CL5.3.2

The practice of initialized subseasonal, seasonal and decadal climate prediction has matured considerably in recent years, with real-time subseasonal and decadal multi-system ensembles joining those established previously for the seasonal to multi-seasonal range. However, substantial scientific, modelling, and informational challenges remain that must be overcome in order to more fully realize the potential for such predictions to serve societal needs. This presentation will examine five such challenges that the World Climate Research Programme’s Working Group on Subseasonal to Interdecadal Prediction (WGSIP) has identified as crucial for further advancing capabilities for translating the inherent predictability of the Earth system into actionable predictive information. Surmounting these challenges will bring nearer an envisaged future in which global users have access to such information specific to individual needs, across Earth system components and on a continuum of time scales, with degrees of confidence, limitations and uncertainties clearly indicated, as well as tools to guide optimal actions.

How to cite: Merryfield, W., Baehr, J., Batté, L., Beraki, A., Hermanson, L., Hudson, D., Johnson, S., Lee, J.-Y., Massonnet, F., Muñoz, Á., Orsolini, Y., Ren, H.-L., Saurral, R., Smith, D., Takaya, Y., and Raghavan, K.: Some key challenges for subseasonal to decadal prediction research, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10621, https://doi.org/10.5194/egusphere-egu22-10621, 2022.

The Kuroshio Extension (KE), an eastward-flowing jet located in the Pacific western boundary current system, exhibits prominent seasonal-to-decadal variability, which is crucial for understanding climate variations in northern midlatitudes. We explore the representation, predictability, and prediction skill for the KE in the GFDL SPEAR (Seamless System for Prediction and EArth System Research) coupled model. Two different approaches are used to generate coupled reanalyses and forecasts: (1) restoring the coupled model’s SST and atmospheric variables toward existing reanalyses, or (2) assimilating SST and subsurface observations into the coupled model without atmospheric assimilation.  Both systems use an ocean model with 1^o resolution and capture the largest sea surface height (SSH) variability over the KE region. Assimilating subsurface observations appears to be critical to reproduce the narrow front and related oceanic variability of the KE jet in the coupled reanalysis. We demonstrate skillful retrospective predictions of KE SSH variability in monthly (up to 1 year) and annual-mean (up to 5 years) KE forecasts in the seasonal and decadal prediction systems, respectively. The prediction skill varies seasonally, peaking for forecasts initialized in January and verifying in September due to the winter intensification of North Pacific atmospheric forcing. We show that strong large-scale atmospheric anomalies generate deterministic oceanic forcing (i.e., Rossby waves), leading to skillful long-lead KE forecasts. These atmospheric anomalies also drive Ekman convergence/divergence that forms ocean memory, by sequestering thermal anomalies deep into the winter mixed layer that re-emerge in the subsequent autumn. The SPEAR forecasts capture the recent negative-to-positive transition of the KE phase in 2017, projecting a continued positive phase through 2022.

How to cite: Joh, Y., Delworth, T., Wittenberg, A., Cooke, W., Yang, X., Zeng, F., Jia, L., Lu, F., Johnson, N., Kapnick, S., Rosati, A., Zhang, L., and McHugh, C.: Seasonal-to-decadal variability and predictability of the Kuroshio Extension in the GFDL Coupled Ensemble Reanalysis and Forecasting system, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6756, https://doi.org/10.5194/egusphere-egu22-6756, 2022.

Climate models are an important tool in our understanding of the climate system. Among other things, we use them together with initialisation procedures to predict the climate from a few weeks to more than a decade. While the community has demonstrated prediction skill for various climate modes on these time scales in the past years, we have also encountered problems. One is the non-stationarity of prediction skill over the past century in seasonal and decadal predictions. It was shown in multiple prediction systems and for multiple variables that prediction skill varies over time. Potential reasons for this non-stationarity was found in the changing state of the North Atlantic system on multi-decadal scales and the limited representation of physical processes within the model. While on the one side this feature of climate predictions leaves uncertainties for future predictions it also highlights windows of opportunity and challenges within climate models. 

