This study employs machine learning algorithms to construct Multi Model Ensembles (MMEs) based on Regional Climate Models (RCMs) within the Esca River basin in the Pyrenees. RCMs are ranked comprehensively based on their performance in simulating precipitation (pr), minimum temperature (tmin), and maximum temperature (tmax), revealing variability across seasons and influenced by the General Circulation Model (GCM) driving each RCM. The top-ranked approach is used to determine the optimal number of RCMs for MME construction, resulting in the selection of seven RCMs. Analysis of MME results demonstrates significant improvements in precipitation on both annual and seasonal scales, while temperature-related enhancements are more subtle at the seasonal level. The effectiveness of the ML–MME technique is highlighted by its impact on hydrological representation using a Temez model, yielding outcomes comparable to climate observations and surpassing results from Simple Ensemble Means (SEMs). The methodology is extended to climate projections under the RCP8.5 scenario, generating more realistic information for precipitation, temperature, and streamflow compared to SEM, thus reducing uncertainty and aiding informed decision-making in hydrological modeling at the basin scale. This study underscores the potential of ML–MME techniques in advancing climate projection accuracy and enhancing the reliability of data for basin-scale impact analyses.

. The Coupled Model Intercomparison Project (CMIP) coordinates community-based efforts to answer key and timely climate science questions, facilitate delivery of relevant multi-model simulations through shared infrastructure, and support national and international climate assessments. Generations of CMIP have evolved through extensive community engagement from punctuated phasing into more continuous support for the design of experimental protocols, infrastructure for data publication and access, and public delivery of climate information. We identify four fundamental research questions motivating a seventh phase of coupled model intercomparison relating to patterns of sea surface temperature change, changing weather, the water–carbon–climate nexus, and tipping points. Key CMIP7 advances include an expansion of baseline experiments, a focus on CO2-emissions-driven experiments, sustained support for community MIPs, periodic updating of historical forcings and diagnostics requests, and a collection of prioritized experiments, or the “Assessment Fast Track”, drawn from community MIPs to support climate research, assessment, and service goals across prediction and projection, characterization, attribution, and process understanding.

. By coordinating the design and distribution of global climate model simulations of the past, current, and future climate, the Coupled Model Intercomparison Project (CMIP) has become one of the foundational elements of climate science. However, the need to address an ever-expanding range of scientific questions arising from more and more research communities has made it necessary to revise the organization of CMIP. After a long and wide community consultation, a new and more federated structure has been put in place. It consists of three major elements: (1) a handful of common experiments, the DECK (Diagnostic, Evaluation and Characterization of Klima) and CMIP historical simulations (1850–near present) that will maintain continuity and help document basic characteristics of models across different phases of CMIP; (2) common standards, coordination, infrastructure, and documentation that will facilitate the distribution of model outputs and the characterization of the model ensemble; and (3) an ensemble of CMIP-Endorsed Model Intercomparison Projects (MIPs) that will be specific to a particular phase of CMIP (now CMIP6) and that will build on the DECK and CMIP historical simulations to address a large range of specific questions and fill the scientific gaps of the previous CMIP phases. The DECK and CMIP historical simulations, together with the use of CMIP data standards, will be the entry cards for models participating in CMIP. Participation in CMIP6-Endorsed MIPs by individual modelling groups will be at their own discretion and will depend on their scientific interests and priorities. With the Grand Science Challenges of the World Climate Research Programme (WCRP) as its scientific backdrop, CMIP6 will address three broad questions: – How does the Earth system respond to forcing? – What are the origins and consequences of systematic model biases? – How can we assess future climate changes given internal climate variability, predictability, and uncertainties in scenarios? This CMIP6 overview paper presents the background and rationale for the new structure of CMIP, provides a detailed description of the DECK and CMIP6 historical simulations, and includes a brief introduction to the 21 CMIP6-Endorsed MIPs.

