Figure 6 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Spread of mean total bias in the 95th quantile (Q95) of CBI and its contribution from T , RH, and their copula for individual CMIP5 models (a), and a scatter plot of the T and RH contributions to Q95 CBI bias, with their Kendall rank correlation coefficient (p value < 0.001) (b). Shown are the results for a grid point in Brazil (Amazonia, 5 • S and 56.5 • W). Bias was calculated as (CMIP5 or transformation minus ERA-Interim). Equal axes are used in panel (b) to highlight the differences in spread between both bias components.
Source publication
Climate models' outputs are affected by biases that need to be detected and adjusted to model climate impacts. Many climate hazards and climate-related impacts are associated with the interaction between multiple drivers, i.e. by compound events. So far climate model biases are typically assessed based on the hazard of interest, and it is unclear h...
Contexts in source publication
Context 1
... in the CMIP5 model CBI is 21 • C, which is large compared to the area-weighted mean CBI in ERA-Interim of 84 • C (i.e. corresponding to 25 %). In fact, the stippling over 75 % of land masses in Fig. 5c indicates that the models differ significantly from ERA-Interim. The bias in RH is the main contributor to total mean bias in extreme CBI values (Fig. 6d-f). The relevance of RH for the bias in CBI is visible from the similarities in magnitude and spatial distribution of bias between Fig. 5c and e. Furthermore, while the area-weighted mean of absolute bias in CBI is 21 • C, the corresponding mean biases due to T , RH, and the dependency between them are 3, 20, and 3 • C, respectively. The ...
Context 2
... in Fig. 5d and f). A closer examination of the bias decomposition results shows, for a site with large positive bias in Brazil, that the results shown in the multimodal mean bias plots (Fig. 5c and e) reflect intermodel model behaviour at the local level. That is, CMIP5 models with high RH bias contributions also show high overall CBI bias (Fig. 6a). At this location, there is a positive intermodel correlation between the biases driven by T and RH (τ = 0.82; Fig. 6b). Such behaviour is due to the combination of the following two reasons: (1) a negative intermodel correlation between the biases in T and RH, i.e. CMIP5 models simulating temperatures that are too high also tend to ...
Context 3
... in Brazil, that the results shown in the multimodal mean bias plots (Fig. 5c and e) reflect intermodel model behaviour at the local level. That is, CMIP5 models with high RH bias contributions also show high overall CBI bias (Fig. 6a). At this location, there is a positive intermodel correlation between the biases driven by T and RH (τ = 0.82; Fig. 6b). Such behaviour is due to the combination of the following two reasons: (1) a negative intermodel correlation between the biases in T and RH, i.e. CMIP5 models simulating temperatures that are too high also tend to simulate relative humidity that is too low (as discussed by Fischer and Knutti, 2013); and (2) the fact that CBI is high ...
Context 4
... compensation of the biases in individual models arises from (1) opposite biases in T and RH (models simulating temperatures that are too high also tend to simulate relative humidity that is too low; Fischer and Knutti, 2013) and (2) the WBGT tendency to be high (low) for humid and warm (dry and cold) conditions. Figure A6 illustrates such a cancellation of the bias in WBGT for a location in South Africa, where the negative dependency between T and RH leads to a small bias in WBGT. In this location, the model biases driven by T are negative; therefore those driven by RH are positive. ...
Context 5
... in the CMIP5 model CBI is 21 • C, which is large compared to the area-weighted mean CBI in ERA-Interim of 84 • C (i.e. corresponding to 25 %). In fact, the stippling over 75 % of land masses in Fig. 5c indicates that the models differ significantly from ERA-Interim. The bias in RH is the main contributor to total mean bias in extreme CBI values (Fig. 6d-f). The relevance of RH for the bias in CBI is visible from the similarities in magnitude and spatial distribution of bias between Fig. 5c and e. Furthermore, while the area-weighted mean of absolute bias in CBI is 21 • C, the corresponding mean biases due to T , RH, and the dependency between them are 3, 20, and 3 • C, respectively. The ...