We investigate the past century for this non-stationarity with a special focus on the North Atlantic Oscillation, and how the North Atlantic sector changes during these low prediction skill periods. We will demonstrate the limited predictability of features of the North Atlantic Oscillation, like the movement of its activity centres, as well as its implication for the Signal-to-Noise paradox. We also discuss the implications of non-stationarity model prediction skill for the development on future prediction systems and which processes are most likely the reason for the current challenges the community faces.

How to cite: Düsterhus, A., Borchert, L., Mayer, B., Koul, V., Pohlmann, H., Brune, S., and Baehr, J.: What can the last century teach us about climate models?, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5377, https://doi.org/10.5194/egusphere-egu22-5377, 2022.

The challenges of climate prediction are varied and complex. On the one hand they include conceptual and mathematical questions relating to the consequences of model error and the information content of observations and models. On the other, they involve practical issues of model and ensemble design, and the statistical processing of data.

A route to understanding the complexity of these challenges is to study them using low-dimensional nonlinear systems that encapsulate the key characteristics of climate and climate change. Doing so facilitates the fast generation of very large ensembles with a variety of designs and target goals. These idealised ensembles can provide a solid foundation for improving the design of ESM/GCM ensembles, making them better suited to evaluating the risks associated with climate change and to providing end-user support through climate services.

The ODESSS project - Optimizing the Design of Ensembles to Support Science and Society - is using low-dimensional nonlinear systems to provide solid foundations for the design of climate change ensembles with climate models. In this presentation I will introduce the project and the concepts behind it.

First I will discuss the essential characteristics required of a low dimensional nonlinear system to be able to capture the process of climate prediction. Results will then be presented from the coupled Lorentz ’84 - Stommel ’61 system; a low-dimensional nonlinear system which has these characteristics. These results will be used to illustrate the dangers of confounding natural variability with the consequences of initial condition uncertainty[1], and to demonstrate why risk assessments require much larger initial condition ensembles than are currently available with today’s ESMs/GCMs.

The difference between micro and macro initial condition ensembles [2,3] will then be introduced, along with an explanation of how this leads to a requirement for ensembles of ensembles: the former exploring macro-initial-condition-uncertainty, the latter micro-initial-conditional-uncertainty. The importance of this distinction will be illustrated with both new results from the Lorentz ‘84 - Stommel ‘61 system, and also a GCM[3]. I will highlight the challenges in designing these ensembles of ensembles to be most informative. These challenges relate closely to the problems of initialization and the optimal use of observations.

Finally the subject of model error, multi-model and perturbed-physics ensembles will be discussed. The impact of model error on climate predictions can only be studied effectively if climate change can be accurately quantified within each model. To begin to explore the consequences of model error for climate predictions therefore requires ensembles of ensembles of ensembles: perturbed-physics or multi-model ensembles which  themselves consist of both macro and micro initial condition ensembles. Some approaches will be presented for how low-dimensional systems can be used to optimise the design of such multi-layered ensembles with ESMs/GCMs where computational constraints are more restrictive.

[1] Daron and Stainforth, On predicting climate under climate change. ERL, 2013.

[2] Stainforth et al., Confidence, uncertainty and decision-support relevance in climate predictions. Phil. Trans Roy. Soc., 2007.

[3] Hawkins et al., Irreducible uncertainty in near-term climate projections. Climatic Change, 2015.

How to cite: Stainforth, D.: Ensembles of ensembles of ensembles: On using low-dimensional nonlinear systems to design climate prediction experiments, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3885, https://doi.org/10.5194/egusphere-egu22-3885, 2022.

