Search

Water and disaster risk management require accurate information about hydrometeorological extremes. Estimation of rare events by extreme value analysis is hampered by short observational records. Probabilistic seasonal forecasts allow assessing the uncertainty in the estimation of extremes. From meteorological seasonal reforecasts and therewith driven hydrological simulations, we create hundred- to thousand-year-long surrogate timeseries across Europe. We identify independent samples based on the assessment of the forecast skill, and extract precipitation and streamflow extremes to explore the impact of sample size on return period estimations. The analysis clearly demonstrates the large uncertainty in long return period estimates with typical available samples of only few decades. The uncertainty is reduced at 100-year samples, and stabilizes at very low uncertainty around 500 years. We discuss the benefits and limitations of this method, and how it can be applied to study climate change and multiple extremes.

Regional climate model simulations have routinely been applied to assess changes in precipitation extremes at daily time steps. However, shorter sub-daily extremes have not received as much attention. This is likely because of the limited availability of high temporal resolution data, both for observations and for model outputs. Here, summertime depth duration frequencies of a subset of the EURO-CORDEX 0.11∘ ensemble are evaluated with observations for several European countries for durations of 1 to 12 h. Most of the model simulations strongly underestimate 10-year depths for durations up to a few hours but perform better at longer durations. The spatial patterns over Germany are reproduced at least partly at a 12 h duration, but all models fail at shorter durations. Projected changes are assessed by relating relative depth changes to mean temperature changes. A strong relationship with temperature is found across different subregions of Europe, emission scenarios and future time periods. However, the scaling varies considerably between different combinations of global and regional climate models, with a spread in scaling of around 1–10 % K−1 at a 12 h duration and generally higher values at shorter durations.

Robust information of hydrometeorological extremes is important for effective risk management, mitigation and adaptation measures by public authorities, civil and engineers dealing for example with water management. Typically, return values of certain variables, such as extreme precipitation and river discharge, are of particular interest and are modelled statistically using Extreme Value Theory (EVT). However, the estimation of these rare events based on extreme value analysis are affected by short observational data records leading to large uncertainties.In order to overcome this limitation, we propose to use the latest seasonal meteorological prediction system of the European Centre for Medium-Range Weather Forecasts (ECMWF SEAS5) and seasonal hydrological forecasts generated with the pan-European E-HYPE model of the original period 1993-2015 and to extend the dataset to longer synthetic time series by pooling single forecast months to surrogate years. To ensure an independent dataset, the seasonal forecast skill is assessed in advance and months (and lead months) with positive skill are excluded. In this study, we simplify the method and work with samples of 6- and 4-month forecasts (instead of the full 7-month forecasts) depending on the statistical independency of the variables. It enables the record to be extended from the original 23 years to 3450 and 2300 surrogate years for the 6- and 4-month forecasts respectively.Furthermore, we investigate the robustness of estimated 50- and 100-year return values for extreme precipitation and river discharge using 1-year block maxima that are fitted to the Generalized Extreme Value distribution. Surrogate sets of pooled years are randomly constructed using the Monte-Carlo approach and different sample sizes are chosen. This analysis reveals a considerable reduction in the uncertainty of all return period estimations for both variables for selected locations across Europe using a sample size of 500 years. This highlights the potential in using the ensembles of meteorological and hydrological seasonal forecasts to obtain timeseries of sufficient length and minimize the uncertainty in the extreme value analysis.

The next generation of climate services needs not only tailoring to specific user needs but to provide, in addition, access to key information in a usable way that satisfies the needs of different users&rsquo, profiles, especially web-based services. Here, we present the outcomes from developing such a new interactive prototype. The service provides data for robust climate analysis to underpin decision-making when planning measures to compensate for climate impact. The goal is to facilitate the communication on climate information between climate modelling communities and adaptation or mitigation initiatives from vulnerable countries that are applying for funds from the Green Climate Fund (GCF). A participatory process was ensured during four workshops in four pilot countries, with an audience of national and international experts. During this process it was made clear that in all countries there is a strong need for knowledge in climate science, while in most countries there was also an increasing need of capacity in hydrological modelling and water management. The active interaction during the workshops was found necessary to facilitate the dialogue between service developers and users. Understanding the users, transparency on potentials and limitations of climate services together with capacity development in climate science and methods were required components in the development of the service.

