An evaluation of current capabilities of modelling low-frequency climate variability

Abstract. A crucial performance test of Earth system models is their ability to simulate the climate mean state and variability. Here we concentrate on representation of inter-annual to multi-decadal variability in 12 CMIP5 climate model simulations. Reference climate is provided by two 20th century reanalysis data sets of monthly mean near-surface air temperature. The spectral decomposition is based on Randomised Multi-Channel Singular Spectrum Analysis (RMSSA). The main results are as follows. First, the total spectra for the two reanalysis data sets are remarkably similar in all time scales, except that spectral power of decadal variability (10–30 yr) differ in these data by about 30 %. Second, none of the 12 coupled climate models closely reproduce all aspects of the reference spectra, although some models represent many aspects well. For instance, the IPSL-CM5B-LR model is close to reanalyses but has too little multi-decadal variability, and the HadGEM2-ES model is close to reanalyses except the notable over-activity at periods at and around 10 yr.

Download Full-text

Multi-annual modes in the 20th century temperature variability in reanalyses and CMIP5 models

Geoscientific Model Development ◽

10.5194/gmd-9-4097-2016 ◽

2016 ◽

Vol 9 (11) ◽

pp. 4097-4109

Author(s):

Heikki Järvinen ◽

Teija Seitola ◽

Johan Silén ◽

Jouni Räisänen

Keyword(s):

Climate Variability ◽

20Th Century ◽

Climate Models ◽

Spectral Power ◽

Singular Spectrum Analysis ◽

Reanalysis Data ◽

Cmip5 Models ◽

Data Sets ◽

Near Surface ◽

Annual Variability

Abstract. A performance expectation is that Earth system models simulate well the climate mean state and the climate variability. To test this expectation, we decompose two 20th century reanalysis data sets and 12 CMIP5 model simulations for the years 1901–2005 of the monthly mean near-surface air temperature using randomised multi-channel singular spectrum analysis (RMSSA). Due to the relatively short time span, we concentrate on the representation of multi-annual variability which the RMSSA method effectively captures as separate and mutually orthogonal spatio-temporal components. This decomposition is a unique way to separate statistically significant quasi-periodic oscillations from one another in high-dimensional data sets.The main results are as follows. First, the total spectra for the two reanalysis data sets are remarkably similar in all timescales, except that the spectral power in ERA-20C is systematically slightly higher than in 20CR. Apart from the slow components related to multi-decadal periodicities, ENSO oscillations with approximately 3.5- and 5-year periods are the most prominent forms of variability in both reanalyses. In 20CR, these are relatively slightly more pronounced than in ERA-20C. Since about the 1970s, the amplitudes of the 3.5- and 5-year oscillations have increased, presumably due to some combination of forced climate change, intrinsic low-frequency climate variability, or change in global observing network. Second, none of the 12 coupled climate models closely reproduce all aspects of the reanalysis spectra, although some models represent many aspects well. For instance, the GFDL-ESM2M model has two nicely separated ENSO periods although they are relatively too prominent as compared with the reanalyses. There is an extensive Supplement and YouTube videos to illustrate the multi-annual variability of the data sets.

Download Full-text

Evaluating temperature extremes in CMIP6 simulations using statistically proper evaluation methods

10.5194/egusphere-egu21-9722 ◽

2021 ◽

Author(s):

Thordis Thorarinsdottir ◽

Jana Sillmann ◽

Marion Haugen ◽

Nadine Gissibl ◽

Marit Sandstad

Keyword(s):

