Richard E. Chandler
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Richard E. Chandler has a total of 188 co-authors. Here are the top co-authors by number of shared publications:
- Valerie Isham
University College London
9 shared publications - Clair Barnes
Centre for Environmental Policy, Imperial College London, London, UK
6 shared publications H. S. Wheater
Department of Civil and Environmental Engineering, Imperial College London, South Kensington Campus, London SW7 2BU, United Kingdom
6 shared publications- Chris Brierley
University College London, London, UK
5 shared publications C. Onof
Environmental and Water Resources Engineering Research Group, Imperial College London, London, UK
4 shared publications- Andrew J. Dowdy
University of Melbourne
3 shared publications
Latest Publications
- Open Access
- 9 September 2026
- Published by Wiley in Wires Climate Change
- Vol. 17 (5) , e70088
Interpreting Ensembles of Climate Projections for Impact Studies
To study the potential impacts of future changes in the earth's climate, projections of these changes are required. These are usually produced using complex computer models, and are subject to many uncertainties which are often sampled using collections (“ensembles”) of model simulations. However, the sampling in any individual ensemble is incomplete and often unrepresentative. This creates challenges for users of the projections, especially when it is important to understand the range of potential impacts that can reasonably be expected. Other challenges include the presence of systematic discrepancies between the model outputs and observations of the real climate system; and the fact that it is often infeasible to work with all members of an ensemble so that an appropriate subset must be chosen. Various approaches have been suggested for addressing each of these issues: a critical review is provided here, to allow an informed choice depending on the requirements of a given application. This article is categorized under: Climate Models and Modeling > Knowledge Generation with Models Assessing Impacts of Climate Change > Evaluating Future Impacts of Climate Change Assessing Impacts of Climate Change > Representing Uncertainty
- 9 September 2026
- Published by Wiley in Wires Climate Change
- Vol. 17 (5) , e70088
Open AccessInterpreting Ensembles of Climate Projections for Impact Studies
To study the potential impacts of future changes in the earth's climate, projections of these changes are required. These are usually produced using complex computer models, and are subject to many uncertainties which are often sampled using collections (“ensembles”) of model simulations. However, the sampling in any individual ensemble is incomplete and often unrepresentative. This creates challenges for users of the projections, especially when it is important to understand the range of potential impacts that can reasonably be expected. Other challenges include the presence of systematic discrepancies between the model outputs and observations of the real climate system; and the fact that it is often infeasible to work with all members of an ensemble so that an appropriate subset must be chosen. Various approaches have been suggested for addressing each of these issues: a critical review is provided here, to allow an informed choice depending on the requirements of a given application. This article is categorized under: Climate Models and Modeling > Knowledge Generation with Models Assessing Impacts of Climate Change > Evaluating Future Impacts of Climate Change Assessing Impacts of Climate Change > Representing Uncertainty
- Open Access
- 5 May 2026
- Published by Copernicus GmbH in Advances in Statistical Climatology, Meteorology and Oceanography
- Vol. 12 (1) , 149-172
Improving multisite precipitation generators based on generalised linear models
Precipitation generators are statistical models used to produce synthetic sequences of (multisite) precipitation for hydrological applications such as flood risk assessment and water resource management. Among these, approaches based on generalized linear models (GLMs) are widely used and often perform competitively with state-of-the-art alternatives, but they face limitations in representing seasonal variation in extremes and in flexibly capturing covariate effects on the precipitation distribution. In this paper, we extend the GLM framework in two directions. First, we introduce generalised additive models for location, scale and shape (GAMLSS) for precipitation generation. These models allow the use of spline-based model terms to flexibly capture covariate effects and allow covariates to influence multiple distributional parameters, thereby increasing flexibility in representing variation in both the mean and variance of the precipitation distribution. Second, we adapt a transformed Gaussian fields approach to jointly account for spatial dependence in both precipitation occurrence and intensity, thus allowing for potential cross-dependence between the two. A further contribution is to investigate the sensitivity of model performance to data resolution, highlighting that rounding in data pre-processing can substantially affect the reproduction of extremes. Using a well-studied daily precipitation dataset, we demonstrate that these extensions improve the realism of simulated sequences, particularly with respect to extremes, and capture spatial dependence well in both occurrence and intensity.
