Societally-Relevant Multi-Year Climate Predictions Workshop, March 28-30, 2022

US CLIVAR (US Climate Variability and Predictability Program)

The workshop will bring together scientists and stakeholders engaged in the modeling and forecasting communities as well as the private sector.

Seasonal predictability research focuses on time scales less than a year, when both initialization and boundary forcing from sea, land, and ice provide operational forecast skill. About a decade ago, driven by high demands from stakeholders and policymakers, a few large modeling centers began a sustained effort to address “near-term” or decadal prediction of averages over several years (e.g., forecasts of Years 1-10, Years 2-5, Years 6-9 means), hoping to exploit the potential for skill revealed by many years of observational and modeling studies into decadal time-scale processes, both within the climate system and externally forced.

More info: https://usclivar.org/meetings/multi-year-workshop

Filter displayed posters (52 keywords)

water management (2) show more... AMOC (1) Arctic (1) CMIP (1) CMIP6 (1) Central Pacific ENSO events (1) Decadal (1) Decadal Pacific variability (1) Decadal Prediction (1) Decadal prediction (1) ENSO prediction (1) Extremes (1) Indian Monsoon (1) Kuroshio Extension (1) Limits of Predictability (1) Model-analog (1) NAO (1) Northeast Pacific marine heatwaves (1) Ocean acidification (1) Pacemakers (1) Pacific (1) Physics guided data-driven models (1) Power Sector (1) SST trend (1) Sea level (1) Sea surface temperature (1) Solar (1) Southeastern South America (1) Southern Ocean (1) Wind (1) agriculture (1) case study (1) climate change adaptation (1) climate forecast skill (1) decadal climate forecast (1) decadal prediction (1) drought (1) high resolution (1) initialized models (1) large-ensemble datasets (1) machine learning (1) model (1) model diversity (1) ocean (1) persistence (1) predictability (1) recalibration (1) reservoir (1) statistical downscaling (1) streamflow (1) temperature predictions (1) user co-production (1)

Show Posters: with Video Chat with Recorded Presentation

▼ Day 1: Ocean Processes/Predictability and Applications Back to top

Multi-Year Predictability of the Kuroshio Extension in the CESM Decadal Prediction Systems

Who M. Kim, Stephen Yeager, Gokhan Danabasoglu, and Ping Chang

Abstract

CESM1 high-resolution decadal prediction (HRDP) simulations show significant skill, up to lead year 4, in predicting the Kuroshio Extension (KE), which is significant improvement compared to 1-2 year predictability of the low-resolution equivalent (DPLE). The source of skill appears to be the initialized sea surface height anomaly in the central North Pacific that propagates westward more obviously in HRDP than DPLE, guided by the well represented KE.

Presented by

Who Kim <whokim@ucar.edu>

Institution

National Center for Atmospheric Research

Keywords

Kuroshio Extension

Chat with Presenter

Available March 28th 12:40-1:40pm MST

Join the Discussion

Multi-year drought predictions at high-resolution for the water management sector in Germany

A. Paxian, K. Reinhardt, K. Pankatz, A. Pasternack , M.P. Lorza Villegas, M. Scheibel, A. Hoff, B. Mannig, P. Lorenz, B. Früh

Abstract

The Wupper catchment water board in north-western Germany requires multi-year predictions of drought indices, strongly connected to dam water levels, such as the Standardized Precipitation Index (SPI). They need predictions at high spatial resolution to manage water level and quality of 14 dams in a total catchment area of 813 km². We analysed different temporal aggregations impacting on different management processes: annual means and multi-year seasonal means of hydrological seasons (February-March-April, May-June-July, August-September-October, November-December-January) for forecast years 1-3. The multi-year forecasts are based on the MPI-ESM-LR decadal prediction system with 16 ensemble members. To reach a higher spatial resolution, the global simulations (200 km resolution) have been statistically downscaled in Germany (11 km). A statistical recalibration has been performed to correct model errors and adjust the ensemble spread. The SPI has been calculated in standardizing 3-year mean and 3-year seasonal mean precipitation. The prediction skill has been evaluated in comparison to HYRAS observations during 1962-2020 based on the correlation and the RPSS. The significance of prediction skill has been evaluated using non-parametric bootstraps. The multi-year skill of the global prediction system could be preserved by statistical downscaling at high resolution and be further improved by statistical recalibration in Germany. Promising results could be found for three-year annual and seasonal means of SPI predictions for several regions and seasons in Germany. The multi-year SPI prediction for forecast years 1-3 has been published within a user-oriented product sheet on the website of the Copernicus Climate Change Service (https://climate.copernicus.eu/decadal-predictions-infrastructure). This ‘sectoral application of decadal predictions’ provides a probabilistic prediction for the Wupper catchment and whole north-western Germany to cover as well needs of neighboring water boards. The display combines the multi-year prediction (via the color of dots) with its prediction skill (via the size of dots). Dry conditions (negative SPI values) have been predicted in most of north-western Germany for 2021-2023, whereas the prediction skill is larger than that of the observed climate mean in several parts of the region. The results of this product sheet have been discussed with the Wupper catchment board and other German water managers on a user workshop. The skill of these high-resolution multi-year predictions is considered to be promising to fulfil their needs. The product sheet is well understandable and can be used within the working routines of German water boards.