Air Temperature ◽

Model Evaluation ◽

Climate Models ◽

Climate Model ◽

Mean Squared Error ◽

Surface Air Temperature ◽

Cmip5 Models ◽

Model Output ◽

Intergovernmental Panel ◽

Near Surface

Reliable projections of extremes in near-surface air temperature (SAT) by climate models become more and more important&#160;as global warming is leading to significant increases in the hottest days and decreases in coldest nights around the world with considerable impacts on various sectors, such as agriculture, health and tourism.Climate model evaluation has traditionally been performed by comparing summary statistics that are derived from simulated model output and corresponding observed quantities using, for instance, the root mean squared error (RMSE) or mean bias as also used in the model evaluation chapter of the fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR5). Both RMSE and mean bias compare averages over time and/or space, ignoring the variability, or the uncertainty, in the underlying values. Particularly when interested in the evaluation of climate extremes, climate models should be evaluated by comparing the probability distribution of model output to the corresponding distribution of observed data.To address this shortcoming, we use the integrated quadratic distance (IQD) to compare distributions of simulated indices to the corresponding distributions from a data product. The IQD is the proper divergence associated with the proper continuous ranked probability score (CRPS) as it fulfills essential decision-theoretic properties for ranking competing models and testing equality in performance, while also assessing the full distribution.The IQD is applied to evaluate CMIP5 and CMIP6 simulations of monthly maximum (TXx) and minimum near-surface air temperature (TNn) over the data-dense regions Europe and North America against both observational and reanalysis datasets. There is not a notable difference between the model generations CMIP5 and CMIP6 when the model simulations are compared against the observational dataset HadEX2. However, the CMIP6 models show a better agreement with the reanalysis ERA5 than CMIP5 models, with a few exceptions. Overall, the climate models show higher skill when compared against ERA5 than when compared against HadEX2. While the model rankings vary with region, season and index, the model evaluation is robust against changes in the grid resolution considered in the analysis.

Download Full-text

Snowfall in the Himalayas: an uncertain future from a little-known past

The Cryosphere ◽

10.5194/tc-9-1147-2015 ◽

2015 ◽

Vol 9 (3) ◽

pp. 1147-1167 ◽

Cited By ~ 21

Author(s):

E. Viste ◽

A. Sorteberg

Keyword(s):

Climate Models ◽

Climate Model ◽

Reanalysis Data ◽

Indus Basin ◽

Data Sets ◽

Anthropogenic Forcing ◽

Temperature And Precipitation ◽

The Mean ◽

Uncertain Future ◽

Relative Change

Abstract. Snow and ice provide large amounts of meltwater to the Indus, Ganges and Brahmaputra rivers. This study combines present-day observations and reanalysis data with climate model projections to estimate the amount of snow falling over the basins today and in the last decades of the 21st century. Estimates of present-day snowfall based on a combination of temperature and precipitation from reanalysis data and observations vary by factors of 2–4. The spread is large, not just between the reanalysis and the observations but also between the different observational data sets. With the strongest anthropogenic forcing scenario (RCP8.5), the climate models project reductions in annual snowfall by 30–50% in the Indus Basin, 50–60% in the Ganges Basin and 50–70% in the Brahmaputra Basin by 2071–2100. The reduction is due to increasing temperatures, as the mean of the models show constant or increasing precipitation throughout the year in most of the region. With the strongest anthropogenic forcing scenario, the mean elevation where rain changes to snow – the rain/snow line – creeps upward by 400–900 m, in most of the region by 700–900 meters. The largest relative change in snowfall is seen in the upper westernmost sub-basins of the Brahmaputra. With the strongest forcing scenario, most of this region will have temperatures above freezing, especially in the summer. The projected reduction in annual snowfall is 65–75%. In the upper Indus, the effect of a warmer climate on snowfall is less extreme, as most of the terrain is high enough to have temperatures sufficiently far below freezing today. A 20–40% reduction in annual snowfall is projected.

Download Full-text

Snowfall in the Himalayas: an uncertain future from a little-known past

The Cryosphere Discussions ◽

10.5194/tcd-9-441-2015 ◽

2015 ◽

Vol 9 (1) ◽

pp. 441-493 ◽

Cited By ~ 2

Author(s):

E. Viste ◽

A. Sorteberg

Keyword(s):

Climate Models ◽

Climate Model ◽

Reanalysis Data ◽

Indus Basin ◽

Data Sets ◽

Anthropogenic Forcing ◽

Temperature And Precipitation ◽

The Mean ◽

Uncertain Future ◽

Relative Change

Abstract. Snow and ice provide large amounts of meltwater to the Indus, Ganges and Brahmaputra rivers. This study combines present-day observations and reanalysis data with climate model projections to estimate the amount of snow falling over the basins today and in the last decades of the 21st century. Estimates of present-day snowfall based on a combination of temperature and precipitation from reanalysis data and observations, vary by factors of 2–4. The spread is large, not just between the reanalysis and the observations, but also between the different observational data sets. With the strongest anthropogenic forcing scenario (RCP 8.5), the climate models project reductions in annual snowfall by 30–50% in the Indus Basin, 50–60% in the Ganges Basin and 50–70% in the Brahmaputra Basin, by 2071–2100. The reduction is due to increasing temperatures, as the mean of the models show constant or increasing precipitation throughout the year in most of the region. With the strongest anthropogenic forcing scenario, the mean elevation where rain changes to snow – the rain/snow line – creeps upward by 400–900 m, in most of the region by 700–900 m. The largest relative change in snowfall is seen in the upper, westernmost sub-basins of the Brahmaputra. With the strongest forcing scenario, most of this region will have temperatures above freezing, especially in the summer. The projected reduction in annual snowfall is 65–75%. In the upper Indus, the effect of a warmer climate on snowfall is less extreme, as most of the terrain is high enough to have temperatures sufficiently far below freezing today. A 20–40% reduction in annual snowfall is projected.