- 5 May 2026
- Published by Copernicus GmbH in Advances in Statistical Climatology, Meteorology and Oceanography
- Vol. 12 (1) , 149-172
Open AccessImproving multisite precipitation generators based on generalised linear models
Precipitation generators are statistical models used to produce synthetic sequences of (multisite) precipitation for hydrological applications such as flood risk assessment and water resource management. Among these, approaches based on generalized linear models (GLMs) are widely used and often perform competitively with state-of-the-art alternatives, but they face limitations in representing seasonal variation in extremes and in flexibly capturing covariate effects on the precipitation distribution. In this paper, we extend the GLM framework in two directions. First, we introduce generalised additive models for location, scale and shape (GAMLSS) for precipitation generation. These models allow the use of spline-based model terms to flexibly capture covariate effects and allow covariates to influence multiple distributional parameters, thereby increasing flexibility in representing variation in both the mean and variance of the precipitation distribution. Second, we adapt a transformed Gaussian fields approach to jointly account for spatial dependence in both precipitation occurrence and intensity, thus allowing for potential cross-dependence between the two. A further contribution is to investigate the sensitivity of model performance to data resolution, highlighting that rounding in data pre-processing can substantially affect the reproduction of extremes. Using a well-studied daily precipitation dataset, we demonstrate that these extensions improve the realism of simulated sequences, particularly with respect to extremes, and capture spatial dependence well in both occurrence and intensity.
- Open Access
- 19 February 2026
- Published by Elsevier in Environmental Modelling & Software
- Vol. 199
Bayesian network-informed conditional random forests for probabilistic multisite downscaling of precipitation occurrence
- 19 February 2026
- Published by Elsevier in Environmental Modelling & Software
- Vol. 199
Open AccessBayesian network-informed conditional random forests for probabilistic multisite downscaling of precipitation occurrence
- Open Access
- 28 October 2025
- Published by Wiley
Improving multisite precipitation generators based on generalised linear models
Precipitation generators are statistical models used to produce synthetic sequences of (multisite) precipitation for hydrological applications such as flood risk assessment and water resource management. Among these, approaches based on generalized linear models (GLMs) are widely used and often perform competitively with state-of-the-art alternatives, but they face limitations in representing seasonal variation in extremes and in flexibly capturing covariate effects on the precipitation distribution. In this paper, we extend the GLM framework in two directions. First, we introduce generalised additive models for location, scale and shape (GAMLSS) for precipitation generation. These models allow the use of spline-based model terms to flexibly capture covariate effects and allow covariates to influence multiple distributional parameters, thereby increasing flexibility in representing variation in both the mean and variance of the precipitation distribution. Second, we adapt a transformed Gaussian fields approach to jointly account for spatial dependence in both precipitation occurrence and intensity, thus allowing for potential cross-dependence between the two. A further contribution is to investigate the sensitivity of model performance to data resolution, highlighting that rounding in data pre-processing can substantially affect the reproduction of extremes. Using a well-studied daily precipitation dataset, we demonstrate that these extensions improve the realism of simulated sequences, particularly with respect to extremes, and capture spatial dependence well in both occurrence and intensity.
- 28 October 2025
- Published by Wiley
Open AccessImproving multisite precipitation generators based on generalised linear models
Precipitation generators are statistical models used to produce synthetic sequences of (multisite) precipitation for hydrological applications such as flood risk assessment and water resource management. Among these, approaches based on generalized linear models (GLMs) are widely used and often perform competitively with state-of-the-art alternatives, but they face limitations in representing seasonal variation in extremes and in flexibly capturing covariate effects on the precipitation distribution. In this paper, we extend the GLM framework in two directions. First, we introduce generalised additive models for location, scale and shape (GAMLSS) for precipitation generation. These models allow the use of spline-based model terms to flexibly capture covariate effects and allow covariates to influence multiple distributional parameters, thereby increasing flexibility in representing variation in both the mean and variance of the precipitation distribution. Second, we adapt a transformed Gaussian fields approach to jointly account for spatial dependence in both precipitation occurrence and intensity, thus allowing for potential cross-dependence between the two. A further contribution is to investigate the sensitivity of model performance to data resolution, highlighting that rounding in data pre-processing can substantially affect the reproduction of extremes. Using a well-studied daily precipitation dataset, we demonstrate that these extensions improve the realism of simulated sequences, particularly with respect to extremes, and capture spatial dependence well in both occurrence and intensity.