Multi-Model Ensembles (MMEs) are used for improving the performance of GCM simulations. This study evaluates the performance of MMEs of precipitation, maximum temperature and minimum temperature over a tropical river basin in India developed by various techniques like arithmetic mean, Multiple Linear Regression (MLR), Support Vector Machine (SVM), Extra Tree Regressor (ETR), Random Forest (RF) and long short-term memory (LSTM). The 21 General Circulation Models (GCMs) from National Aeronautics Space Administration (NASA) Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) dataset and 13 GCMs of Coupled Model Inter-comparison Project, Phase 6 (CMIP6) are used for this purpose. The results of the study reveal that the application of a LSTM model for ensembling performs significantly better than models in the case of precipitation with a coefficient of determination (R) value of 0.9. In case of temperature, all the machine learning (ML) methods showed equally good performance, with RF and LSTM performing consistently well in all the cases of temperature with R value ranging from 0.82 to 0.93. Hence, based on this study RF and LSTM methods are recommended for creation of MMEs in the basin. In general, all ML approaches performed better than mean ensemble approach.© 2022. The Author(s).

Heat stress harms human health, agriculture, the economy, and the environment more broadly. Exposure to heat stress is increasing with rising global temperatures. While most studies assessing future heat stress have focused on surface air temperature, compound extremes of heat and humidity are key drivers of heat stress. Here, we use atmospheric reanalysis data and a large initial-condition ensemble of global climate model simulations to evaluate future changes in daily compound heat-humidity extremes as a function of increasing global-mean surface air temperature (GSAT). The changing frequency of heat-humidity extremes, measured using wet bulb globe temperature (WBGT), is strongly related to GSAT and, conditional upon GSAT, nearly independent of forcing pathway. The historical ∼1°C of GSAT increase above preindustrial levels has already increased the population annually exposed to at least one day with WBGT exceeding 33°C (the reference safety value for humans at rest per the ISO-7243 standard) from 97 million to 275 million. Maintaining the current population distribution, this exposure is projected to increase to 508 million with 1.5°C of warming, 789 million with 2.0°C of warming, and 1.22 billion with 3.0°C of warming (similar to late-century warming projected based on current mitigation policies).

North Africa is considered a climate change hot spot. Existing studies either focus on the physical aspects of climate change or discuss the social ones. The present article aims to address this divide by assessing and comparing the climate change vulnerability of Algeria, Egypt, Libya, Morocco, and Tunisia and linking it to its social implications. The vulnerability assessment focuses on climate change exposure, water resources, sensitivity, and adaptive capacity. The results suggest that all countries are exposed to strong temperature increases and a high drought risk under climate change. Algeria is most vulnerable to climate change, mainly due to the country’s high sensitivity. Across North Africa, the combination of climate change and strong population growth is very likely to further aggravate the already scarce water situation. The so-called Arab Spring has shown that social unrest is partly caused by unmet basic needs of the population for food and water. Thus, climate change may become an indirect driver of social instability in North Africa. To mitigate the impact of climate change, it is important to reduce economic and livelihood dependence on rain-fed agriculture, strengthen sustainable land use practices, and increase the adaptive capacity. Further, increased regional cooperation and sub-national vulnerability assessments are needed.

This study compares the bias correction techniques of empirical quantile mapping (QM) and the Long Short-Term Memory (LSTM) machine learning model for summertime daily rainfall simulation focusing on precipitation-dependent bias and temporal variation. Numerical experiments using Weather Research and Forecasting (WRF) were conducted over South Korea with lateral boundary conditions of ERA5 reanalysis data. For the spatial distribution of mean summertime rainfall, the bias-uncorrected WRF simulation (WRF_RAW) showed dry bias for most of the region of South Korea. The WRF results corrected by QM and LSTM (WRF_QM and WRF_LSTM, respectively) were improved for the mean summer rainfall simulation with the root mean square error values of 0.17 and 0.69, respectively, which were smaller than those of the WRF_RAW (1.10). Although the WRF_QM performed better than the WRF_LSTM in terms of the summertime mean and monthly precipitation, the WRF_LSTM presented a closer interannual rainfall variation to the observation than the WRF_QM. The coefficient of determination for calendar-day mean rainfall was the highest in the following order: the WRF_LSTM (0.451), WRF_QM (0.230), and WRF_RAW (0.201). However, the WRF_LSTM had a limitation in reproducing extreme rainfall exceeding 50 mm/day due to the few cases of extreme precipitation in training data. Nevertheless, the WRF_LSTM better simulated the observed light-to-moderate precipitation (10–50 mm/day) than the others.

