Points of Contact

Co-chairs of sub-project: Amy Butler and Chaim Garfinkel

S2S committee member: Daniela Domeisen 

Overview

The stratosphere is one of the major contributors to sub-seasonal to seasonal predictability of surface climate in winter and spring. Our activity is continuing to lead international efforts to quantify and understand this predictability so that, ultimately, we might improve our models and forecasting systems so they can better exploit it.

Summary of Previous Activities

We have contributed an overview chapter led by Amy Butler to the S2S prediction book “The gap between weather and climate forecasting” on subseasonal predictability and the stratosphere:  https://www.sciencedirect.com/science/article/pii/B9780128117149000115

We have produced two major review papers led by Daniela Domeisen to gather the international community on the topic of predictability of the stratosphere and stratosphere-troposphere coupling on sub-seasonal timescales using the S2S project database. Both studies are published in the JGR special issue on “Bridging Weather and Climate: S2S Prediction”. This was a major milestone for our project.

• Stratospheric predictability: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2019JD030920
• Stratosphere - troposphere coupling: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2019JD030923

Another milestone was a joint meeting held with SPARC DynVar in Madrid that brought together a wide range of scientists working on dynamics and predictability, including but not limited to stratosphere topics. Another joint meeting was held in Munich in October 2023 (Conference webpage).

Ongoing Activities (1 of 2):

Our initial analysis of stratospheric predictability and processes in the S2S prediction systems revealed issues with biases in these models that vary as a function of lead-time and initialization time. A community project is ongoing to evaluate stratosphere-troposphere coupling biases in S2S prediction systems, and how these biases may influence predictive skill.     We currently have researchers from 11 countries contributing to this analysis (Switzerland, Israel, Spain, United States, Finland, South Korea, United Kingdom, Japan, Australia, Norway, Argentina).  Core questions we seek to answer include:

a.      What are the lead time dependent mean biases in the stratosphere and in stratosphere-troposphere coupling processes, and how do they compare among models? What about biases in variability?

b.     Which biases have the greatest impact on predictive skill? How large are the impacts?

c.      What are the sources of the biases? Can they be linked to biases in the troposphere, and/or can biases in the troposphere be linked to those in the stratosphere?

d.     Are biases linked to stratospheric processes/variability? E.g., are biases larger before or after SSWs, vortex intensifications, phases of the QBO, etc?

 These efforts are being led by Zachary Lawrence. The first paper was published in 2022, and can be found at

https://wcd.copernicus.org/articles/3/977/2022/wcd-3-977-2022.html

A second community paper will be submitted in June 2024. A third community paper is planned. If you would like to join these efforts, please contact the SNAP co-chairs (Amy Butler, NOAA and Chaim Garfinkel, Hebrew University).

Ongoing Activities (2 of 2):
Stratospheric Nudging And Predictable Surface Impacts (SNAPSI - experiments for operational centres to perform that isolate the role of the stratosphere on predictive skill)

Recent studies including our review papers have highlighted that stratosphere-troposphere coupling in the extra-tropics contributes to S2S predictability. However, not all SSWs are followed by impactful weather events; furthermore, there is substantial spread in the ability of models to represent this coupling. Similarly, in the tropics, a focus on the links between the QBO and MJO predictability is very important for surface predictability. The Stratospheric Nudging And Predictable Surface Impacts (SNAPSI) project aims to isolate the role of the stratosphere for surface climate and predictability, and also explore the role of stratospheric biases for inter-model spread in the representation of stratosphere-troposphere coupling. Specifically SNAPSI targets three SSW case study events, in which the stratospheric state can be either freely-evolving or nudged towards a climatological or observational state.  The test cases will be NH SSWs during two recent winters with very different tropospheric impact (2018 and 2019) and the 2019 SH warming, which has likely contributed to the extreme wildfires over Australia in 2019/20. The goals of this project are:

a.     To assess the contribution of the stratospheric evolution to forecast skill in a controlled fashion.

b.     To assess the representation of coupling processes across different operational models.

c.      Attribute particular meteorological events to stratospheric conditions

d.     To assess the representation of stratospheric wave driving.

The basic experimental protocol consists of a set of forecast ensembles: (1) a standard, free running forecast ensemble, (2) a 'perfect stratosphere' forecast in which the stratosphere is relaxed towards the observed evolution, and (3) a 'control' forecast in which the stratosphere is relaxed towards climatology. The paper documenting the Experiment Protocol for damping experiments was published in 2022. To date, twelve modeling groups at eleven centers have contributed integrations following this protocol, and most data is archived on CEDA. This allows for an unprecedented, multi-model comparison of the dynamics underlying the surface responses to sudden stratospheric warmings. Moreover, by including 'counterfactual' forecasts in which the stratospheric circulation remains in a climatological state, the experimental protocol allows for formal attribution statements to be made regarding the surface extremes that followed the stratospheric anomalies.

In addition to the four aforementioned goals, two working groups are analyzing stratosphere-troposphere coupling in the tropics in these experiments in collaboration with with QBOi and SATIO-TCS. Each of the six working groups are working on papers documenting the results - these publications will be submitted in Q3 and Q4 of 2024.

Anyone interested in participating in the analysis for the 6 community papers is encouraged to contact Peter Hitchcock (aph28@cornell.edu). The data embargo for the broader community will tentatively be lifted in 2024.