- Open Access
- 8 October 2025
- Published by Copernicus GmbH
- Vol. 2025, 1-30
Potentially interesting, but doesn't read like a HESS paper and doesn't make a strong enough case at present
Weather generators (WGs) are important tools for downscaling General Circulation Model (GCM) output for climate impact modelling. This study introduces a precipitation WG conditioned on a recent storm types dataset and outlines a methodology for evaluating WGs using proper scoring rules. The storm types are a set of discrete weather types that use atmospheric variables to categorise ERA5 grid cells into fronts, cyclones and thunderstorms or combinations of these. The WG is a Bayesian Generalised Linear Model (GLM) with vertical velocity and humidity as covariates, conditioned on the storm types, and trained on 6 hourly precipitation accumulations. This approach contrasts with previous WGs based on weather types, which use clustering methods such as k-means to generate weather types. A Bayesian model framework is used, instead of typical maximum likelihood approaches, and an informed prior choice is made on observable quantities using the prior predictive distribution. The WG is assessed using proper scoring rules and Diebold-Mariano (DM) significance tests. The use of a DM test to assess the statistical significance of average proper score differences is a key addition to typical WG evaluation approaches, as it helps model developers avoid changes that improve an average score by chance. Calibration is assessed using the probability integral transform histogram and by comparing draws from the posterior predictive distribution to observations. Compared to the same WG not conditioned on storm types the inclusion of storm types improved the average Continuous Ranked Probably Score (CRPS) by a statistically significant amount across the stations considered. When storm types are used as an alternative to continuous atmospheric variables, they provide 33 % of the improvement in average CRPS that the atmospheric variables do, averaged over the stations. To quantify the WG's ability to represent extremes the threshold weighted CRPS (twCRPS) is explored. For three different thresholds the twCRPS corroborates the results for the CRPS. The use of proper scoring rules in conjunction with a DM test is highlighted as a powerful tool for assessing WG skill.
- 8 October 2025
- Published by Copernicus GmbH
- Vol. 2025, 1-30
Open AccessPotentially interesting, but doesn't read like a HESS paper and doesn't make a strong enough case at present
Weather generators (WGs) are important tools for downscaling General Circulation Model (GCM) output for climate impact modelling. This study introduces a precipitation WG conditioned on a recent storm types dataset and outlines a methodology for evaluating WGs using proper scoring rules. The storm types are a set of discrete weather types that use atmospheric variables to categorise ERA5 grid cells into fronts, cyclones and thunderstorms or combinations of these. The WG is a Bayesian Generalised Linear Model (GLM) with vertical velocity and humidity as covariates, conditioned on the storm types, and trained on 6 hourly precipitation accumulations. This approach contrasts with previous WGs based on weather types, which use clustering methods such as k-means to generate weather types. A Bayesian model framework is used, instead of typical maximum likelihood approaches, and an informed prior choice is made on observable quantities using the prior predictive distribution. The WG is assessed using proper scoring rules and Diebold-Mariano (DM) significance tests. The use of a DM test to assess the statistical significance of average proper score differences is a key addition to typical WG evaluation approaches, as it helps model developers avoid changes that improve an average score by chance. Calibration is assessed using the probability integral transform histogram and by comparing draws from the posterior predictive distribution to observations. Compared to the same WG not conditioned on storm types the inclusion of storm types improved the average Continuous Ranked Probably Score (CRPS) by a statistically significant amount across the stations considered. When storm types are used as an alternative to continuous atmospheric variables, they provide 33 % of the improvement in average CRPS that the atmospheric variables do, averaged over the stations. To quantify the WG's ability to represent extremes the threshold weighted CRPS (twCRPS) is explored. For three different thresholds the twCRPS corroborates the results for the CRPS. The use of proper scoring rules in conjunction with a DM test is highlighted as a powerful tool for assessing WG skill.