Context 6
... closer examination of the bias decomposition results shows, for a site with large positive bias in Brazil, that the results shown in the multimodal mean bias plots (Fig. 5c and e) reflect intermodel model behaviour at the local level. That is, CMIP5 models with high RH bias contributions also show high overall CBI bias (Fig. 6a). At this location, there is a positive intermodel correlation between the biases driven by T and RH (τ = 0.82; Fig. 6b). Such behaviour is due to the combination of the following two reasons: (1) a negative intermodel correlation between the biases in T and RH, i.e. CMIP5 models simulating temperatures that are too high also tend to ...
Context 7
... in Brazil, that the results shown in the multimodal mean bias plots (Fig. 5c and e) reflect intermodel model behaviour at the local level. That is, CMIP5 models with high RH bias contributions also show high overall CBI bias (Fig. 6a). At this location, there is a positive intermodel correlation between the biases driven by T and RH (τ = 0.82; Fig. 6b). Such behaviour is due to the combination of the following two reasons: (1) a negative intermodel correlation between the biases in T and RH, i.e. CMIP5 models simulating temperatures that are too high also tend to simulate relative humidity that is too low (as discussed by Fischer and Knutti, 2013); and (2) the fact that CBI is high ...
Context 8
... compensation of the biases in individual models arises from (1) opposite biases in T and RH (models simulating temperatures that are too high also tend to simulate relative humidity that is too low; Fischer and Knutti, 2013) and (2) the WBGT tendency to be high (low) for humid and warm (dry and cold) conditions. Figure A6 illustrates such a cancellation of the bias in WBGT for a location in South Africa, where the negative dependency between T and RH leads to a small bias in WBGT. In this location, the model biases driven by T are negative; therefore those driven by RH are positive. ...
Citations
... Multiple impacts of climate 48 hazards such as mean warming, heatwaves, drought, floods, wildfires or changes in 49 natural land cover are also traceable on human-related sectors such as human health, 50 water, food, economy, infrastructure or security [26]. Nevertheless, many papers have 51 exhibited large uncertainties related to these cross-sectoral climate impacts and 52 September 13, 2023 2/19 population exposure analysis. It was shown that larger uncertainties arise from the 53 impact models than from the climate models [27], and the spread across different impact 54 models seems to be a major component of climate impacts simulations uncertainty [28]. ...
... Previous papers have shown that climate models with similar global mean warming 362 can lead to different aggregated impacts, so that climate model uncertainties remain a 363 significant contributor to sectoral and cross-sectoral analysis uncertainties [43]. Though 364 out of the scope of this paper, a further step could be to investigate these uncertainties 365 and assess ESMs × GIMs simulations contribution to the total multi-model TOE changes [48,51,52]. ...
Consequently to global warming, multi-sectoral impacts are observed and should intensify in the future, affecting sectors of water resources, agriculture, weather extremes and health. Related projected change signs, their possible emergences from the historical variability, and how these emergences may cumulate in time and space could result in severe risks or great benefits for local populations. Using the world's largest cross-sectoral climate-related impact multi-model simulations database Inter-Sectoral Impacts Model Intercomparison Project (ISIMIP), here we quantify for the first time the Time of Emergence (TOE) of historical and future simulated changes in multiple climate-related indicators at a global scale. We assess how both adverse and positive changes (i.e. risks and benefits) could cumulate during the century. Based on ISIMIP2b (ISIMIP phase 2b) and a low mitigation future scenario (RCP6.0), we find that most land areas are characterized by a multi-model TOE earlier than 2010 for the majority of the 10 analyzed multi-sectoral indicators. This illustrates an already reached new equilibrium-state resulting from global warming for many sectors including hydrology, agriculture and weather extremes. However, the TOE varies depending on the region and the sectoral indicator, encompassing both projected risks and benefits. The largest number of cumulated emergences of cross-sectoral risks is projected in the tropics and include detected TOE for higher heat stress extremes, increased flood and drought risks, and declining crop yields. Conversely, northern mid-and high-latitudes experience the greatest number of cumulated emergences of benefits, primarly associated with future crop yields increase. These cumulative emergences of risks (in the tropics) and benefits (in northern latitudes) reach their peak before 2050, indicating an early impact and related emergence on multiple sectors. Nevertheless, substantial uncertainties are exhibited in this multi-model assessment of TOE, particularly in the tropics due to a cascade of uncertainties combining climate models, impacts models, aggregated cross-sectoral impacts, and TOE detection. This study brings time constrains of multiple climate-related changes. It particularly highlights the heightened and early cumulated risks projected in tropical regions that may further exacerbate disparities and inequalities with northern latitudes. These results confirm the tropics paradox where fewer greenhouse gas emissions correspond to more adverse impacts from global warming. Without further mitigation and adaptation strategies, the vulnerable socioeconomic conditions and limited resources in these areas will amplify the negative consequences arising from early and cumulative cross-sectoral emergences, both for populations and ecosystems.