In the last decade, high demands from stakeholders and policymakers have driven unprecedented research efforts directed to improve climate predictability. Nevertheless, attempts to get operational climate predictions on seasonal time scales have been far from skillful for a long time. Based on sources of predictability from the ocean, atmosphere and land processes, current state-of-the-art prediction systems are approaching operational predictability. This work examines and compares the ability of different prediction systems to simulate the variability of sea surface temperatures (SSTs) associated with El Niño-Southern Oscillation (ENSO) and the ENSO-forced response of hydroclimate variability in the North Atlantic-Europe (NAE) region. Seasonal hindcasts derived from two generations of the Norwegian Earth System Model (NorESM1-ME and NorESM2-MM) are used in addition to C3S data to generate time series of year-to-year variability that are validated against observational data. Our results reveal both the advantages and the limitations of these prediction systems to simulate ENSO-related variability, identifying model biases that prevent skillful predictability. Further efforts must be aimed at mitigating these biases in order to achieve fully operational predictions of paramount importance for the benefit of society.

How to cite: Suarez-Moreno, R., Svendsen, L., Bethke, I., King, M. P., Chiu, P.-G., and Bilge, T. A.: Towards operational climate prediction: ENSO-related variability as simulated in a set of state-of-the-art seasonal prediction systems, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7652, https://doi.org/10.5194/egusphere-egu22-7652, 2022.

The Horn of Africa is known to be prone to climate impacts; the frequent occurrence of droughts and floods creates vulnerable conditions in the region. Gaining knowledge on (sub-)seasonal weather prediction and generating more reliable long-term forecasts is an important asset in building resilience. Most of the region is characterized by a bimodal precipitation cycle with rainfall seasons in boreal spring (March–May), termed the long rains, and boreal autumn (October–November), termed the short rains. Previous studies on seasonal forecasting focused mostly on empirical linear regression methods using information from ocean–atmosphere modes. To date, the potential of more complex methods, such as machine learning approaches, in improving seasonal precipitation predictability in the Horn of Africa still remains understudied. 

In this study, machine learning models targeting precipitation during the long rains are developed. The focus on the long rains is motivated by the fact that it is the main rain season in the region and the sources of predictability have proven to be more difficult to pin down. The long rain season has a weak internal coherence and looking at the months separately has proven to enhance prediction skill. Therefore, machine learning models are constructed for the different months (March, April, and May) separately at lead times of 1–3 months. Following an extensive survey of literature, the predictors of the long rain precipitation at seasonal timescales selected in this study include coupled oceanic-atmospheric oscillation indices (such as MJO, ENSO and PDO), regions of zonal winds over 200mb and 850mb and sea-surface temperature (SST) regions with strong correlation to long rain precipitation. Further, a selection of additional terrestrial and oceanic predictors is guided by Lagrangian transport modeling, used to identify the regions sourcing moisture during the long rains. This set of predictors include soil moisture, land surface temperature, normalized vegetation index (NDVI), leaf area index (LAI) and SST, which are averaged over the climatological source region of long rain precipitation. Finally, we provide new insights into the predictability of long rain precipitation at seasonal timescales by analyzing the relative importance of the different predictors used for developing the machine learning model.

How to cite: Deman, V., Koppa, A., and Miralles, D.: Seasonal Forecasting of Horn of Africa’s Long Rains Using Physics-Guided Machine Learning, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9719, https://doi.org/10.5194/egusphere-egu22-9719, 2022.

A novel temperature swing index (TSI) is formed to measure the extreme surface temperature variations associated with the winter extratropical storms. The seasonal prediction skill of the winter TSI over North America was assessed versus ERA5 data using GFDL’s new SPEAR seasonal prediction system. The location with the skillful TSI prediction shows distinctive geographic pattern from that with skillful seasonal mean temperature prediction, thus the skillful prediction of TSI provides additive predictable climate information beyond the traditional seasonal mean temperature prediction. The source of the seasonal TSI prediction can be attributed to year-to-year variations of ENSO, North Pacific Oscillation and NAO. These results point towards providing skillful prediction of higher-order statistical information related to winter temperature extremes, thus enriching the seasonal forecast products for the research community and decision makers beyond the seasonal mean.

How to cite: Yang, X., delworth, T., Jia, L., Johnson, N., Lu, F., and MacHugh, C.: On the seasonal prediction and predictability of winter temperature swings over North America, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10340, https://doi.org/10.5194/egusphere-egu22-10340, 2022.