Presented by

Andreas Paxian <andreas.paxian@dwd.de>

Institution

Deutscher Wetterdienst (DWD), Germany

Keywords

decadal prediction, high resolution, drought, statistical downscaling, recalibration, case study, water management, user co-production

Chat with Presenter

Available March 28th 12:40-14:10 UMD

Join the Discussion

Diversity in North Atlantic variability and potential predictability on interannual to decadal timescales across CMIP6 models – with a focus on NAO-AMOC interactions

Annika Reintges (1), Jon I. Robson (1), Rowan T. Sutton (1), and Stephen G. Yeager (2)

Abstract

The variations of the winter climate in Europe are influenced by the North Atlantic Oscillation (NAO). Therefore, the ability to predict the NAO is of great value. Predictability of the NAO can be enabled through oceanic processes that are characterized by relatively long time scales, for example interannual to decadal. An important variable for the interannual to (multi-)decadal variability in the North Atlantic is the Atlantic Meridional Overturning Circulation (AMOC). The NAO and the AMOC are known to interact, but observational records of the AMOC are short and the details of this interaction are unknown. Thus, our understanding largely relies on climate model simulations. However, the interaction of NAO and AMOC is very model dependent. Here, we present the diversity across CMIP6 models in pre-industrial control experiments. The focus lies on variability and potential predictability of the NAO, the AMOC, their interaction, and related variables on interannual to decadal timescales. Regarding the NAO-AMOC interaction, there are large differences in the strength of their relationship, in the location (like the latitude of the AMOC), its periodicity and in the time-lag between both variables. Furthermore, we propose hypotheses of the causes for this diversity in the models. Specific processes involved in NAO-AMOC interaction might be of varying relative importance from model to model, for example, NAO-related buoyancy versus wind-forcing affecting the AMOC. Also, mean state difference like in the North Atlantic sea surface temperature might play an important role for causing differences in the variability and potential predictability across models.

Presented by

Annika Reintges <a.reintges@reading.ac.uk>

Institution

(1) National Centre for Atmospheric Science, University of Reading, Reading, UK; (2) National Center for Atmospheric Research, Boulder, CO, USA

Keywords

CMIP6, AMOC, NAO, model diversity

Chat with Presenter

Available Monday, 28 March 2022, 12:40-13:10 AND 13:45-14:10 MDT (UTC-6)

Join the Discussion

An Optimal Precursor of Northeast Pacific Marine Heatwaves and Central Pacific El Niño events

Antonietta Capotondi, Matthew Newman, Tongtong Xu, Emanuele Di Lorenzo

Abstract

Northeast Pacific marine heatwaves (MHWs) have devastating ecological consequences, so that the ability to predict them is of large societal value. The intensity of Northeast Pacific MHWs has been related to local stochastic atmospheric forcing with limited predictability, but their evolution and persistence may be controlled by large-scale climate influences. A Linear Inverse Model (LIM) containing both sea surface temperature (SST) and height (SSH) anomalies is used to identify the “optimal” conditions for observed Northeast Pacific MHWs events that developed two-to-four seasons later. These optimal initial conditions include SSH anomalies that are responsible for most of the MHW growth, suggesting the relevance of subsurface ocean dynamics. Moreover, Northeast Pacific MHW growth occurs as part of a basin-scale dynamical mode that links the North Pacific to central equatorial Pacific El Niño events, whose subsequent development may lengthen MHW duration.

Presented by

Antonietta Capotondi <antonietta.capotondi@noaa.gov>

Institution

University of Colorado/CIRES and NOAA/PSL

Keywords

Northeast Pacific marine heatwaves, Decadal Pacific variability, Central Pacific ENSO events

Chat with Presenter

Available 1:15-2:15pm MST

Join the Discussion

Seasonal to decadal marine ecological forecasting using the Community Earth System Model

Kristen M. Krumhardt, Nicole S. Lovenduski, Matthew Long, Jessica Luo, Keith Lindsay, Steven Yeager, and Cheryl Harrison

Abstract

Seasonal and interannual variations in ecological metrics, such as marine net primary production (NPP), contribute to the variability of available living marine resources, as well as influence critical carbon cycle processes. Here we provide a global overview of near-term (seasonal to 10 years) potential predictability of marine NPP using two ensembles of initialized retrospective multi-year forecasts from the Community Earth System Model (CESM). Interannual variations in marine NPP are potentially predictable in many areas of the ocean 1 to 3 years in advance, from temperate waters to the tropics, showing a substantial improvement over a simple persistence forecast. However, some regions, such as the subpolar Southern Ocean, show low potential predictability. We analyze how bottom-up drivers of marine NPP (nutrients, light, and temperature) contribute to its predictability. Regions where NPP is primarily driven by the physical supply of nutrients (e.g., subtropics) retain higher potential predictability than high-latitude regions where NPP is controlled by light and/or temperature (e.g., the Southern Ocean). We further examine NPP predictability in the world's Large Marine Ecosystems. With a few exceptions, we show that initialized forecasts improve potential predictability of NPP in Large Marine Ecosystems over a persistence forecast and may aid to manage living marine resources.