It is common practice to apply some form of bias correction to climate models before use in impact modelling, such as hydrology. The standard method is to evaluate the correction method based on a cross validation procedure with two or more sub-periods. This allows the method to be assessed on data not previously seen in the calibration step. However, with standard split-sample setups, the data is most likely in a similar climate regime as the calibration data. In effect, the method is evaluated in the same climate regime as it is calibrated, and informs little about the performance outside the current climate regime.To address this issue, a discrete split sample test (DSST) was set up so that as diverse climate regimes as possible were sampled. The simplest climate analogue would be to perform the DSST on the coldest years and evaluate on the warmest, to mimic a changing temperature. Here, the tests are extended to more exotic indicators, such as snow pack, the joint probability of wet and cold seasons, the number of hot days in a year, the convective activity during summer; all related to a specific case study issue. Six different bias correction methods of both standard quantile mapping and other approaches to scale the reference time series are included. The methods are applied in a pseudo-reality framework to six climate model projections from Euro-CORDEX 12.5 km simulations. The analysis is focused on comparing the DSST performance with the impact on the climate change signals, and to the reliability of each method when applied to different climate regimes.

The World Meteorological Organization (WMO), the Green Climate Fund (GCF) and the Swedish Meteorological and Hydrological Institute (SMHI) are collaborating on a project providing expert services for enhancing the climate science basis of GCF-funded activities. The goal is to ensure that the causal links between climate and climate impacts, and between climate action and societal benefits, are fully grounded in the best available climate data and science. Five pilot countries are participating in this phase of the project: St Lucia, Democratic Republic of Congo, Cape Verde, Cambodia, and Paraguay, with an audience of national experts, international stakeholders, and policy and decision makers.The scientific framework which we follow here is a compendium of available data, methods and tools for analysing and documenting the past, present and potential future climate conditions which a GCF-funded project or adaptation plan might seek to address. Through the WMO-GCF-SMHI project, the methodology, scientific framework, data, methods and tools to link global to local data are complemented by hands-on support, backed by access to relevant data and tools through a structured access platform.In this presentation we elaborate on the lessons learnt from a number of workshops that were designed for the five pilot countries. The main focus of the workshops was a hands-on opportunity of national experts and international stakeholders to work with the WMO methodology in order to develop a GCF proposal for future funding. The participants in each country worked intensively during a five-day workshop on each step of the methodology: Problem definition, Identification of climate science basis, Interpretation of data analysis, selection of best adaptation/mitigation options, and assessment of adaptation/mitigation effectiveness.Assessing past and current climate and climate projections is the basis for inferring real and potential climate change and related impacts. For this, SMHI has developed a new interactive online platform/service (https://climateinformation.org/) to facilitate the communication between the GCF and developing countries and provide access to state of the art climate data to be used in impact assessment planning. The new service provides data for robust climate analysis to underpin decision-making when planning measures for climate adaptation or mitigation. Readily available climate indicators will help defining future problems, assess climatic stressors, and analyse current and future risks. This makes a climate case, which is the basis for developing interventions and propose investments. In particular the service provides:Easy access to many climate indicators, based on state-of-the-art climate science. Instant summary reports of climate change for any site on the globe. Guidance on how to link global changes to local observations.

