5 and 8.5 scenarios. This study is expected to enhance our understanding of the spatiotemporal variability in the extreme temperature and precipitation along the Sichuan–Tibet Railway, and to provide scientific basis to advance the Sichuan–Tibet Railway construction and operation. The results show that temperatures in the Sichuan–Tibet region have shown a noticeable warming trend in the past 40 years, and the increase of minimum temperature is significantly higher than that of maximum temperature in the northwest of the region. The significant increase of precipitation is mainly distributed in the northwest of the Tibetan Plateau, and there is no significant change in precipitation in other areas along the Sichuan–Tibet Railway, except in Lhasa and its surrounding areas. However, precipitation along the Sichuan–Tibet Railway has shown a remarkable decrease in the past 20 yr. The warm days and nights have clearly increased by 6 and 5 day decade−1 for 1979–2019, while cold days and nights have markedly decreased by about 6.6 and 3.6 day decade−1, respectively. In the past 20 yr, the areas with increased precipitation from very wet days and extremely wet days are mainly distributed to the north of the Sichuan–Tibet Railway, while in the areas along the railway itself the very wet days and extremely wet days are decreasing. Under the RCP 4.5 and 8.5, the temperature in the Sichuan–Tibet region will increase significantly, and the frequency of extreme high (low) temperature events in the late 21st century (2070–2099) will greatly increase (decrease) by about 50%–80% (10%) compared with occurrences in the late 20th century (1970–1999). Meanwhile, the frequency of very wet days and extremely wet days in the Sichuan–Tibet region will increase by ~2%–19% and 2%–5%, respectively, and the areas along the Sichuan–Tibet Railway will be affected by more extreme high temperature and extreme precipitation events.

Global climate change will change the temporal and spatial distribution of precipitation, as well as the intensity and frequency of extreme precipitation. The Yangtze River basin is one of the world’s largest basins, and understanding the future precipitation changes should be vital to flood control, water resources supply, and hydropower electricity generation in this basin. In this study, projected future characteristics of precipitation are analyzed in the upper Yangtze river basin (UYRB). To this end, based on the observed data from national meteorological stations, the bias correction spatial downscaling (BCSD) of five models from the Coupled Model Intercomparison Project Phase 6 (CMIP6) is carried out. Then, based on the results of multi model ensemble (MME), we find that, relative to the historical period (1988–2014), the mean annual precipitation in the whole UYRB during 2015–2064 increases by 4.23%, 1.11%, 1.24% under SSP1-2.6, SSP2-4.5, SSP5-8.5, respectively, and it increases more in the long term (2040-2064) than that in the near term (2015–2039). Under SSP1–2.6, the precipitation will increase more significantly, which means lower emission of aerosols and greenhouse gases may increase the risk of flood disaster in the future over the UYRB. Interdecadal precipitation variability is more intense than interannual precipitation variability. Future precipitation changes in four seasons are spatially heterogeneous under three scenarios. Three extreme precipitation indices, including R95p, Rx1day and R10 mm, generally increase in the UYRB. R95p and Rx1day increase more in the WR and YBYCR basins with relatively high mean annual precipitation than that in other three sub-basins. R10 mm changes slightly in all sub-basins. The results reveal that the lower region of the UYRB may face greater risk of extreme precipitation. This study provides a timely updated finding about future changes in precipitation in the UYRB based on more accurate climate projections and ground-based observation.