Presented by

Kristen Krumhardt <kristenk@ucar.edu>

Institution

National Center for Atmospheric Research

Keywords

Chat with Presenter

Available password: 1Xt1mO

Join the Discussion

Pacific contributions to multidecadal variability in the Arctic: A multi-model intercomparison

Lea Svendsen, Yu Kosaka, Bunmei Taguchi

Abstract

Instrumental records show multidecadal variability in Arctic winter surface temperature throughout the 20th century. This variability is thought to be caused by a combination of external forcing and internal variability, but their relative importance is not clear. Since decadal variability in the Pacific has been linked to ‘hiatuses’ and accelerated warming trends in global temperatures, we hypothesize that the Pacific could also impact decadal temperature trends in the Arctic. To investigate this, we compare two ensembles of historical all-forcing 20th century simulations with the Norwegian Earth System Model (NorESM): one fully coupled ensemble and one ensemble where momentum flux anomalies from reanalysis are prescribed over the Indo-Pacific Ocean to constrain Pacific sea surface temperature variability. We find that the combination of decadal variability in the tropical and extratropical Pacific contributed to up to 50% of the decadal Arctic warming trends during the 20th century, and we identify a mechanism for this connection involving both horizontal advection in the lower troposphere and adiabatic heating through a stratospheric pathway, mediated by Aleutian Low variations. To investigate the model diversity of the Pacific impact on multidecadal variability of Arctic surface temperature, we also compare CMIP5 and available CMIP6 preindustrial control simulations. The results suggest that the simulated strength of the Pacific-Arctic link is sensitive to the mean sea ice cover and the location and strength of the climatological Aleutian Low. These results have implications for predictability of Arctic climate on decadal timescales, as well as understanding present and future Arctic warming.

Reference: Svendsen, L., N. Keenlyside, M. Muilwijk, I. Bethke, N.-E. Omrani, Y. Gao (2021) Pacific contribution to decadal surface temperature trends in the Arctic during the 20th century. Climate Dynamics, doi:10.1007/s00382-021-05868-9

Presented by

Lea Svendsen <lea.svendsen@uib.no>

Institution

Geophysical Institute, University of Bergen, and Bjerknes Centre for Climate Research

Keywords

Pacific, Arctic, Decadal, CMIP, Pacemakers

Chat with Presenter

Available March 28th, 1240-1410 MDT / 2040-2210 CET

Join the Discussion

Roles of meridional overturning circulation (MOC) in subpolar Southern Ocean (SO) SST trends: Insights from ensemble simulations

Liping Zhang

Abstract

One of the most puzzling observed features of recent climate has been a multidecadal surface cooling trend over the subpolar Southern Ocean (SO). In this study we use large ensembles of simulations with multiple climate models to study the role of the SO meridional overturning circulation (MOC) in these sea surface temperature (SST) trends. We find that multiple competing processes play prominent roles, consistent with multiple mechanisms proposed in the literature for the observed cooling. Early in the simulations (20th century and early 21st century) internal variability of the MOC can have a large impact, in part due to substantial simulated multidecadal variability of the MOC. Ensemble members with initially strong convection (and related surface warming due to convective mixing of subsurface warmth to the surface) tend to subsequently cool at the surface as convection associated with internal variability weakens. A second process occurs in the late 20th and 21st centuries, as weakening of oceanic convection associated with global warming and high latitude freshening can contribute to the surface cooling trend by suppressing convection and associated vertical mixing of subsurface heat. As the simulations progress, the multidecadal SO variability is suppressed due to forced changes in the mean state and increased oceanic stratification. As a third process, the shallower mixed layers can then rapidly warm due to increasing forcing from greenhouse gas warming. Also, during this period the ensemble spread of SO SST trend partly arises from the spread of the wind-driven Deacon cell strength. Thus, different processes could conceivably have led to the observed cooling trend, consistent with the range of possibilities presented in the literature. To better understand the causes of the observed trend it is important to better understand the characteristics of internal low-frequency variability in the SO and the response of that variability to global warming.

Presented by

Liping Zhang <Liping.Zhang@noaa.gov>

Institution

UCAR and GFDL/NOAA

Keywords

Southern Ocean, SST trend

Chat with Presenter

Available March 28 2:40:3:10pm EST

Join the Discussion

Initialized and Uninitialized ENSO Predictability in Year 2+

Nathan Lenssen, Lisa Goddard, Simon Mason, Yochanan Kushnir

Abstract

While studies have shown robust skill in predicting ENSO and its resulting impacts on precipitation at leads extending into the second year, there are still many questions as to the initial conditions and forecast system needed to achieve such second year skill. In this study, initialized predictions from the CMIP6 decadal prediction project and uninitialized predictions using an analogue prediction method similar to (Ding et al. 2018, 2019, 2020) are compared to assess the role of model biases in climatology and variability on long-lead ENSO predictability. Comparable probabilistic skill is found in the first year between the uninitialized analogs and the initialized dynamical forecasts, but only the initialized forecasts show skill past 12 months. The presence of skill in the initialized dynamical forecasts in spite of large initialization shocks suggest that initialization of the subsurface ocean may be a key component of multi-year ENSO skill.