... tion of their individual biases to the associated multivariate hazard index may lead to different results. For instance, biases in wet-bulb globe temperature (WBGT) are found to be smaller than in the Chandler burning index (CBI) for a given model output, yet both indices are based on temperature and relative humidity (Villalobos-Herrera et al., 2021). This is attributed to the construction of the index, since bias in CBI is mainly driven by the bias in relative humidity, whereas bias in WBGT interplays between biases in temperature and relative humidity, which compensate each other. ...
Bias correction (BC) is often a necessity to improve the applicability of global and regional climate model (GCM and RCM, respectively) outputs to impact assessment studies, which usually depend on multiple potentially dependent variables. To date, various BC methods have been developed which adjust climate variables separately (univariate BC) or jointly (multivariate BC) prior to their application in impact studies (i.e., the component-wise approach). Another possible approach is to first calculate the multivariate hazard index from the original, biased simulations and bias-correct the impact model output or index itself using univariate methods (direct approach). This has the advantage of circumventing the difficulties associated with correcting the inter-variable dependence of climate variables which is not considered by univariate BC methods. Using a multivariate drought index (i.e., standardized precipitation evapotranspiration index – SPEI) as an example, the present study compares different state-of-the-art BC methods (univariate and multivariate) and BC approaches (direct and component-wise) applied to climate model simulations stemming from different experiments at different spatial resolutions (namely Coordinated Regional Climate Downscaling Experiment (CORDEX), CORDEX Coordinated Output for Regional Evaluations (CORDEX-CORE), and 6th Coupled Intercomparison Project (CMIP6)). The BC methods are calibrated and evaluated over the same historical period (1986–2005). The proposed framework is demonstrated as a case study over a transboundary watershed, i.e., the Upper Jhelum Basin (UJB) in the Western Himalayas. Results show that (1) there is some added value of multivariate BC methods over the univariate methods in adjusting the inter-variable relationship; however, comparable performance is found for SPEI indices. (2) The best-performing BC methods exhibit a comparable performance under both approaches with a slightly better performance for the direct approach. (3) The added value of the high-resolution experiments (CORDEX-CORE) compared to their coarser-resolution counterparts (CORDEX) is not apparent in this study.
... SMILEs have already helped substantially to advance our understanding of climate variability 24 ; however, climate models can have biases that should be considered carefully. That means climate model skills need to be evaluated 43,86,99 . Assuming that any difference between model-and observation-based estimates is only due to biases can be misleading as, especially for multivariate relationships, large differences may arise from internal climate variability 43,54 . ...
Societally relevant weather impacts typically result from compound events, which are rare combinations of weather and climate drivers. Focussing on four event types arising from different combinations of climate variables across space and time, here we illustrate that robust analyses of compound events — such as frequency and uncertainty analysis under present-day and future conditions, event attribution to climate change, and exploration of low-probability-high-impact events — require data with very large sample size. In particular, the required sample is much larger than that needed for analyses of univariate extremes. We demonstrate that Single Model Initial-condition Large Ensemble (SMILE) simulations from multiple climate models, which provide hundreds to thousands of years of weather conditions, are crucial for advancing our assessments of compound events and constructing robust model projections. Combining SMILEs with an improved physical understanding of compound events will ultimately provide practitioners and stakeholders with the best available information on climate risks.
... Cannon, 2018;François et al., 2020). This could have major consequences on the ability of climate models to simulate CEs accurately Villalobos-Herrera et al., 2021;Vrac et al., 2022a;Ridder et al., 2021), and then on the resulting analyses involved in decision-making processes. A few multivariate bias correction methods, i.e. statistical methods that are able to adjust both univariate and multivariate properties of simulations with respect to reference datasets, have been recently developed (e.g. ...