Over East Asia, reliable forecasts of boreal spring droughts and pluvials can provide time window of opportunities to mitigate their adverse effects. Here, we aim to assess the seasonal prediction skill of boreal spring droughts and pluvials over East Asia (EA), using NMME and atmospheric-only global climate model (AGCM) simulations. Results show that NMME models show a better prediction skill of pluvials than that of droughts, indicating asymmetry in the prediction skill. This asymmetric tendency is also found in the prediction skill of sea surface temperature (SST) during the corresponding drought and pluvial years. Results from the AGCM simulations show asymmetry in the prediction skills of spring droughts and pluvials, indicating the limited predictability of SST-teleconnections in the model physics. The findings of this study prioritize a need to improve the representation of sea-air interactions during drought years in the current climate models.

How to cite: Kim, B.-H. and Kam, J.: Asymmetry in the prediction skills of NMME models for springtime droughts and pluvials over East Asia, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10950, https://doi.org/10.5194/egusphere-egu22-10950, 2022.

Northern Hemisphere anthropogenic aerosols influence Southeast and East Asian summer monsoon precipitation. In the late 20^th century, both the East Asian and the South Asian summer monsoons weakened because of increased emissions of anthropogenic aerosols over Asia, counteracting the warming effect of increased greenhouse gases (GHGs). Changes in the anthropogenic aerosols burden in the Northern Hemisphere, and specifically over the Asian continent, may also have affected the sub-seasonal evolution of the summer monsoon. During the spring 2020, when restrictions to contain the spread of the coronavirus were implemented worldwide, reduced emissions of gases and aerosols were detected also over Asia.

Following on from the above and using the EC-Earth3 coupled model, a case-study forecast for summer 2020 (May 1st start date) has been designed and produced with and without the reduced atmospheric forcing due to covid-19 in the SSP2-4.5 baseline scenario, as estimated and adopted within CMIP6 DAMIP covidMIP experiments (hereinafter “covid-19 forcing”). The forecast ensembles (sensitivity and control experiments, meaning with and without covid-19 forcing) consist of 60 members each to better account for the internal variability (noise) and to maximize the capability to identify the effects of the reduced emissions.

The analysis focuses on  the effects of the covid-19 forcing, in particular the reduction of anthropogenic aerosols, on the forecasted evolution of the monsoon, with a specific focus on the performance in predicting the summer precipitation over India and over other parts of  South and East Asia. Changes in the performance of the prediction for specific aspects of the monsoon, like the onset and the length of the season, are evaluated as well.

How to cite: Cherchi, A., Alessandri, A., Tourigny, E., Acosta Navarro, J. C., Ortega, P., Davini, P., Volpi, D., Catalano, F., and van Noije, T.: Effects of aerosols reduction on the Asian summer monsoon prediction: the case of summer 2020, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11562, https://doi.org/10.5194/egusphere-egu22-11562, 2022.

Decadal climate prediction is a scientific endeavour of potentially large societal impacts. Yet such predictions remain challenging, as they predict climate skilfully only under certain circumstances or in specific regions. Moreover, decadal climate prediction simulations rely on dedicated coupled climate model simulations that are particularly expensive. In this study, we build upon earlier research by Menary et al. (2021) in search of a method to make skilful and cheap decadal climate predictions by constructing predictions from existing climate model simulations using the so-called analogue method.

The analogue method draws on the idea that there is decadal memory in the climatic state at the start of a prediction. This method identifies the observed state of the climate system at the start of a prediction and then screens the archive of available model simulations for comparable climatic states. It then selects a number of modelled climate states that are similar to the observed situation, and uses the years after the selected simulated climate states as prediction. Using a simple analogue method based on temperature trends in the North Atlantic basin, Menary et al. (2021) demonstrated skilful prediction of North Atlantic SST on par with dynamical decadal prediction simulations. In this study, we refine the original method by using more sophisticated algorithms to select the analogues, and choosing decadal prediction of seasonal European climate as our target. These new selection algorithms include multivariate regression at different time lags as well as non-linear methods.

