A Cooperative Agreement to Analyze variability, Change and Predictability in the Earth System (CATALYST)
Task / Institutional Lead
The “Cooperative Agreement To Analyze variabiLity, change and predictabilitY in the earth SysTem” (CATALYST) performs foundational coordinated research in a team-oriented collaborative effort aimed at advancing a robust understanding of modes of Earth system variability and change. The research team does this by using models, observations, and process studies. In addition, innovative machine learning techniques are developed and applied to advance predictability and quantify connections between synoptic weather systems and precipitation extremes.
CATALYST RESEARCH TASKS
- Using machine learning (ML) and artificial intelligence (AI) methods, identify what types of synoptic weather features produce extreme precipitation, and quantify how those synoptic features are affected by modes of variability and possible future changes.
- Assess predictability of modes of variability.
- Expand limits of predictability by applying machine learning methods.
- Quantify the interactions between the quasi-biennial oscillation and the Madden-Julian Oscillation (MJO).
- Identify interactions between MJO and El Nino-Southern Oscillation (ENSO).
- Examine teleconnections among ENSO, semi-annual oscillation, southern annular mode, and interdecadal Pacific oscillation (IPO).
- Benchmark modes of variability and climate states in the Energy Exascale Earth System Model (E3SM), Community Earth System Model (CESM), and other climate models.
- Quantify the role of external forcing in driving changes in modes of variability in E3SM and CESM using large- and single-forcing ensembles.
- Quantify the couplings between modes of variability in E3SM and CESM and tipping points in the climate system.
- Examine connections between modes of variability and atmospheric rivers.
- Quantify linkages between modes of variability and tropical cyclones.
- Connect modes of variability with precipitation extremes through the development and application of innovative machine learning methodologies.
CATALYST RESEARCH OBJECTIVES
Four interrelated research objectives (ROs) address key science questions that focus on modes of variability:
Research Objective 1 (Team lead: Aixue Hu): RO1 uses a combination of Earth System Models (ESMs) and machine learning (ML) methods to understand modes of variability and their limits of predictability on subseasonal to decadal timescales. RO1 provides research themes that tie together the other three ROs in that RO1 uses state of the art modeling tools and ML methods to quantify the limits of predictability for different modes of variability to address the processes and mechanisms that contribute to the predictability of those modes on different timescales. Understanding the limits of predictability requires knowledge of processes and mechanisms that produce such modes of variability, and thus RO1 is the starting point for research in the other three ROs. All ultimately tie together to provide a comprehensive research plan to advance our fundamental knowledge of modes of variability and change in the Earth system.
Research Objective 2 (Team lead: Brian Medeiros): RO2 investigates interactions between modes of variability and sensitivity of modes of variability to model configuration. Following from RO1 where the limits of predictability of modes of variability are examined, RO2 specifically targets interactions of modes of variability and the fundamental processes that underpin them. This work focuses on the interactions of the Madden-Julian Oscillation (MJO) and El Nino-Southern Oscillation (ENSO), the quasi-biennial oscillation (QBO) and MJO, and ENSO and the southern annular mode (SAM). Modes of variability in the climate system are intricately connected, which may have profound implications for predictability in both present and future climates. In this RO, we target several relationships between modes of variability that span time scales: on the subseasonal scale, the QBO modulates the MJO, on seasonal to interannual scales the MJO and ENSO and SAM interact, and these relationships can be modulated on decadal scales via the Interdecadal Pacific Oscillation (IPO). One tool we propose to use is a configuration of a climate model with highly enhanced spatial resolution in the tropics (TRBELT) to help quantify these interactions.
Research Objective 3 (Team lead: John Fasullo): RO3 is designed to benchmark the simulation of modes of variability, examine their changes in a changing climate, and assess connections to tipping points. This RO uses a range of diagnostic packages to evaluate simulation performance across ROs, quantify responses in modes of variability due to external forcings, and explore mutual connections between modes of variability and tipping points. Following from research in RO1 related to limits of predictability of modes of variability, to RO2 where key processes involved with understanding and predicting modes of variability are explored, RO3 then goes on to specifically evaluate and understand the model representation of modes of variability, their response to external forcing, and connections to tipping points.
Research Objective 4 (Team lead: Christine Shields): RO4 uses high-resolution ESMs, regionally refined models (RRMs), and ML methods to investigate the relationships between high-impact events (e.g., flash droughts and precipitation extremes, atmospheric rivers (ARs), tropical cyclones TCs)), the synoptic systems that produce them, and their interactions with modes of variability. The interconnected research in the first three ROs—RO1 (limits of predictability of modes of variability), RO2 (key processes involved with understanding and predicting modes of variability), and RO3 (evaluating and understanding the representation of modes of variability and tipping points, and connections between them)—leads to RO4, in which predictability and the processes and mechanisms of modes of variability delineated in the first three ROs are brought to bear where high-impact events are related to modes of variability. RO4 focuses on high-resolution simulations and utilizes feature detection and ML techniques across all the proposed tasks. RO4 will build on past work and look beyond diagnostics and metrics, to better understand feedbacks and how extreme weather events interact with specific modes of variability at regional scales, including potential changes for future climates.