- Open Access
- 1 February 2024
- Published by American Meteorological Society in Journal of Climate
- Vol. 37 (3) , 1053-1064
Characterizing Spatial Structure in Climate Model Ensembles
This paper presents a methodology that is designed for rapid exploratory analysis of the outputs from ensembles of climate models, especially when these outputs consist of maps. The approach formalizes and extends the technique of “intermodel empirical orthogonal function” analysis, combining multivariate analysis of variance techniques with singular value decompositions (SVDs) of structured components of the ensemble data matrix. The SVDs yield spatial patterns associated with these components, which we call ensemble principal patterns (EPPs). A unique hierarchical partitioning of variation is obtained for balanced ensembles in which all combinations of factors, such as GCM and RCM pairs in a regional ensemble, appear with equal frequency: suggestions are also proposed to handle unbalanced ensembles without imputing missing values or discarding runs. Applications include the selection of ensemble members to propagate uncertainty into subsequent analyses, and the diagnosis of modes of variation associated with specific model variants or parameter perturbations. The approach is illustrated using outputs from the EuroCORDEX regional ensemble over the United Kingdom.
- 1 February 2024
- Published by American Meteorological Society in Journal of Climate
- Vol. 37 (3) , 1053-1064
Open AccessCharacterizing Spatial Structure in Climate Model Ensembles
This paper presents a methodology that is designed for rapid exploratory analysis of the outputs from ensembles of climate models, especially when these outputs consist of maps. The approach formalizes and extends the technique of “intermodel empirical orthogonal function” analysis, combining multivariate analysis of variance techniques with singular value decompositions (SVDs) of structured components of the ensemble data matrix. The SVDs yield spatial patterns associated with these components, which we call ensemble principal patterns (EPPs). A unique hierarchical partitioning of variation is obtained for balanced ensembles in which all combinations of factors, such as GCM and RCM pairs in a regional ensemble, appear with equal frequency: suggestions are also proposed to handle unbalanced ensembles without imputing missing values or discarding runs. Applications include the selection of ensemble members to propagate uncertainty into subsequent analyses, and the diagnosis of modes of variation associated with specific model variants or parameter perturbations. The approach is illustrated using outputs from the EuroCORDEX regional ensemble over the United Kingdom.
- Open Access
- 10 January 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Atmospheres
- Vol. 129 (2)
A Comparison of Regional Climate Projections With a Range of Climate Sensitivities
To investigate the extent to which differences in regional model projections can be explained by differences in the warming rates of their driving models, we compare projections of temperature and precipitation over the UK from two regional climate ensembles—the EuroCORDEX multi‐model ensemble and UKCP18 perturbed parameter ensemble—along with projections produced by the “parent” GCMs from which boundary conditions were taken. We evaluate the ensembles in terms of their representation of recent climate, then compare the changes simulated between 1981–2010 and 2050–2079. While both ensembles exhibit seasonal biases with similar magnitudes and spatial patterns during the evaluation period, the UKCP18 ensemble exhibits a somewhat stronger change signal in future simulations, due to a combination of higher climate sensitivity of the driving models, variations in the forcings applied, and—in the regional simulations—the inclusion of time‐varying aerosols. In order to reconcile the two sets of projections, we compare two periods corresponding to fixed global warming levels in the driving models, to constrain the variability within and between the ensembles which can be ascribed to differing rates of global warming: the discrepancy between the ensembles is greatly reduced, although some differences in the local response remain, with the UKCP18 runs slightly warmer and drier than the EuroCORDEX runs, particularly in summer. We also highlight potential pitfalls of comparing warming levels with a reference time period, due to uncertainty about the warming that has already occurred in the driving models prior to the reference period.