Many climate-related disasters often result from a combination of several climate phenomena, also referred to as “compound events’’ (CEs). By interacting with each other, these phenomena can lead to huge environmental and societal impacts, at a scale potentially far greater than any of these climate events could have caused separately. Marginal and dependence properties of the climate phenomena forming the CEs are key statistical properties characterising their probabilities of occurrence. In this study, we propose a new methodology to assess the time of emergence of CE probabilities, which is critical for mitigation strategies and adaptation planning. Using copula theory, we separate and quantify the contribution of marginal and dependence properties to the overall probability changes of multivariate hazards leading to CEs. It provides a better understanding of how the statistical properties of variables leading to CEs evolve and contribute to the change in their occurrences. For illustrative purposes, the methodology is applied over a 13-member multi-model ensemble (CMIP6) to two case studies: compound wind and precipitation extremes over the region of Brittany (France), and frost events occurring during the growing season preconditioned by warm temperatures (growing-period frost) over central France. For compound wind and precipitation extremes, results show that probabilities emerge before the end of the 21st century for six models of the CMIP6 ensemble considered. For growing-period frosts, significant changes of probability are detected for 11 models. Yet, the contribution of marginal and dependence properties to these changes in probabilities can be very different from one climate hazard to another, and from one model to another. Depending on the CE, some models place strong importance on both marginal properties and dependence properties for probability changes. These results highlight the importance of considering changes in both marginal and dependence properties, as well as their inter-model variability, for future risk assessments related to CEs.
... A multivariate drought index i.e., standardized precipitation evapotranspiration index-SPEI (Vicente-Serrano et al., 2010), is widely used to monitor and assess drought and their sectorial impacts under global warming conditions. It can be interpreted as the number of standard deviations by which the observed anomaly deviates from the long-term mean. ...
... It can be interpreted as the number of standard deviations by which the observed anomaly deviates from the long-term mean. Various researchers highlighted its suitability to detect the onset and spatio-temporal evolution of drought at 35 the regional to global scales Ansari and Grossi, 2022), and recommended it for operational drought monitoring (Vicente-Serrano et al., 2010). ...
... critical (Vicente-Serrano et al., 2010). A study conducted by Beguería et al. (2014) compared the SPEI values using three different methods for PET estimation (Penman-Monteith, Hargreaves, and Thornthwaite) and found small differences in humid regions. ...
Bias correction (BC) is often a necessity to improve the applicability of global and regional climate model (GCM and RCM, respectively) outputs to impact assessment studies, which usually depend on multiple potentially dependent variables. To date, various BC methods have been developed which adjust climate variables separately (univariate BC) or jointly (multivariate BC) prior to their application in impact studies (i.e., the component-wise approach). Another possible approach is to first calculate the multivariate hazard index from the original, biased simulations, and bias-correct the impact model output or index itself using univariate methods (direct approach). This has the advantage of circumventing the difficulties associated with correcting the inter-variable dependence of climate variables which is not considered by univariate BC methods. Using a multivariate drought index (i.e., SPEI) as an example, the present study compares different state-ofthe- art BC methods (univariate and multivariate) and BC approaches (direct and component-wise) applied to climate model simulations stemming from different experiments at different spatial resolutions (namely CORDEX, CORDEX-CORE and CMIP6). The BC methods are calibrated and evaluated over the same historical period (1986–2005). The proposed framework is demonstrated as a case study over a transboundary watershed, i.e. the Upper Jhelum Basin (UJB) in the Western Himalaya. Results show that (1) there is some added value of multivariate BC methods over the univariate methods in adjusting the inter-variable relationship, however, comparable performance is found for SPEI indices. (2) The best performing BC methods exhibits a comparable performance under both approaches with a slightly better performance for the direct approach. (3) The added value of the high-resolution experiments (CORDEX-CORE) compared to their coarser resolution counterparts (CORDEX) are not apparent in this study.