Presented by

Nathan Lenssen <lenssen@iri.columbia.edu>

Institution

International Research Institute for Climate and Society, Columbia University

Keywords

Chat with Presenter Join the Discussion

Predicting decadal monsoon precipitation variability

Paul-Arthur Monerie

Abstract

Decadal changes in monsoon precipitation have strong effects on the economy, agriculture, and human health. Therefore, understanding and predicting multidecadal changes in monsoon precipitation is of relevant societal importance. However, the ability of climate models to predict precipitation over land, on decadal timescales (2-5 year and 2–9-year lead time), is still not well known. Therefore, we assess the ability of the CMIP6 climate models to predict land monsoon precipitation, for all monsoon domains (over South and North America, southern and northern Africa, South and East Asia, and Australia/Indonesia) and in a global monsoon framework. We explore how a higher skill (Anomaly coefficient correlation) could be achieved by combining model’s outputs and observations, and we assess source of skill at predicting monsoon precipitation (e.g. external forcing, sea surface temperature variability).

Presented by

Paul-Arthur Monerie <p.monerie@reading.ac.uk>

Institution

University of Reading/NCAS

Keywords

Chat with Presenter

Available March 28; 13:00-14:00 MDT (UTC-6)

Join the Discussion

Comparison of Signal-based Decadal Predictions with CMIP5 Predictability

Rishi Sahastrabuddhe, Subimal Ghosh

Abstract

Decadal predictions are near-perfect real-world future realizations in order to assess climate variability and upcoming trends. However, owing to the predictability limits of any geophysical processes, these models fail to capture skillful prediction beyond a certain point of initialization. We show that the limit of predictability to the observed sea surface temperature monthly time series is 12 months maximum. We hypothesize that a signal-based approach, Multi-variate Singular Spectrum Analysis (MSSA), can forecast predictive signals from interannual variability, imitating skill of global climate models (GCM) projections of Coupled Model Intercomparison Project Phase 5 (CMIP5). The analysis shows that MSSA forecasts are quite competitive to the capability of 8 selected GCMs while evaluating Anomaly Correlation Coefficient (ACC) skill scores. Moreover, decadal forecasts from a data-driven regression model using MSSA forecasts for seasonal averages March-May and June-August SST can fulfill skillful multi-year predictions of Indian Summer Monsoon Rainfall.

Presented by

Rishi Sahastrabuddhe <rishi.sahastrabuddhe01@gmail.com>

Institution

IIT Bombay

Keywords

Limits of Predictability, Decadal Prediction, Physics guided data-driven models, Sea surface temperature, Indian Monsoon

Chat with Presenter

Available March 28, 12:40 - 14:10 MDT

Join the Discussion Watch Presentation

Predictions of multiple ocean stressors in CESM-SMYLE in Coral Reef Regions

Samuel Mogen, Nicole S Lovenduski

Abstract

Marine resource planning has the potential to benefit from near-term predictions of ocean physical and biogeochemical state in important ecosystems. Here, we use the novel Community Earth System Model (CESM) Seasonal-to-Multiyear Large Ensemble (SMYLE) to quantify multi-month predictive skill of surface ocean aragonite saturation state, a variable related to ocean acidification, over 1985-2018 in regions with important coral reefs. We find that CESM-SMYLE shows high predictive skill multiple seasons in advance across various regions for physical variables, but that this skill is lower for biogeochemical tracers. Here, predictive skill for sea surface temperatures surpasses skill from persistence and uninitialized forecasts. We argue that improved ecosystem and biogeochemical models can help better predict carbonate chemistry in the near-term in important regions.

Presented by

Samuel Mogen <samuel.mogen@colorado.edu>

Institution

University of Colorado

Keywords

Ocean acidification, initialized models

Chat with Presenter

Available March 28th, 12:30-2:00 PM

Join the Discussion

A hybrid dynamical method for seasonal prediction of sea-level anomalies with potential applicability for multi-year prediction

Thomas Frederikse (1,2), Tong Lee (1), Ou Wang (1), Ben Kirtman (3), Emily Becker (3), Ben Hamlington (1), John Krasting (4), Daniel Limonadi (1), Kristopher Rand (4), Duane Waliser (1)