Download Full-text

Future Extreme Climates Projection over Huang-Huai-Hai Region of China

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.955-959.3887 ◽

2014 ◽

Vol 955-959 ◽

pp. 3887-3892 ◽

Cited By ~ 1

Author(s):

Huang He Gu ◽

Zhong Bo Yu ◽

Ji Gan Wang

Keyword(s):

Climate Changes ◽

Climate Model ◽

Regional Climate ◽

Surface Air Temperature ◽

Daily Rainfall ◽

Heavy Rain ◽

Near Surface ◽

Temperature And Precipitation ◽

Huai River ◽

The Future

This study projects the future extreme climate changes over Huang-Huai-Hai (3H) region in China using a regional climate model (RegCM4). The RegCM4 performs well in “current” climate (1970-1999) simulations by compared with the available surface station data, focusing on near-surface air temperature and precipitation. Future climate changes are evaluated based on experiments driven by European-Hamburg general climate model (ECHAM5) in A1B future scenario (2070-2099). The results show that the annual temperature increase about 3.4 °C-4.2 °C and the annual precipitation increase about 5-15% in most of 3H region at the end of 21st century. The model predicts a generally less frost days, longer growing season, more hot days, no obvious change in heat wave duration index, larger maximum five-day rainfall, more heavy rain days, and larger daily rainfall intensity. The results indicate a higher risk of floods in the future warmer climate. In addition, the consecutive dry days in Huai River Basin will increase, indicating more serve drought and floods conditions in this region.

Download Full-text

Regional Climate Model Inter-Comparison for Antarctica within a Data Science Framework

10.5194/egusphere-egu21-4892 ◽

2021 ◽

Author(s):

Jeremy Carter ◽

Amber Leeson ◽

Andrew Orr ◽

Christoph Kittel ◽

Melchior van Wessem

Keyword(s):

Regional Climate Model ◽

Climate Models ◽

Climate Model ◽

Regional Climate ◽

Regional Climate Models ◽

Future Climate Change ◽

Near Surface ◽

Virtual Lab ◽

Model Variability ◽

The Antarctic

Understanding the surface climatology of the Antarctic ice sheet is essential if we are to adequately predict its response to future climate change. This includes both primary impacts such as increased ice melting and secondary impacts such as ice shelf collapse events. Given its size, and inhospitable environment, weather stations on Antarctica are sparse. Thus, we rely on regional climate models to 1) develop our understanding of how the climate of Antarctica varies in both time and space and 2) provide data to use as context for remote sensing studies and forcing for dynamical process models. Given that there are a number of different regional climate models available that explicitly simulate Antarctic climate, understanding inter- and intra model variability is important.Here, inter- and intra-model variability in Antarctic-wide regional climate model output is assessed for: snowfall; rainfall; snowmelt and near-surface air temperature within a cloud-based virtual lab framework. State-of-the-art regional climate model runs from the Antarctic-CORDEX project using the RACMO, MAR and MetUM models are used, together with the ERA5 and ERA-Interim reanalyses products. Multiple simulations using the same model and domain boundary but run at either different spatial resolutions or with different driving data are used. Traditional analysis techniques are exploited and the question of potential added value from more modern and involved methods such as the use of Gaussian Processes is investigated. The advantages of using a virtual lab in a cloud based environment for increasing transparency and reproducibility, are demonstrated, with a view to ultimately make the code and methods used widely available for other research groups.

Download Full-text

Model cascade from meteorological drivers to river flood hazard: flood-cascade v1.0

Geoscientific Model Development ◽

10.5194/gmd-14-4865-2021 ◽

2021 ◽

Vol 14 (8) ◽

pp. 4865-4890

Author(s):

Peter Uhe ◽

Daniel Mitchell ◽

Paul D. Bates ◽

Nans Addor ◽

Jeff Neal ◽

...