... Therefore, formal detection and attribution analysis concerning compound wind and precipitation extremes are still lacking and worthy of further efforts. Such studies should rely on models that satisfactorily simulate not only the statistics of univariate extremes but also, more importantly, their complex dependence for the right physics (Villalobos-Herrera et al., 2021;Zscheischler et al., 2021). In this context, a compound event-oriented model evaluation is needed to optimize the model selection, which would also help constrain societally relevant CWPE projections (Villalobos-Herrera et al., 2021). ...
... Such studies should rely on models that satisfactorily simulate not only the statistics of univariate extremes but also, more importantly, their complex dependence for the right physics (Villalobos-Herrera et al., 2021;Zscheischler et al., 2021). In this context, a compound event-oriented model evaluation is needed to optimize the model selection, which would also help constrain societally relevant CWPE projections (Villalobos-Herrera et al., 2021). ...
Plain Language Summary
Co‐occurrences of wind extremes and precipitation extremes, termed compound wind and precipitation extremes (CWPEs), can disrupt and endanger shipment and shipping logistics. The associated winds and floods may cause severe socio‐economic impacts in coastal and inland areas, such as paralyzed public transportation, critical infrastructure damages, and fatalities. CWPEs substantially impact the Indo‐Pacific region as a result of frequent occurrences of extreme weather conditions. Results show that the northwestern Pacific Ocean and its coasts have experienced the most frequent, severe, and longest‐lasting CWPEs in summer in recent decades, which are induced by cyclones. Landfalling atmospheric rivers are one of the main drivers for frequent occurrences of CWPEs in central and western China and the northwestern Indo‐China Peninsula in both boreal summer and winter. The equatorial tropical area features significant increasing trends in the frequency and intensity of CWPEs, mainly resulting from other processes free of cyclones and atmospheric rivers, while southern China shows significant decreasing trends for both seasons. Overall, the mapping of CWPE hotspots and understanding of long‐term changes and drivers of CWPEs help vulnerable communities better prepare for combined and amplified hazards related to wind and precipitation extremes.
... We found that the use of univariate bias adjustment can lead to an over-or underestimation of future climate-related fire risks, despite the overall clear sign of climate-change induced risks, regardless of the bias-correction method employed. In fact, the way how climate models' biases affect the final risks is still a topic of discussion in the scientific community [59] and the way how changes in the dependence of the drivers will affect compound events in a warming world is a challenging task in assessing future climate risk of complex events. ...
Tropical fire activity closely follows the co-occurrence of multiple climate stressors. Yet, it remains challenging to quantify how changes in climate alter the likelihood of fire risks associated with compound events. Recent abrupt changes in fire regimes in iconic landscapes in Brazil (namely the Pantanal and Xingu) provide a key opportunity to evaluate how extremely dry and hot conditions, both together and individually, have influenced the probability of large fires. Here we quantify the relationships between climate and fire across these regions and provide evidence on the extent to which fire risk and the associated impacts could be constrained if anthropogenic global warming is limited. We investigate the burned area, differentiating between fire types according to land use (forest fires, savanna fires, farming fires and grassland and wetland fires), and derive present and future fire risks linked to multiple climate variables. We show that concurrent air dryness (high vapour-pressure deficit – VPD) and low precipitation have driven fire occurrence in both Xingu and the Pantanal, with VPD playing a dominant role. Historical climatic change has already increased compound event-related (CE-related) fire risks of all fire types (5%-10%), and these risks are likely to increase in the future due to global warming. The likelihood of CE-related increase in fire risk may be reduced by up to 14.4% if global warming is constrained to +1.5°C instead of +3°C. Nevertheless, substantially increased CE-related fire risks are still expected even if restricting global mean warming to 1.5 °C, particularly in the Pantanal. We thus conclude that climate action should be coordinated with environmental protection to reduce ignition sources and promote effective conservation measures to preserve these biomes.
... Climate model evaluation is usually univariate without considering the multivariate nature of multiple hazards, it is thus important to evaluate the biases in the dependence between the contributing variables in climate models (Vezzoli et al., 2017). However, rare studies have evaluated the climate model multivariate representation of hazard indicators Villalobos-Herrera et al., 2021;Zscheischler et al., 2021). ...