Abstract

Predictions on seasonal and longer time scales using coupled ocean-atmosphere models are typically initialized from an ocean state obtained from ocean data assimilation. Coupled model biases and initialization drifts due to the incompatibility of the initial ocean state with coupled model dynamics can negatively impact the prediction skill. Here we present a hybrid dynamical prediction method aiming to alleviate these issues. As a first step, we applied the method for seasonal (1-12 month) prediction of sea-level anomalies (SLA) in Charleston at the US east coast and San Diego at the US west coast. We use adjoint sensitivity maps, which show the sensitivity of sea level at our target location to atmospheric forcings (wind stress and heat/freshwater fluxes) at lead times up to one year. By convolving these sensitivities with actual forcings we can reconstruct sea level. The sensitivities are pre-computed by the adjoint model of the Estimating the Circulation and Climate of the Ocean (ECCO) state estimation system. The forcings are a concatenation of (1) ECCO forcing estimates before each prediction initialization time and (2) ensemble forcing predictions from coupled seasonal prediction systems (CCSM4 and GFDL SPEAR used thus far) after the prediction initialization time. We will present results showing the encouraging skill of our hybrid dynamical predictions of SLA. This approach has the unique advantage of attributing prediction skill to particular forcings, and can provide useful feedbacks to the coupled model centers. We also performed SLA predictions only using ECCO forcing before the initialization: these predictions show the “dynamical persistence”: the predictability due to oceanic memory of past forcing. For both San Diego and Charleston, this delayed response to oceanic forcing explains a substantial fraction of the observed variability. This dynamical persistence has a better prediction skill than damped persistence and can serve as an additional benchmark for the assessment of coupled model prediction skills. Because certain regions of the ocean can be influenced by slow oceanic adjustment processes (e.g., open-ocean Rossby waves and oceanic advection) on multi-year time scales, our hybrid prediction method and the resultant dynamical persistence may have potential values for multi-year prediction efforts, not only for SLA, but for other variables such as sea surface temperature and salinity as well.

Presented by

Thomas Frederikse <thomas.frederikse@jpl.nasa.gov>

Institution

1 Jet Propulsion Laboratory, California Institute of Technology, 2 University of California Los Angeles, 3 University of Miami, 4 NOAA Geophysical Fluid Dynamics Laboratory

Keywords

Sea level, persistence, model, ocean

Chat with Presenter

Available March 28, 2022 @ 12:40-14:10 MDT (UTC-6)

Join the Discussion

High Resolution Satellite Study to Characterize Physical & Dynamical Processes in Coupled Climate System & Correlation of Ocean-Atmosphere- Cryosphere Interactions with Climate Variability to Develop Ocean-Atmosphere- Cryosphere Prediction Models (OACPM)

Prof. (Dr.) Virendra Goswami, Noida (Delhi), India.

Abstract

There are places on Earth that are so cold that water is frozen solid. These areas of snow or ice compose the cryosphere. To Characterize physical and dynamical processes in the coupled climate system & Correlation of Ocean-Atmosphere- Cryosphere interactions with Climate Variability & to Develop Ocean-Atmosphere- Cryosphere Prediction Models (OACPM); the understanding of impacts of multiple stressors on the ocean and the associated risks of abrupt state shifts are to be explored through the comprehensive studies of Ocean Systems Interactions, Risks, Instabilities and Synergies (OSIRIS) in order to develop Ocean-Atmosphere Climate Predicting Models (OACPM), over the (Arctic/Antarctica) regions. It’s imperative to investigate submesoscale dynamics of Arctic/Antarctic ice sheet stability, ice and bedrock coring, ice sheet modeling, and ice sheet processes viz. physical, chemical, and biological oceanographic for climate modeling study through the computation of Correlation of Cryosphere Ice Sheet Stability with Sea-level Variability Mechanism , Sub-Mesoscale Dynamics and Climate variability by developing Ocean Systems Interactions, Risks, Instabilities and Synergies(OSIRIS), and Ocean-Atmosphere Climate Predicting Models (OACPM), over the regions. The kinematic features of the mesoscale convective systems over (Arctic/Antarctic) Ocean-Atmosphere regions would be correlated with ocean-atmosphere-cryosphere variability on time and Space Scales; at the local, regional, and global levels through the extracted Sea Surface Temperature (SSTs) over the grid box, attributing the regional change to natural and anthropogenic radiative forcing agents, and to bring out a few optimum values Instabilities and Synergies to develop Ocean-Atmosphere Climate Predicting Models (OACPM), by using Satellite imageries, HPC and cloud computing for real-time analysis.

Presented by

Virendra Goswami <vk_goswami1@rediffmail.com>

Institution

IIT & Environment and Peace Foundation

Keywords

Join the Discussion

Decadal predictability of the North Atlantic eddy-driven jet in winter within CMIP6

Andrea Marcheggiani, Jon Robson, Paul-Arthur Monerie, Thomas Bricegirdle, Doug Smith

Abstract

Recently it has been shown that initialised climate predictions capture the decadal variability of the winter NAO with high skill. However, the signal from models is often hidden among their large internal variability, which results in a low signal-to-noise ratio. In this study, we quantify the skill of the North Atlantic eddy-driven jet’s location and intensity, both in summer and winter. We focus on multi-model decadal predictions made for CMIP6. Overall, we find that models feature a higher skill (as featured by the Anomaly Correlation Coefficient) in predicting the intensity of the jet than its location. For years 2-9, the high winter NAO skill is largely associated with skilful prediction of the jet speed. However, skill in summer is considerably worse than in winter, with models consistently failing to capture the observed southward shift of the Jet between the 1970s and 2010s. Finally, we also show that the skill for the winter NAO is sensitive to the period over which it is computed, and skill drops considerably when evaluating up to the present day, as models fail to capture the observed northern shift and strengthening of the winter eddy-driven jet over the period 2005-2020, as well as the positive trend in the winter NAO.