Keyword(s):

Climate Change ◽

Climate Models ◽

River Flow ◽

Climate Model ◽

Climate Modeling ◽

Flood Hazard ◽

Spatial Scales ◽

Computational Cost ◽

Data Sets ◽

Flood Inundation

Abstract. Riverine flood hazard is the consequence of meteorological drivers, primarily precipitation, hydrological processes and the interaction of floodwaters with the floodplain landscape. Modeling this can be particularly challenging because of the multiple steps and differing spatial scales involved in the varying processes. As the climate modeling community increases their focus on the risks associated with climate change, it is important to translate the meteorological drivers into relevant hazard estimates. This is especially important for the climate attribution and climate projection communities. Current climate change assessments of flood risk typically neglect key processes, and instead of explicitly modeling flood inundation, they commonly use precipitation or river flow as proxies for flood hazard. This is due to the complexity and uncertainties of model cascades and the computational cost of flood inundation modeling. Here, we lay out a clear methodology for taking meteorological drivers, e.g., from observations or climate models, through to high-resolution (∼90 m) river flooding (fluvial) hazards. Thus, this framework is designed to be an accessible, computationally efficient tool using freely available data to enable greater uptake of this type of modeling. The meteorological inputs (precipitation and air temperature) are transformed through a series of modeling steps to yield, in turn, surface runoff, river flow, and flood inundation. We explore uncertainties at different modeling steps. The flood inundation estimates can then be related to impacts felt at community and household levels to determine exposure and risks from flood events. The approach uses global data sets and thus can be applied anywhere in the world, but we use the Brahmaputra River in Bangladesh as a case study in order to demonstrate the necessary steps in our hazard framework. This framework is designed to be driven by meteorology from observational data sets or climate model output. In this study, only observations are used to drive the models, so climate changes are not assessed. However, by comparing current and future simulated climates, this framework can also be used to assess impacts of climate change.

Download Full-text

Regional Climate Model Emulator based on Deep Learning: Concept and First Evaluation of a Novel Hybrid Downscaling Approach

10.21203/rs.3.rs-725819/v1 ◽

2021 ◽

Author(s):

Antoine Doury ◽

Samuel Somot ◽

Sébastien Gadat ◽

Aurélien Ribes ◽

Lola Corre

Keyword(s):

Climate Change ◽

Network Architecture ◽

Large Scale ◽

Climate Models ◽

Climate Model ◽

Regional Climate ◽

Regional Climate Models ◽

Future Climate ◽

Near Surface ◽

The Way

Abstract Providing reliable information on climate change at local scale remains a challenge of first importance for impact studies and policymakers. Here, we propose a novel hybrid downscaling method combining the strengths of both empirical statistical downscaling methods and Regional Climate Models (RCMs). The aim of this tool is to enlarge the size of high-resolution RCM simulation ensembles at low cost.We build a statistical RCM-emulator by estimating the downscaling function included in the RCM. This framework allows us to learn the relationship between large-scale predictors and a local surface variable of interest over the RCM domain in present and future climate. Furthermore, the emulator relies on a neural network architecture, which grants computational efficiency. The RCM-emulator developed in this study is trained to produce daily maps of the near-surface temperature at the RCM resolution (12km). The emulator demonstrates an excellent ability to reproduce the complex spatial structure and daily variability simulated by the RCM and in particular the way the RCM refines locally the low-resolution climate patterns. Training in future climate appears to be a key feature of our emulator. Moreover, there is a huge computational benefit in running the emulator rather than the RCM, since training the emulator takes about 2 hours on GPU, and the prediction is nearly instantaneous. However, further work is needed to improve the way the RCM-emulator reproduces some of the temperature extremes, the intensity of climate change, and to extend the proposed methodology to different regions, GCMs, RCMs, and variables of interest.