This paper demonstrates two important aspects of regional dynamical downscaling of multidecadal atmospheric reanalysis. First, that in this way skillful regional descriptions of multidecadal climate variability may be constructed in regions with little or no local data. Second, that the concept of large-scale constraining allows global downscaling, so that global reanalyses may be completed by additions of consistent detail in all regions of the world. Global reanalyses suffer from inhomogeneities. However, their large-scale componenst are mostly homogeneous; Therefore, the concept of downscaling may be applied to homogeneously complement the large-scale state of the reanalyses with regional detail-wherever the condition of homogeneity of the description of large scales is fulfilled. Technically, this can be done by dynamical downscaling using a regional or global climate model, which's large scales are constrained by spectral nudging. This approach has been developed and tested for the region of Europe, and a skillful representation of regional weather risks-in particular marine risks-was identified. We have run this system in regions with reduced or absent local data coverage, such as Central Siberia, the Bohai and Yellow Sea, Southwestern Africa, and the South Atlantic. Also, a global simulation was computed, which adds regional features to prescribed global dynamics. Our cases demonstrate that spatially detailed reconstructions of the climate state and its change in the recent three to six decades add useful supplementary information to existing observational data for midlatitude and subtropical regions of the world.

A detailed analysis is carried out to assess the HadGEM3-A global atmospheric model skill in simulating extreme temperatures, precipitation and storm surges in Europe in the view of their attribution to human influence. The analysis is performed based on an ensemble of 15 atmospheric simulations forced with observed sea surface temperature of the 54 year period 1960–2013. These simulations, together with dual simulations without human influence in the forcing, are intended to be used in weather and climate event attribution. The analysis investigates the main processes leading to extreme events, including atmospheric circulation patterns, their links with temperature extremes, land–atmosphere and troposphere-stratosphere interactions. It also compares observed and simulated variability, trends and generalized extreme value theory parameters for temperature and precipitation. One of the most striking findings is the ability of the model to capture North-Atlantic atmospheric weather regimes as obtained from a cluster analysis of sea level pressure fields. The model also reproduces the main observed weather patterns responsible for temperature and precipitation extreme events. However, biases are found in many physical processes. Slightly excessive drying may be the cause of an overestimated summer interannual variability and too intense heat waves, especially in central/northern Europe. However, this does not seem to hinder proper simulation of summer temperature trends. Cold extremes appear well simulated, as well as the underlying blocking frequency and stratosphere-troposphere interactions. Extreme precipitation amounts are overestimated and too variable. The atmospheric conditions leading to storm surges were also examined in the Baltics region. There, simulated weather conditions appear not to be leading to strong enough storm surges, but winds were found in very good agreement with reanalyses. The performance in reproducing atmospheric weather patterns indicates that biases mainly originate from local and regional physical processes. This makes local bias adjustment meaningful for climate change attribution.

Regional climate model simulations have routinely been applied to assess changes in precipitation extremes at daily time steps. However, shorter sub-daily extremes have not received as much attention. This is likely because of the limited availability of high temporal resolution data, both for observations and for model outputs. Here, summertime depth duration frequencies of a sub-set of the EURO-CORDEX 0.11° ensemble is evaluated with observations for several European countries for durations of one to 12 h. Most of the model simulations strongly underestimate 10-year depths for durations up to a few hours, but do better on 12 h durations. All models fail in reproducing observed spatial patterns over Germany for durations shorter than 12 h, but all reproduce the pattern at least partly at 12 h duration. Large-scale driven spatial patterns, such as the extreme depths in southern France are better captured also at shorter durations, albeit severely underestimated. Projected changes are assessed by relating relative depth changes to mean temperature changes. A strong relationship with temperature is found across sub-regions of Europe, the emission scenario and future time period. However, there is an equally strong dependency on the global and regional model applied, with a spread in scaling of around 1–10 %/K at 12 h duration, and generally higher values at shorter durations.