Presented by

Andrea Marcheggiani <andrea.marcheggiani@reading.ac.uk>

Institution

National Centre for Atmospheric Science, Department of Meteorology, University of Reading

Keywords

Chat with Presenter

Available March 28th 12:40-14:00 MDT (19:40-21:00 BST)

Join the Discussion

▼ Day 2: Land/Atmosphere Processes/Predictability, and Applications; Numerical and Empirical Modeling Techniques; Current and Planned Multi-Year Prediction Capabilities Back to top

The Nonlinear Climate Response to CO2 and Aerosol Forcing for Decadal Climate Simulations

Abdullah A. Fahad1,3; Andrea Molod1; Dimitris Menemenlis2

Abstract

To predict the future climate on seasonal to decadal timescales, it is crucial to understand how the changing CO2 and aerosol concentration in the atmosphere drive the climate and impact the skill of inter-seasonal to inter-annual variability prediction. In this study, we used a 1-degree configuration of the GEOS-MITgcm coupled ocean-atmosphere general circulation model to understand how an equilibrium climate simulation forced with constant CO2 and aerosol levels differ from transient climate model simulations with varying CO2 and aerosol levels. For the equilibrium climate simulations, we ran perpetual 1992, perpetual 2000, and perpetual 2020 experiments with 30 ensemble members for each, forced with their respective year’s CO2 and aerosol concentration levels. To simulate the transient climate simulations, we ran two sets of experiments (5 ensembles each) where one set of ensembles starts with warm initial conditions and the second set starts with cold initial conditions. Examination of the 150 hPa -10 hPa and 55N – 90N averaged DJF temperature from MERRA-2, ERA5, and CFSR (Figure 1a) reveals a positive trend during 1992-2000, and a negative trend during 2000-2020, despite the general expectation that the stratosphere cools as the sea surface and troposphere warm. We found this opposite trend is also present in the GEOS-MITgcm coupled model simulation.

Presented by

Abdullah A. Fahad, <a.fahad@nasa.gov>

Institution

1NASA, GMAO, Goddard Space Flight Center, MD, US; 2Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, US; 3ESSIC, University of Maryland, College Park, MD, US

Keywords

Chat with Presenter

Available March 29, 2022 @ 12:25-13:55 MDT (UTC-6)

Join the Discussion Watch Presentation

Multi-year prediction of drought and heat stress for decision making in the agricultural sector

Balakrishnan Solaraju-Murali, Nube González-Reviriego, Louis-Philippe Caron, Andrej Ceglar, Andrea Toreti, Matteo Zampieri, Pierre-Antoine Bretonnière, Margarita Samsó Cabré and Francisco J. Doblas-Reyes

Abstract

Drought and heat stress negatively impact crop production and food security globally and the impact of these climate hazards is expected to increase over the upcoming decades due to anthropogenic climate change. Therefore, there is a growing need for effective planning and adaptive actions to reduce the impact and the amount of related losses incurred at all timescales relevant to the stakeholders in the agricultural sector. This work aims at assessing the skill in predicting the evolution of drought and heat stress by using user-relevant agro-climatic indices such as SPEI (Standardized Potential Evapotranspiration Index) and HSI (Heat Stress Index) on a multi-annual timescale, as this time horizon coincides with the strategic planning of many stakeholders. In particular, we present the forecast skill of initialized decadal hindcasts to predict these indices for the months preceding the harvest of wheat on a global spatial scale. This study also illustrates the added-value of initialized decadal climate information compared to standard non-initialized climate simulations and demonstrates the potential applicability of decadal forecasts for enhancing the adaptation and mitigation activities for wheat producing regions.

Presented by

Balakrishnan Solaraju-Murali <balakrishnan.solaraju@bsc.es>

Institution

Barcelona Supercomputing Center (BSC), European Commission, Joint Research Centre (JRC), Institució Catalana de Recerca i Estudis Avançats (ICREA)

Keywords

decadal climate forecast, agriculture, climate change adaptation

Chat with Presenter Join the Discussion

Multi-year climate data and modeling needs across the power sector

Erik Smith and Delavane Diaz

Abstract

The weather has a major impact on the power sector. Extreme temperature events can lead to high energy demand and storms can damage infrastructure leading to outages. Transitioning to a decarbonized society means more energy will be supplied by renewables and the power system will be even more reliant on the weather. For this reason, it is critical to determine the ability of wind and solar to meet decarbonization goals and to quantify the inter-annual variability of wind and solar could which present challenges in meeting these demands. An improved understanding of both extreme weather and wind and solar variability will be an important component of near-term system strategy and operational planning as they can have significant impacts on reliable delivery of electricity as well as capital investment decisions. More research is needed to bridge the gap in the near-term (1-10 years) from historical data to climate projections. The power system would benefit from a common framework for tackling this research gap.