Menary, MB, J Mignot, J Robson (2021) Skilful decadal predictions of subpolar North Atlantic SSTs using CMIP model-analogues. Environ. Res. Lett. 16 064090. https://doi.org/10.1088/1748-9326/ac06fb

How to cite: Borchert, L., Menary, M., and Mignot, J.: An analogue approach to predicting European climate, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1448, https://doi.org/10.5194/egusphere-egu22-1448, 2022.

Ireland is bordering the North Atlantic, and its climate is dominated by its climate modes on short to longer timescales. The Atlantic low-pressure systems, Jetstream variabilities and airmasses are features of the atmospheric circulation, which also contribute to the climate this region.  So, a long-term climate prediction of Ireland is majorly controlled by the ocean, and by other atmospheric components.

The Ocean has shown good capabilities for decadal to multi-decadal climate predictions, hence, our study adapted a coupled model to investigate seasonal changes in the climate on annual to multi-annual timescales within the Max Planck Institute for Meteorology Earth System Model (MPI-ESM).  Initialized prediction is extended to multi-decadal timescale up onto twenty lead years, and we study prediction capabilities for common climate variables in and around , by identifying major drivers and documenting their prediction skills.  Our results have shown prediction skill for surface temperature over longer timescales, and we explore these capabilities for other variables of interest.  This study opens new opportunities for better long-term predictions of climate components in the region, and our results are relevant for strategic planning.

How to cite: Ogungbenro, S., O'Beirne, C., and Düsterhus, A.: Long-term climate prediction for Ireland and its surrounding, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6767, https://doi.org/10.5194/egusphere-egu22-6767, 2022.

Anthropogenic CO2 emissions are associated with global warming in the late 20th century and beyond. Climate-carbon feedbacks will likely result in a higher airborne fraction of emitted CO2 in the future. However, the variability in atmospheric CO2 growth rate is largely controlled by natural variability and is poorly understood. This can interfere with the attribution  of slowing CO2 growth rates  to reducing emissions during the implementation of the Paris Agreement. There is thus a need to both improve our understanding of the processes controlling the global carbon cycle and establish a near-term prediction system of the climate and carbon cycle.

As part of the 4C (Carbon Cycle Interactions in the Current Century) project, the Barcelona Supercomputing Center is implementing a new system for near-term prediction of the climate and carbon cycle interactions using EC-Earth3-CC, the Carbon Cycle version of the EC-Earth3 Earth System Model. This new system is based on the existing operational climate prediction system developed by the BSC, contributing to the WMO Global Annual to Decadal Climate Update. EC-Earth3-CC comprises the IFS atmospheric model, the NEMO ocean model, the PISCES ocean biogeochemistry model, the LPJ-GUESS dynamic vegetation model, the TM5 global atmospheric transport model and the OASIS3 coupler. The system uses initial conditions from in-house ocean biogeochemical and land/vegetation reconstructions based on global atmospheric/ocean reanalyses. By performing retrospective decadal predictions of ocean and land carbon uptake we are able to evaluate the performance of the system in predicting CO2 fluxes and atmospheric CO2 concentrations.

We will present results from the latest concentration- and emission-driven retrospective predictions (or hindcasts) using our system, highlighting the skill and biases of the carbon fluxes and atmospheric CO2. We will also present future predictions for 2022 and beyond, a prototype for the operational system for prediction of future atmospheric CO2.

How to cite: Tourigny, E., Bernardello, R., Sicardi, V., Ortega, P., Ruprich Robert, Y., Lapin, V., Acosta Navarro, J. C., Bilbao, R., Meier, A., Li, H., and Ilyina, T.: Near-term prediction of the global carbon cycle using EC-Earth3-CC, the Carbon Cycle version of the EC-Earth3 Earth System Model, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10228, https://doi.org/10.5194/egusphere-egu22-10228, 2022.