- 10 January 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Atmospheres
- Vol. 129 (2)
Open AccessA Comparison of Regional Climate Projections With a Range of Climate Sensitivities
To investigate the extent to which differences in regional model projections can be explained by differences in the warming rates of their driving models, we compare projections of temperature and precipitation over the UK from two regional climate ensembles—the EuroCORDEX multi‐model ensemble and UKCP18 perturbed parameter ensemble—along with projections produced by the “parent” GCMs from which boundary conditions were taken. We evaluate the ensembles in terms of their representation of recent climate, then compare the changes simulated between 1981–2010 and 2050–2079. While both ensembles exhibit seasonal biases with similar magnitudes and spatial patterns during the evaluation period, the UKCP18 ensemble exhibits a somewhat stronger change signal in future simulations, due to a combination of higher climate sensitivity of the driving models, variations in the forcings applied, and—in the regional simulations—the inclusion of time‐varying aerosols. In order to reconcile the two sets of projections, we compare two periods corresponding to fixed global warming levels in the driving models, to constrain the variability within and between the ensembles which can be ascribed to differing rates of global warming: the discrepancy between the ensembles is greatly reduced, although some differences in the local response remain, with the UKCP18 runs slightly warmer and drier than the EuroCORDEX runs, particularly in summer. We also highlight potential pitfalls of comparing warming levels with a reference time period, due to uncertainty about the warming that has already occurred in the driving models prior to the reference period.
- Open Access
- 16 April 2023
- Published by Wiley
Bayesian Network-Informed Conditional Random Forests for Probabilistic Multisite Downscaling of Precipitation Occurrence
This work presents Bayesian network-informed conditional random forests (BNICRF), a novel multiresponse classification technique able to downscale the joint probability distribution of precipitation occurrence at a set of geographical locations of interest from large-scale predictors. BNICRFs combine a Bayesian network, which provides a description of the multisite/spatial dependence structure of the target locations; with an ensemble of conditional random forests, which extract the relevant information from the large-scale predictors to produce accurate probabilistic predictions. This work extends two previous articles that explored Bayesian networks (Legasa & Gutiérrez, 2020) and random forests (Legasa et al., 2022) in the context of statistical climate downscaling. Building on the experimental and validation framework proposed in the experiment 1c of the COST action VALUE (the largest, most exhaustive intercomparison study of statistical downscaling methods to date) and under the perfect prognosis approach, we thoroughly assess the performance of the proposed methodology focusing on its capability to reproduce several multisite and single-site statistics of the observed series. BNICRFs accurately capture the relevant spatial relationships while keeping the same single-site predictive performance than both single-site random forests and a Generalized Linear Model that exhibited a good performance in VALUE. Moreover, we compare BNICRFs against a robust multisite methodology proposed in Chandler (2020), obtaining better predictive capability while keeping similar spatial performance. Since the assessed models still underestimate autocorrelation, we also propose a straightforward extension of BNICRFs that incorporates the temporal structure at each location of interest able to produce both temporarily and spatially realistic precipitation occurrence fields
- 16 April 2023
- Published by Wiley
Open AccessBayesian Network-Informed Conditional Random Forests for Probabilistic Multisite Downscaling of Precipitation Occurrence
This work presents Bayesian network-informed conditional random forests (BNICRF), a novel multiresponse classification technique able to downscale the joint probability distribution of precipitation occurrence at a set of geographical locations of interest from large-scale predictors. BNICRFs combine a Bayesian network, which provides a description of the multisite/spatial dependence structure of the target locations; with an ensemble of conditional random forests, which extract the relevant information from the large-scale predictors to produce accurate probabilistic predictions. This work extends two previous articles that explored Bayesian networks (Legasa & Gutiérrez, 2020) and random forests (Legasa et al., 2022) in the context of statistical climate downscaling. Building on the experimental and validation framework proposed in the experiment 1c of the COST action VALUE (the largest, most exhaustive intercomparison study of statistical downscaling methods to date) and under the perfect prognosis approach, we thoroughly assess the performance of the proposed methodology focusing on its capability to reproduce several multisite and single-site statistics of the observed series. BNICRFs accurately capture the relevant spatial relationships while keeping the same single-site predictive performance than both single-site random forests and a Generalized Linear Model that exhibited a good performance in VALUE. Moreover, we compare BNICRFs against a robust multisite methodology proposed in Chandler (2020), obtaining better predictive capability while keeping similar spatial performance. Since the assessed models still underestimate autocorrelation, we also propose a straightforward extension of BNICRFs that incorporates the temporal structure at each location of interest able to produce both temporarily and spatially realistic precipitation occurrence fields