Presented by

Erik Smith <esmith@epri.com>

Institution

Electric Power Research Institute

Keywords

Power Sector, Wind, Solar, Extremes

Chat with Presenter

Available March 29, 12:25 pm - 1:55 pm MT.

Join the Discussion

Incorporating Multi-Year Temperature Predictions into Streamflow Forecasts and Operational Reservoir Projections in the Colorado River Basin

Erin Towler (1), David Woodson (2), Sarah Baker (3), Ming Ge (1), James Prairie (3), Balaji Rajagopalan (2), Seth Shanahan (4), & Rebecca Smith (3)

Abstract

Skillful multi-year temperature predictions (up to five years out) offer a potential opportunity for water managers, especially in the Colorado River Basin (CRB), where streamflows are sensitive to temperature. The purpose of this study is to demonstrate how multi-year temperature predictions can be incorporated into streamflow forecasting as well as into operational projections. The framework has three steps. First, 5-year average temperature predictions are obtained from two large ensemble climate model datasets. Second, hindcasts from the Ensemble Streamflow Predictions (ESP), an operationally used forecast method in the CRB, are post-processed using the 5-year average temperature predictions; specifically, a tercile-based block bootstrap resampling approach generates weighted streamflow ensembles called WeighESP. Third, ESP and WeighESP are run through an operational model, the Colorado River Mid-term Modeling System (CRMMS). Compared to ESP, WeighESP marginally improves streamflow forecast accuracy in the multi-year hindcasts up to five years out (i.e., years 1-5, 2-5, 2-4, and 2-3). The multi-year hindcasts show median annual root mean square error (RMSE) improvements between 437,000 and 771,000 cubic meters (354 and 625 thousand acre-feet). Improvements in streamflow accuracy are more pronounced for the more recent hindcast run dates through 2016, partially due to ESP being run with climate time series from 1981-2010. Next, CRMMS translates the streamflow forecasts into operational projections of end of calendar year (EOCY) pool elevations. WeighESP improves the accuracy of EOCY predictions, but mainly for longer leads of 3- and 4-years. For the 4-year lead, the median RMSE improves by 1.1 and 0.7 meters (3.5 and 2.3 feet) for Lakes Powell and Mead, respectively. Although marginal improvements in pool elevations could be beneficial, not being realized until longer leads is a limitation. The study shows the need for better prediction skill at the mid-term timescale and underscores the importance of evaluating improvements in streamflow forecasts in decision-relevant terms.

Presented by

Erin Towler

Institution

1. National Center for Atmospheric Research, Boulder, CO, US; 2. University of Colorado, Boulder, CO, US; 3. U.S. Bureau of Reclamation, Boulder, CO, US; 4. Southern Nevada Water Authority, Las Vegas, NV, US

Keywords

streamflow, reservoir, water management, temperature predictions, large-ensemble datasets

Chat with Presenter Join the Discussion

Fair assessment of climate forecasts

J. Risbey, D. Squire, A. Black, T. DelSJ. Risbey, D. Squire, A. Black, T. DelSole, C. Lepore, R. Matear, D. Monselesan, D. Richardson, A. Schepen, M. Tippett, C. Tozer

Abstract

The assessment of climate forecast skill relies on large hindcasts in order to obtain large samples. Assessments of hindcast skill can be unfairly advantaged over actual forecast skill if they use information in the forecast period that would not be available to an actual forecast. This is the case when the model forecast climatology (used to calculate anomalies or bias corrections) is calculated over periods that include the forecasts tested.

Presented by

James Risbey <james.risbey@csiro.au>

Institution

CSIRO

Keywords

climate forecast skill

Join the Discussion

Multi-year variations of ENSO skill in the past century using model-analog technique

Jiale Lou; Mathew Newman; Andrew Hoell

Abstract

Seasonal-to-interannual hindcasts using state-of-the-art global circulation models (GCMs) are accessible through different operational climate prediction centers, but most of the hindcasts only date back to two to three decades ago, which is insufficient to understand the long-term forecast skill variations. Here, we applied a machine learning approach, called a model-analog (MA) technique, to make hindcasts for the whole twentieth century to evaluate the long-term variations of forecast skills in the tropical climate system with specific focuses on El Niño-Southern Oscillation (ENSO). We made use of the already-existing long simulations of 9 models from sixth phase of the Coupled Model Intercomparison Project (CMIP6) to determine the model states that are close to the observations. The subsequent evolution of those selected MAs then forms a forecast ensemble without additional model integration. In this study, monthly tropical sea surface temperature (SST) anomalies are used to determine MAs at each calendar month. The monthly-initialized hindcasts are then made for forecast lead times of up to 24 months during the 20th century. Different deterministic and probabilistic scores have been used to verify the evolution of predictive skill of ENSO throughout the twentieth century. Our results show that the MAs selected from CMIP6 models can well represent the observations. However, a pronounced ENSO westward extension is readily apparent in those CMIP6 simulations. The predictive skill derived from MA technique is comparable to those state-of-the-art GCMs, and has even higher skill over the long time leads. We show that the predictive skill undergoes decadal variations throughout the whole period with the highest skill in the late-twentieth and lowest in the mid-twentieth century. This MA technique provides a benchmark to investigate the forecast skill without further optimizations of sophisticated GCMs and extra computational burden.