The predictability of phytoplankton abundance in the Barents Sea is explored in the CMIP6 decadal prediction runs with the Norwegian Climate Prediction Model (NorCPM1), together with satellite data and in situ measurements. The model successfully predicts a maximum in the observed phytoplankton abundance in 2007 up to five years in advance, which is associated with a strong predictive skill of 2007 minimum extent of the summer sea ice concentration. The underlying mechanism is an event of anomalously high heat transport into the Barents Sea that is seen both in the model and in situ observations. These results are an important step towards marine ecosystem predictions.

How to cite: Fransner, F., Årthun, M., Bethke, I., Counillon, F., Samuelsen, A., Tjiputra, J., Olsen, A., and Keenlyside, N.: Skillful Prediction of Barents Sea Phytoplankton Concentration, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12989, https://doi.org/10.5194/egusphere-egu22-12989, 2022.

In the Northwest Atlantic (NWA), including the Labrador Sea, interactions between the atmosphere, ocean circulation, and sea ice play a critical role in regulating the global climate system. The ocean and climate in this region observe rapid and unprecedented, anthropogenically forced changes to the physical environment and biosphere with downstream effects. Future projections of NWA circulation and sea ice can help address pressing questions about these changes and mitigate their potential impacts on the global carbon cycle, coastal communities, and transportation. However, the spatial resolution of current climate models is often insufficient to accurately represent important features in the NWA, such as the location and strength of the Gulf Stream and Labrador Current and their dynamical interactions. This can lead to biases in the model’s mean state, and a misrepresentation of the temporal and spatial scales of ocean variability, e.g., mesoscale eddies, deep convection. Regional ocean models with grid spacing <10 km, forced by global climate simulations, can be used to improve estimates of historical and future circulation and hydrography. However, given the limited spatial resolution and biases in global climate models, a challenge of downscaling their simulations is the appropriate reconstruction of the forcing fields.

Here, we present preliminary results of future projections of NWA circulation and sea ice based on downscaled global climate simulations. These projections are performed using an eddy-resolving, coupled circulation-sea ice model based on the Regional Ocean Modeling System (ROMS) and the Los Alamos Sea Ice Model (CICE). We will focus on the value of correcting biases in the mean and variance of the forcing. We further explore the need of including missing spatial and temporal scales in the atmospheric forcing that are not captured by the global models. Implications for the design of model experiments for future projections will be discussed.

How to cite: Renkl, C. and Oliver, E.: Bias Correction and Spatiotemporal Scales for Downscaling Future Projections of Northwest Atlantic Circulation and Sea Ice, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10467, https://doi.org/10.5194/egusphere-egu22-10467, 2022.

In order to improve our ability to predict the near-term evolution of climate, it may be important to accurately predict the evolution of atmospheric CO2, and thus carbon sinks. Following on from process-driven improvements of decadal predictions in physical oceanography, we focus on improving our understanding of the internal processes and variables driving CO2 uptake by the North Atlantic ocean. Specifically, we use the CMIP6 model IPSLCM6A to investigate the drivers of ocean-atmosphere CO2 flux variability in the North Atlantic subpolar gyre (NA SPG) on seasonal to decadal timescales. We find that DpCO2 (CO2 partial pressure difference between atmosphere and ocean) variability dominates over sea surface temperature (SST) and sea surface salinity (SSS) variability on all timescales within the NA SPG. Meanwhile, at the ice-edge, there are significant roles for both ice concentration and surface winds in driving the overall CO2 flux changes. Investigating the interannual DpCO2 variability further, we find that this variability is itself driven largely by variability in simulated mixed layer depths in the northern SPG. On the other hand, SSTs show an important contribution to DpCO2 variability in the southern SPG and on longer (decadal) timescales. Initial extensions into a multi-model context show similar results. By determining the key regions and processes important for skilful decadal predictions of ocean-atmosphere CO2 fluxes, we aim to both improve confidence in these predictions as well as highlight key targets for climate model improvement. 

How to cite: Menary, M., Mignot, J., Bopp, L., and Kwiatkowski, L.: Processes of interannual internal variability of the CO2 flux at the air-sea interface in IPSLCM6A, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9618, https://doi.org/10.5194/egusphere-egu22-9618, 2022.