A diagram has been devised that can provide a concise statistical summary of how well patterns match each other in terms of their correlation, their root‐mean‐square difference, and the ratio of their variances. Although the form of this diagram is general, it is especially useful in evaluating complex models, such as those used to study geophysical phenomena. Examples are given showing that the diagram can be used to summarize the relative merits of a collection of different models or to track changes in performance of a model as it is modified. Methods are suggested for indicating on these diagrams the statistical significance of apparent differences and the degree to which observational uncertainty and unforced internal variability limit the expected agreement between model‐simulated and observed behaviors. The geometric relationship between the statistics plotted on the diagram also provides some guidance for devising skill scores that appropriately weight among the various measures of pattern correspondence.

Two different bias correction methods, the quantile mapping (QM) and quantile delta mapping (QDM), are applied to simulated daily temperature and precipitation over China from a set of 21st century regional climate model (the ICTP RegCM4) projections. The RegCM4 is driven by five different general circulation models (GCMs) under the representative concentration pathway RCP4.5 at a grid spacing of 25 km using the CORDEX East Asia domain. The focus is on mean temperature and precipitation in December–January–February (DJF) and June–July–August (JJA). The impacts of the two methods on the present day biases and future change signals are investigated. Results show that both the QM and QDM methods are effective in removing the systematic model biases during the validation period. For the future changes, the QDM preserves the temperature change signals well, in both magnitude and spatial distribution, while the QM artificially modifies the change signal by decreasing the warming and modifying the patterns of change. For precipitation, both methods preserve the change signals well but they produce greater magnitude of the projected increase, especially the QDM. We also show that the effects of bias correction are variable- and season-dependent. Our results show that different bias correction methods can affect in different way the simulated change signals, and therefore care has to be taken in carrying out the bias correction process.

Evaluating global climate model (GCM) outputs is essential for accurately simulating future hydrological cycles using hydrological models. The GCM multi-model ensemble (MME) precipitation simulations of the Climate Model Intercomparison Project Phases 5 and 6 (CMIP5 and CMIP6, respectively) were spatially and temporally downscaled according to a multi-site statistical downscaling method for the Hanjiang River Basin (HRB), China. Downscaled precipitation accuracy was assessed using data collected from 14 meteorological stations in the HRB. The spatial performances, temporal performances, and seasonal variations of the downscaled CMIP5-MME and CMIP6-MME were evaluated and compared with observed data from 1970–2005. We found that the multi-site downscaling method accurately downscaled the CMIP5-MME and CMIP6-MME precipitation simulations. The downscaled precipitation of CMIP5-MME and CMIP6-MME captured the spatial pattern, temporal pattern, and seasonal variations; however, precipitation was slightly overestimated in the western and central HRB and precipitation was underestimated in the eastern HRB. The precipitation simulation ability of the downscaled CMIP6-MME relative to the downscaled CMIP5-MME improved because of reduced biases. The downscaled CMIP6-MME better simulated precipitation for most stations compared to the downscaled CMIP5-MME in all seasons except for summer. Both the downscaled CMIP5-MME and CMIP6-MME exhibit poor performance in simulating rainy days in the HRB.