Presented by

Jiale Lou

Institution

CIRES University of Colorado Boulder, and Physical Sciences Laboratory, NOAA

Keywords

Model-analog, machine learning, ENSO prediction, predictability

Chat with Presenter

Available 12:00-14:00 on Mar 29 and in-person during the workshop

Join the Discussion

Prediction skill assessment of large-scale variability influence in summer southeastern South America rainfall in multi-model CMIP decadal predictions

Díaz, Leandro Baltasar; Vera, Carolina Susana; Saurral, Ramiro Ignacio; Doblas-Reyes, Francisco

Abstract

The leading co-variability mode (SVD1) between summer southeastern South America (SESA) rainfall and tropical sea surface temperatures (SST) anomalies over the 1962-2013 period exhibits significant variability ranging from the interannual scale to long-term trends. It shows a clear global warming signal, mainly related to warming in the Pacific and Indian Oceans, in association with a rainfall increase in SESA. After detrending the series, the spatial distribution of both SST and SESA precipitation anomalies associated with the first mode resembles that typically related with El Niño-Southern Oscillation (ENSO). The objective of this work is to assess the prediction skill of SVD1 in decadal hindcast simulations from the Coupled Model Intercomparison Project Phase 5 (CMIP5) of the World Climate Research Programme, using different methodologies to deal with the multi-member/model ensemble. Four methodologies to perform SVD1 will be presented that made use of multi-member/model ensemble information for prediction skill: i) SVD1 computed for the multi-model mean anomalies; ii) SVD1 calculated after concatenating all ensemble members; iii) SVD1 computed projecting all ensemble members in the spatial modes obtained using the multi-model mean anomalies in methodology i); iv) SVD1 computed projecting all ensemble members in the spatial modes obtained using the observed anomalies. Methodologies ii), iii) and iv) allow to obtain probabilistic prediction information, so that internal variability uncertainties could be also assessed. The three methodologies were applied to both detrended and undetrended anomalies. It was found that initialized CMIP5 decadal hindcasts are able to represent SVD1 spatial structures with and without considering trends for the different methodologies, improving results from analogous uninitialized simulations. Although detrended SVD1 activity shows skill in the first two prediction years, differences between methodologies will be discussed. These facts represent a promising result for the predictions of rainfall in the SESA region on interannual and longer time scales.

Presented by

Leandro Baltasar Díaz <ldiaz@cima.fcen.uba.ar>

Institution

Centro de Investigaciones del Mar y la Atmósfera (CIMA, CONICET/UBA)

Keywords

Southeastern South America, Decadal prediction

Chat with Presenter

Available Day 2: March 29, 2022 @ 12:25-13:55 MDT (UTC-6)

Join the Discussion Watch Presentation

Potential reemergence of seasonal soil moisture anomalies in North America

Sanjiv Kumar, Matthew Newman, Yan Wang, and Ben Livneh

Abstract

Soil moisture anomalies within the root zone (roughly, soil depths down to ~0.4 m) typically persist only a few months. Consequently, land surface–related climate predictability research has often focused on subseasonal to seasonal time scales. However, in this study of multidecadal in situ datasets and land data assimilation products, we find that root zone soil moisture anomalies can recur several or more seasons after they were initiated, indicating potential interannual predictability. Lead–lag correlations show that this recurrence often happens during one fixed season and also seems related to the greater memory of soil moisture anomalies within the layer beneath the root zone, with memory on the order of several months to over a year. That is, in some seasons, notably spring and summer when the vertical soil water potential gradient reverses sign throughout much of North America, deeper soil moisture anomalies appear to return to the surface, thereby restoring an earlier root zone anomaly that had decayed. We call this process “reemergence,” in analogy with a similar seasonally varying process (with different underlying physics) providing winter-to-winter memory to the extratropical ocean surface layer. Pronounced spatial and seasonal dependence of soil moisture reemergence is found that is frequently, but not always, robust across datasets. Also, some of its aspects appear sensitive to spatial and temporal sampling, especially within the shorter available in situ datasets, and to precipitation variability. Like its namesake, soil moisture reemergence may enhance interannual-to-decadal variability, notably of droughts. Its detailed physics and role within the climate system, however, remain to be understood.

Presented by

Matthew Newman

Institution

University of Colorado/CIRES and NOAA/PSL

Keywords

Chat with Presenter

Available March 29, 12:45-1:15 PM MDT

Join the Discussion