- Open Access
A comparison of regional climate projections with a range of climate sensitivities
We compare projections of temperature and precipitation over the UK land region from two regional climate ensembles - the EuroCORDEX multi-model ensemble and UKCP18 perturbed physics ensemble - along with projections produced by the 'parent' GCMs from which boundary conditions were taken for the regional runs. The ensembles are first evaluated in terms of their representation of recent climate, before comparing the changes simulated between a reference period from 1980-2010 and the future period from 2050-2080. While the two ensembles exhibit fairly similar biases during the evaluation period, the UKCP18 ensemble exhibits a somewhat stronger change signal in the future simulations, due to the higher climate sensitivity of the driving models. In order to reconcile the two sets of projections, we compare two periods corresponding to fixed global warming levels in the driving models, with the intention of removing the variability within and between the ensembles which can be ascribed to differing rates of global warming. We also highlight the potential pitfalls of comparing warming levels with a reference time period, due to uncertainty about the warming that has already occurred in each of the driving models prior to the reference period. We expect that these findings will become more widely relevant as ensembles increase in size, and particularly in ensembles including models with a higher range of climate sensitivities, as is seen already in CMIP6.
Open AccessA comparison of regional climate projections with a range of climate sensitivities
We compare projections of temperature and precipitation over the UK land region from two regional climate ensembles - the EuroCORDEX multi-model ensemble and UKCP18 perturbed physics ensemble - along with projections produced by the 'parent' GCMs from which boundary conditions were taken for the regional runs. The ensembles are first evaluated in terms of their representation of recent climate, before comparing the changes simulated between a reference period from 1980-2010 and the future period from 2050-2080. While the two ensembles exhibit fairly similar biases during the evaluation period, the UKCP18 ensemble exhibits a somewhat stronger change signal in the future simulations, due to the higher climate sensitivity of the driving models. In order to reconcile the two sets of projections, we compare two periods corresponding to fixed global warming levels in the driving models, with the intention of removing the variability within and between the ensembles which can be ascribed to differing rates of global warming. We also highlight the potential pitfalls of comparing warming levels with a reference time period, due to uncertainty about the warming that has already occurred in each of the driving models prior to the reference period. We expect that these findings will become more widely relevant as ensembles increase in size, and particularly in ensembles including models with a higher range of climate sensitivities, as is seen already in CMIP6.
- Open Access
- 28 March 2022
- Published by Copernicus GmbH
Identifying patterns of spatial variability within the EuroCORDEX ensemble
Ensembles of regional climate projections provide information about the range of possible scenarios of future climate change at the local scale, with more detail and better representation of fine-scale processes than can be provided by lower-resolution global circulation models (GCMs). The CORDEX ensembles are multi-model ensembles, with each member obtained by using a GCM to drive a higher-resolution regional climate model (RCM). Due to resource limitations however, users of regional climate information typically do not want to use an entire ensemble and must select a sample of its members for their purposes. To preserve as much information as possible, such a sample should be chosen to be representative of the variation within the ensemble.Analysis of variance (ANOVA) has often been used to characterise ensemble variation by apportioning the total variation to differences between the GCMs or between the RCMs (Yip et al., 2011; Déqué et al., 2012), and to produce maps of the geographical regions where variance between the runs is ascribed to one or other model component (Christensen and Kjellström, 2020). However, traditional ANOVA methods require a balanced ensemble in which all possible GCM-RCM pairs are available. The analysis of unbalanced ensembles therefore typically proceeds either by discarding surplus runs or imputing