International audience; An attribution study has been performed to investigate the degree to which the unusually cold European winter of 2009/10 was modified by anthropogenic climate change. Two different methods have been included for the attribution: one based on large HadGEM3-A ensembles and one based on a statistical surrogate method. Both methods are evaluated by comparing simulated winter temperature means, trends, standard deviations, skewness, return periods, and 5% quantiles with observations. While the surrogate method performs well, HadGEM3-A in general underestimates the trend in winter by a factor of ⅔. It has a mean cold bias dominated by the mountainous regions and also underestimates the cold 5% quantile in many regions of Europe. Both methods show that the probability of experiencing a winter as cold as 2009/10 has been reduced by approximately a factor of 2 because of anthropogenic changes. The method based on HadGEM3-A ensembles gives somewhat larger changes than the surrogate method because of differences in the definition of the unperturbed climate. The results are based on two diagnostics: the coldest day in winter and the largest continuous area with temperatures colder than twice the local standard deviation. The results are not sensitive to the choice of bias correction except in the mountainous regions. Previous results regarding the behavior of the measures of the changed probability have been extended. The counterintuitive behavior for heavy-tailed distributions is found to hold for a range of measures and for events that become more rare in a changed climate

This study analyzes the added value of a regional climate model hindcast with respect to snow water equivalent (SWE) for Siberia when compared to SWE estimates from forcing NCEP-R. In addition, we examine the discrepancies of simulated SWE to several recent reanalysis products (NCEP-R2 NCEP-CFSR, ERA-Interim). We apply the regional climate model COSMO-CLM (CCLM) to a 50 km grid spacing using NCEP-R1 as driving force to obtain a 63 yr (1948 to 2010) gridded dataset of historical SWE. Simulated regional climate data is necessary because of the absence of station data in that region. To perform large-scale assessments we use the satellite-derived daily SWE product of ESA DUE GlobSnow from 1987 to 2010. Russian station SWE data is used for cross-checking the findings. In January (mid-winter), the SWE hindcast is in good agreement with GlobSnow, whereas it overestimates SWE during the melting season. CCLM shows a clear added value in providing realistic SWE information compared to the driving reanalysis. The temporal consistency of CCLM is higher than that presented by ERA-Interim and NCEP-R2.

This paper evaluates the simulated Arctic land snow cover duration, snow water equivalent, snow cover fraction, surface albedo, and land surface temperature in the regional climate model HIRHAM5 during 2008–2010, compared with various satellite and reanalysis data and one further regional climate model (COSMO-CLM). HIRHAM5 shows a general agreement in the spatial patterns and annual course of these variables, although distinct biases for specific regions and months are obvious. The most prominent biases occur for east Siberian deciduous forest albedo, which is overestimated in the simulation for snow covered conditions in spring. This may be caused by the simplified albedo parameterization (e.g., nonconsideration of different forest types and neglecting the effect of fallen leaves and branches on snow for deciduous tree forest). The land surface temperature biases mirror the albedo biases in their spatial and temporal structures. The snow cover fraction and albedo biases can explain the simulated land surface temperature bias of ca. −3°C over the Siberian forest area in spring.

This study analyzes the added value of a regional climate model hindcast of CCLM compared to global reanalyses in providing a reconstruction of recent past snow water equivalent (SWE) for Siberia. Consistent regional climate data in time and space is necessary due to lack of station data in that region. We focus on SWE since it represents an important snow cover parameter in a region where snow has the potential to feed back to the climate of the whole Northern Hemisphere. The simulation was performed in a 50 km grid spacing for the period 1948 to 2010 using NCEP Reanalysis 1 as boundary forcing. Daily observational reference data for the period of 1987–2010 was obtained by the satellite derived SWE product of ESA DUE GlobSnow that enables a large scale assessment. The analyses includes comparisons of the distribution of snow cover extent, example time series of monthly SWE for January and April, regional characteristics of long-term monthly mean, standard deviation and temporal correlation averaged over subregions. SWE of CCLM is compared against the SWE information of NCEP-R1 itself and three more reanalyses (NCEP-R2, NCEP-CFSR, ERA-Interim). We demonstrate a significant added value of the CCLM hindcast during snow accumulation period shown for January for many subregions compared to SWE of NCEP-R1. NCEP-R1 mostly underestimates SWE during whole snow season. CCLM overestimates SWE compared to the satellite-derived product during April – a month representing the beginning of snow melt in southern regions. We illustrate that SWE of the regional hindcast is more consistent in time than ERA-Interim and NCEP-R2 and thus add realistic detail.