... 4. Advanced physically based models, such as crop and hydrodynamical flooding models, are important for studying the dynamics of many compound events, especially under climate change conditions. In fact, simple statistical models, such as regression, may not provide accurate information about future changes that are outside the observational range, and more sophisticated approaches from extreme value theory are required (Engelke & Ivanovs, 2021). 5. Compound event occurrence probabilities and their future changes can be highly uncertain due to combined uncertainties in multiple drivers and associated interplay (Bevacqua, Vousdoukas, Zappa, et al., 2020;Santos et al., 2021;Villalobos-Herrera et al., 2021). In this context, event-based storylines, which explore consequences of high-impact plausible events either or both under present and future climates putting emphasis on plausibility rather than probability, can provide very effective information for improving emergency preparedness to compound events (Sillmann et al., 2021). ...
Compound weather and climate events are combinations of climate drivers and/or hazards that contribute to societal or environmental risk. Studying compound events often requires a multidisciplinary approach combining domain knowledge of the underlying processes with, for example, statistical methods and climate model outputs. Recently, to aid the development of research on compound events, four compound event types were introduced, namely (1) preconditioned, (2) multivariate, (3) temporally compounding, and (4) spatially compounding events. However, guidelines on how to study these types of events are still lacking. Here, we consider four case studies, each associated with a specific event type and a research question, to illustrate how the key elements of compound events (e.g., analytical tools and relevant physical effects) can be identified. These case studies show that (1) impacts on crops from hot and dry summers can be exacerbated by preconditioning effects of dry and bright springs. (2) Assessing compound coastal flooding in Perth (Australia) requires considering the dynamics of a non-stationary multivariate process. For instance, future mean sea-level rise will lead to the emergence of concurrent coastal and fluvial extremes, enhancing compound flooding risk. (3) In Portugal, deep-landslides are often caused by temporal clusters of moderate precipitation events. Finally, (4) crop yield failures in France and Germany are strongly correlated, threatening European food security through spatially compounding effects. These analyses allow for identifying general recommendations for studying compound events. Overall, our insights can serve as a blueprint for compound event analysis across disciplines and sectors.
... Climate model evaluation is usually univariate without considering the multivariate nature of multiple hazards, it is thus important to evaluate the biases in the dependence between the contributing variables in climate models (Vezzoli et al., 2017). However, rare studies have evaluated the climate model multivariate representation of hazard indicators (Bevacqua et al., 2019;Villalobos-Herrera et al., 2021;Zscheischler et al., 2021). ...
Compound extremes pose immense challenges and hazards to communities, and this is particularly true for compound hydrometeorological extremes associated with deadly floods, surges, droughts, heat waves. To mitigate and better adapt to compound hydrometeorological extremes, we need to better understand the state of knowledge of such extremes. Here we review the current advances in understanding compound hydrometeorological extremes: compound heat wave and drought (hot-dry), compound heat stress and extreme precipitation (hot-wet), cold-wet, cold-dry and compound flooding. We focus on the drivers of these extremes and methods used to investigate and quantify their associated risk. Overall, hot-dry compound extremes are tied to subtropical highs, blocking highs, atmospheric stagnation events, and planetary wave patterns, which are modulated by atmosphere-land feedbacks. Compared with hot-dry compound extremes, hot-wet events are less examined in the literature with most works focusing on case studies. The cold-wet compound events are commonly associated with snowfall and cold frontal systems. Although cold-dry events have been found to decrease, their underlying mechanisms require further investigation. Compound flooding encompasses storm surge and high rainfall, storm surge and sea level rise, storm surge and riverine flooding, and coastal and riverine flooding. Overall, there is a growing risk of compound flooding in the future due to changes in sea level rise, storm intensity, storm precipitation, and land-use-land-cover change. To understand processes and interactions underlying compound extremes, numerical models have been used to complement statistical modeling of the dependence between the components of compound extremes. While global climate models can simulate certain types of compound extremes, high-resolution regional models coupled with land and hydrological models are required to simulate the variability of compound extremes and to project changes in the risk of such extremes. In terms of statistical modeling of compound extremes, previous studies have used empirical approach, event coincidence analysis, multivariate distribution, the indicator approach, quantile regression and the Markov Chain method to understand the dependence, greatly advancing the state of science of compound extremes. Overall, the selection of methods depends on the type of compound extremes of interests and relevant variables.