missing ones, or by using computationally intensive Bayesian methods to account for the lack of balance.We here propose two enhancements to the existing techniques for analysis of ensemble variation. The first is a modification of the standard ANOVA approach, based on the underlying statistical model, that can be applied directly to unbalanced ensembles: the modification is computationally cheap and hence suitable for routine application, and provides ranges of variation that are potentially attributable to the different sources.The second enhancement adds further detail to the partitioning of variation, using an eigenanalysis that characterises the principal spatial modes of variation within an ensemble. As well as identifying the dominant spatial patterns of variation associated with the GCMs and RCMs, the analysis characterises the contribution from each model, for example by identifying models with different treatments of orography, rain shadows, or urban heat island effects. As well as informing the selection of subsets of ensemble members, this enhancement offers the possibility of emulating missing ensemble members where the GCM-RCM matrix is only partially filled. The method is applied to the EuroCORDEX ensemble with a focus on the UK. ReferencesChristensen, O. and Kjellström, E. (2020). Partitioning uncertainty components of mean climate and climate change in a large ensemble of European regional climate model projections. Climate Dynamics, 54:4293–4308.Déqué, M., Somot, S., Sanchez-Gomez, E. et al. (2012). The spread amongst ENSEMBLES regional scenarios: regional climate models, driving general circulation models and interannual variability. Climate Dynamics, 38:951–964 (2012).Yip, S., Ferro, C. A. T., Stephenson, D. B., and Hawkins, E. (2011). A simple, coherent framework for partitioning uncertainty in climate predictions. Journal of Climate, 24(17):4634–4643.
- 28 March 2022
- Published by Copernicus GmbH
Open AccessIdentifying patterns of spatial variability within the EuroCORDEX ensemble
Ensembles of regional climate projections provide information about the range of possible scenarios of future climate change at the local scale, with more detail and better representation of fine-scale processes than can be provided by lower-resolution global circulation models (GCMs). The CORDEX ensembles are multi-model ensembles, with each member obtained by using a GCM to drive a higher-resolution regional climate model (RCM). Due to resource limitations however, users of regional climate information typically do not want to use an entire ensemble and must select a sample of its members for their purposes. To preserve as much information as possible, such a sample should be chosen to be representative of the variation within the ensemble.Analysis of variance (ANOVA) has often been used to characterise ensemble variation by apportioning the total variation to differences between the GCMs or between the RCMs (Yip et al., 2011; Déqué et al., 2012), and to produce maps of the geographical regions where variance between the runs is ascribed to one or other model component (Christensen and Kjellström, 2020). However, traditional ANOVA methods require a balanced ensemble in which all possible GCM-RCM pairs are available. The analysis of unbalanced ensembles therefore typically proceeds either by discarding surplus runs or imputing missing ones, or by using computationally intensive Bayesian methods to account for the lack of balance.We here propose two enhancements to the existing techniques for analysis of ensemble variation. The first is a modification of the standard ANOVA approach, based on the underlying statistical model, that can be applied directly to unbalanced ensembles: the modification is computationally cheap and hence suitable for routine application, and provides ranges of variation that are potentially attributable to the different sources.The second enhancement adds further detail to the partitioning of variation, using an eigenanalysis that characterises the principal spatial modes of variation within an ensemble. As well as identifying the dominant spatial patterns of variation associated with the GCMs and RCMs, the analysis characterises the contribution from each model, for example by identifying models with different treatments of orography, rain shadows, or urban heat island effects. As well as informing the selection of subsets of ensemble members, this enhancement offers the possibility of emulating missing ensemble members where the GCM-RCM matrix is only partially filled. The method is applied to the EuroCORDEX ensemble with a focus on the UK. ReferencesChristensen, O. and Kjellström, E. (2020). Partitioning uncertainty components of mean climate and climate change in a large ensemble of European regional climate model projections. Climate Dynamics, 54:4293–4308.Déqué, M., Somot, S., Sanchez-Gomez, E. et al. (2012). The spread amongst ENSEMBLES regional scenarios: regional climate models, driving general circulation models and interannual variability. Climate Dynamics, 38:951–964 (2012).Yip, S., Ferro, C. A. T., Stephenson, D. B., and Hawkins, E. (2011). A simple, coherent framework for partitioning uncertainty in climate predictions. Journal of Climate, 24(17):4634–4643.