Based on climate model outputs from Coupled Model Intercomparison Project Phase 5 (CMIP5), we investigated the global temperature and precipitation changes when global mean temperature rises by 1.5 and 2°C relative to the pre- industrial period (1861–1900). Multi-model ensemble mean (MME) shows that for the scenarios RCP2.6, RCP4.5, RCP6.0 and RCP8.5, global mean near-surface temperature (GMST) may reach 1.5°C warming relative to the pre- industrial level (1861–1900) around the year 2036, 2028, 2033 and 2025, respectively; and reach 2°C warming in 2049 (RCP4.5), 2056 (RCP6.0) and 2039 (RCP8.5). For RCP2.6 scenario, the 2°C warming could not be seen in the multi-model ensemble mean before 2100, but can be reached in different years in single model simulations including GFDL-CM3, IPSL-CM5A-LR, IPSL-CM5A-MR and MIROC-ESM. The timing when the GMST reaches a specific warming threshold is primarily related to radiative forcing and CO2-equivalent concentrations, which show similar value when the GMST rises by the same value. Projections under all RCPs from 25 CMIP5 models show that only 6 models under RCP2.6 scenario can project a warming lower than 1.5°C relative to the pre-industrial era before 2100, while the temperature increase under other three scenarios will be more than 1.5°C. It is therefore necessary to develop a set of lower emission scenarios to limit the global warming below 1.5°C. The investigations on inter-model differences show that the model with large transient climate response (TCR), also known as warm models, may reach the 1.5 and 2°C temperature increase earlier than those with low TCR (cold models), though other factors can also affect the model to reach 1.5 or 2°C temperature increase. For the future changes of temperature and precipitation at global scale, the MME results show little distinction among different scenarios when GMST rises by the same value, indicating that the global and regional response characteristics of temperature and precipitation are independent of the RCP scenarios definition (defined by the radiative forcing by 2100). Therefore, it is possible to investigate the changes in temperature increase 0.5 under RCP8.5 scenarios. The results show that regional responses are almost the same when GMST rises by each additional 0.5°C, indicating that the temperature and precipitation will essentially change linearly. These changes are characterized by more temperature increase in higher latitudes than in low latitudes, more temperature rise in land than in ocean, and increased precipitation in wet areas and decreased precipitation in dry areas, as was commonly detected in multiple studies. This suggests that the global warming impacts could be evaluated based on the multi-model ensemble projections under any RCP scenario with as much as a collection of models. This study also shows that in China, the regional mean temperature and precipitation changes are larger than the global mean when the GMST rises by 1.5 and 2°C. The temperature may increase across all China, with the warming increases from southeast to northwest. The precipitation will increase in most areas but it may decrease in the eastern part south of 30°N, based on the MME results.

Abstract. Single model initial-condition large ensembles (LEs) are a useful approach to understand the roles of forced responses and internal variability in historical and future climate change. Here, we produce one of the largest ensembles thus far using the MIROC6 coupled atmosphere–ocean global climate model (MIROC6-LE). The total experimental period of MIROC6-LE is longer than 76 000 years. MIROC6-LE consists of a long preindustrial control run, 50-member historical simulations, 8 single forcing historical experiments with 10 or 50 members, 5 future scenario experiments with 50 members and 3 single forcing future experiments with 50 members. Here, we describe the experimental design. The output data of most of the experiments are freely available to the public. This dataset would be useful to a wide range of research communities. We also demonstrate some examples of initial analyses. Specifically, we confirm that the linear additivity of the forcing-response relationship holds for the 1850–2020 trends of the annual mean values and extreme indices of surface air temperature and precipitation by analyzing historical fully forced runs and the sum of single forced historical runs. To isolate historical anthropogenic signals of annual mean and extreme temperature for 2000–2020 relative to 1850–1900, ensemble sizes of 4 and 15, respectively, are sufficient in most of the world. Historical anthropogenic signals of annual mean and extreme precipitation are significant with the 50-member ensembles in 76 % and 69 % of the world, respectively. Fourteen members are sufficient to examine differences in changes in annual mean values and extreme indices of temperature and precipitation between the shared socioeconomic pathways (ssp), ssp585 and ssp126, in most of the world. Ensembles larger than 50 members are desirable for investigations of differences in annual mean and extreme precipitation changes between ssp126 and ssp119. Historical and future changes in internal variability, represented by departures from the ensemble mean, are analyzed with a focus on the El Niño/Southern Oscillation (ENSO) and global annual mean temperature and precipitation. An ensemble size of 31 is large enough to detect ENSO intensification from preindustrial conditions to 1951–2000, from 1951–2000 to 2051–2100 in all future experiments, and from low- to high-emission future scenario experiments. The single forcing historical experiments with 27 members can isolate ENSO intensification due to anthropogenic greenhouse gas and aerosol forcings. Future changes in the global mean temperature variability are discernible with 23 members under all future experiments, while 50 members are not sufficient for detecting changes in the global mean precipitation variability in ssp119 and ssp126. We also confirm that these temperature and precipitation variabilities are not precisely analyzed when detrended anomalies from the long-term averages are used due to interannual climate responses to the historical natural forcing, which highlights the importance of large ensembles for assessing internal variability.