Projecting the climate change impacts on hydrology and water resources relies on the climate scenarios simulated by general circulation models (GCMs), which requires a systematic and comprehensive assessment of the GCMs’ simulation performances at a regional scale. This study evaluates the performances of precipitation simulation over the Hanjiang River Basin (HRB) by six climate models from Phase 6 of the Coupled Model Intercomparison Project (CMIP6), the corresponding six previous models from the CMIP5, and their multi-model ensemble (MME) based on the observational data in the CN05.1. To our knowledge, this is the first preliminary study in the HRB. The Taylor diagram (including standard deviation, root-mean-square difference, and correlation coefficient) and Taylor skill score are used for the evaluation of GCMs’ precipitation simulation performances. The spatial pattern and temporal pattern over the HRB simulated by CMIP6 and CMIP5 models are compared by relative biases. The results of the Taylor diagram and skill score show that CMIP6 models don't necessarily perform better than the corresponding previous CMIP5 models in simulating precipitation over the HRB. The MME exhibits superior performance compared to that of any individual model, and the CMIP6-MME is more skillful than the CMIP5-MME. As to the spatial and temporal variation characteristics, the precipitation biases are both present in CMIP6 and CMIP5 models, and the bias of the CMIP6-MME is lower than that of the CMIP5-MME. The CMIP6 and CMIP5 models overestimate the precipitation from January to June, and simulate larger precipitation biases in the areas and seasons with less precipitation, while they are lower with more precipitation over the HRB. The findings obtained in this study could provide a scientific reference for the research of future hydrological cycle predictions over the HRB.

. The major droughts and floods in the Hanjiang River\nBasin, central China, have a significant impact on flood prevention and\ncontrol in the middle reaches of the Yangtze River and water resources\nmanagement in the areas of the South–North Water Diversion Middle Line\nProject. However, there is a lack of understanding of the multi-decadal to\ncentennial-scale patterns of extreme droughts and floods in the area.\nApplying the yearly drought and flood records from historical documents and\nprecipitation data in the period of instrumental measurements, this study\nconstructs a time series of extreme droughts and floods in the Hanjiang\nRiver Basin from 1426–2017 and analyzes the temporal and spatial\ncharacteristics of the extreme drought and flood event variations. The results\nshow that there were a total of 45 extreme droughts and 52 extreme floods in the\nbasin over the past 592 years. Extreme droughts and floods were highly\nvariable on a multi-decadal to centennial scale, and the frequencies were\nhigher in the first and last 100 years or so of the study period and\nlower in between. Spatially, the frequencies of extreme droughts and floods\nwere generally higher in the middle and lower reaches than in the upper\nreaches. It was also found that there is a good correlation of drought and\nflood frequencies between the upper Hanjiang River Basin and North China.\nThese results are informative for the study of mechanisms and predictability\nof multi-decadal to centennial-scale variability of extreme hydroclimatic\nevents in the river basin.

Based on precipitation extremes calculated from high-resolution daily observational data in East China during 1961–2014, the performance of 34 climate models from phase 6 of the Coupled Model Intercomparison Project (CMIP6) are assessed in terms of climatology and interannual variability. Four extreme precipitation indices, including the total precipitation (Prcptot), the total precipitation for events exceeding the 95th percentile (R95p), and the maximum of 1-day (Rx1day) and 5-day (Rx5day) precipitation, are analyzed. Results show that the CMIP6 models demonstrate good performances in reproducing the climatological spatial distribution and interannual variability of precipitation extremes, with the best from Prcptot. Based on an integrated assessment of the above two factors, the models that perform relatively well for all four extreme precipitation indices are GFDL-CM4, MIROC6, EC-Earth3-Veg, EC-Earth3, and EC-Earth3-CC. Furthermore, the optimal multi-model ensemble (A-MME) constructed from a selection of the most skillful models shows improved behavior compared to the all-model ensemble. The wet (dry) biases over the northern (southern) region of East China are all decreased. This may benefit from the improvement that A-MME can reproduce well the characteristics of moisture and vertical velocity.