Download Full-text

Comparison of different infiltration schemes in the regional climate model REMO

10.5194/egusphere-egu21-9874 ◽

2021 ◽

Author(s):

Daniel Abel ◽

Katrin Ziegler ◽

Felix Pollinger ◽

Heiko Paeth

Keyword(s):

Regional Climate Model ◽

Climate Models ◽

Climate Model ◽

Root Zone ◽

Regional Climate ◽

Climate Projections ◽

Soil Hydrology ◽

Spatiotemporal Resolution ◽

Development Fund ◽

Near Surface

The European Regional Development Fund-Project BigData@Geo aims to create highly resolved climate projections for the model region of Lower Franconia in Bavaria, Germany. These projections are analyzed and made available to local stakeholders of agriculture, forestry, and viniculture as well as general public. Since regional climate models&#8217; spatiotemporal resolution often is too coarse to deal with such local issues, the regional climate model REMO is improved within the frame of the project in cooperation with the Climate Service Center Germany (GERICS).Accurate and highly resolved climate projections require realistic modeling of soil hydrology. Thus, REMO&#8217;s original bucket scheme is replaced by a 5-layer soil scheme. It allows for the representation of water below the root zone. Evaporation is possible solely from the top layer instead of the entire bucket and water can flow vertically between the layers. Consequently, the properties and processes change significantly compared to the bucket scheme. Both, the bucket and the 5-layer scheme, use the improved Arno scheme to separate throughfall into infiltration and surface runoff.In this study, we examine if this scheme is suitable for use with the improved soil hydrology or if other schemes lead to better results. For this, we (1) modify the improved Arno scheme and further introduce the infiltration equations of (2) Philip as well as (3) Green and Ampt. First results of the comparison of these four different schemes and their influence on soil moisture and near-surface atmospheric variables are presented.

Download Full-text

A global decadal mode in a high-end climate model and in observations: Any connection to the solar cycle?

10.5194/egusphere-egu21-4961 ◽

2021 ◽

Author(s):

Michael Ghil ◽

Yizhak Feliks ◽

Justin Small

Keyword(s):

Solar Cycle ◽

Sea Level ◽

North Atlantic ◽

Climate Model ◽

Decadal Variability ◽

Singular Spectrum Analysis ◽

Coupled System ◽

The North ◽

Spatio Temporal ◽

Decadal Mode

The present work addresses two persistent quandaries of the climate sciences: (i) the existence of global oscillatory modes in the coupled ocean&#8211;atmosphere system; and (ii) solar effects on this coupled system. Interannual oscillatory modes, atmospheric and oceanic, are present in several large regions of the globe. We examine here interannual-to-decadal variability over the entire globe in the Community Earth System Model (CESM) and in the NCEP-NCAR reanalysis, and apply multichannel singular spectrum analysis (MSSA) to these two datasets.In the fully coupled CESM1.1 model, with its resolution of 0.1 &#215; 0.1 degrees in the ocean and 0.25 &#215; 0.25 degrees in the atmosphere, the &#64257;elds analyzed are surface temperatures, sea level pressures and&#160; the 200-hPa geopotential. The simulation is 100-yr long and the last 66 yr are used in the analysis. The two statistically signi&#64257;cant periodicities in this IPCC-class model are 11 and 3.4 yr.In the reanalysis, the &#64257;elds of sea level pressure and of 200-hPa geopotential are analyzed at its resolution of 2.5 &#215; 2.5 degrees over the 68-yr interval 1949&#8211;2016. Oscillations with periods of 12 and 3.6 yr are found to be statistically signi&#64257;cant in this dataset. The spatio-temporal patterns&#160; of the oscillations in the two datasets are quite similar. The spatial pattern of these&#160; global oscillations over the North Paci&#64257;c and North Atlantic resemble the Paci&#64257;c Decadal Oscillation and the interannual variability found in the western North Atlantic, respectively.The two global modes, with their 11&#8211;12-yr and 3.4&#8211;3.6-yr periodicities, are quite robust, suggesting potential contributions of both to predictability at 1&#8211;3-yr horizons. On the other hand, the CESM run has no year-to-year changes in the prescribed insolation, excluding any role of the solar cycle in the model&#8217;s 11-yr mode. The solar cycle is present, however, in the reanalysis, since it is present in nature and hence it does affect the observations. We speculate, therefore, that regional oscillations &#8212; with their distinct near-periodicities and spatial patterns &#8212; are synchronized&#160; over the globe, thus yielding both the global oscillatory modes found in CESM. In nature, the decadal mode could be further synchronized with the solar cycle, but that does not seem to be the case, given the slight difference in period &#8212; 12 yr for the reanalysis and 11 yr for the solar cycle, which makes them drift in and out of phase.The work&#8217;s tentative conclusion is, therefore: (i) yes, there are global oscillatory modes in the climate system, especially a decadal mode; but (ii) no, this mode has little or nothing to do with the solar cycle.

Download Full-text