| 10-12 March 2020
Due to the measures associated with the COVID-19 virus, this workshop was held virtually for all external participants.
Warm Conveyor Belts (WCBs) are cloudy regions of ascent and strong diabatic forcing along the cold front of a synoptic depression. They can lead to heavy local precipitation and can have downstream impacts such as the onset and maintenance of blocking. They are therefore important in both weather and climate prediction. However, WCBs are also associated with inherently increased error growth-rates, and are difficult to constrain in analyses - due partly to clouds having a strong non-linear impact on satellite observations and moist processes leading to larger model errors. Indeed, forecast “busts” are often associated with the existence of WCBs. In the climate text, model deficiencies in WCBs and in their large-scale drivers are likely to be particularly relevant; with implications for the future statistics of precipitation, heatwaves and droughts for example. This workshop brought together observation, assimilation, model, forecast and research communities to explore these aspects. The aim was to improve understanding and to help develop optimal strategies to improve weather and climate prediction - a goal which would be very difficult for a single such community to achieve on its own. The workshop included invited and submitted talks, and had a strong focus on posters.
Programme and key questions
1. WCBs and downstream impacts
What do we know about the formation, dynamics and physics of WCBs and their downstream impacts? Is convection and upscale error growth important for the evolution and predictability of WCBs – with implications for the scales we need to represent in the model and constrain in data assimilation? Relationship to Atmospheric Rivers. Conceptual models.
Numerical weather prediction assimilates a wealth of observations; some sensitive to cloud and precipitation through the use of "all sky" methods. What are the key observations which currently constrain WCBs? What are the limits to how well they could constrain the relevant scales and parameters? Do WCBs strengthen the case for additional observations in future? EarthCARE, Aeolus. Learn from NAWDEX and AR campaigns.
3. Models and model uncertainty
How well do model climates represent the dynamics and physics of WCBs? What are the key sensitivities in model formulation and resolution (in the absence of initialisation)? Comparison with observations and reanalyses. Multi-model comparisons. Formulation and impact of model uncertainty.
4. Data assimilation
While WCBs might not highlight useful developments in DA methodology per se, there is a lot that diagnostics of data assimilation can tell us. How well do current assimilation schemes constrain WCBs? Where might the largest achievable improvements be made amongst the prior (background), model (non-linear, tangent linear and model uncertainty) and observational components? Ensemble data assimilation. Adjoint sensitivity. Forecast Sensitivity - Observation Impact (FSOI). Initial process tendencies and analysis increments. Multi-analysis comparisons.
5. Weather forecasting
How well are the dynamical evolution (including downstream impacts) and physical aspects predicted at present? Comparison with observational campaign data. Evaluation of ensemble forecast reliability, refinement and sharpness. Role of model uncertainty. What are the limits and challenges? Multi-model comparisons of ensemble forecasts (including TIGGE).
6. Climate variability and change
From observations/reanalyses what broader-scale features are associated with variations in WCB statistics? How well do models at seasonal/climate resolution represent these links? What can we infer about the statistics of WCBs (and their downstream impacts) in seasonal/climate predictions?
7. Break out groups and plenary
Further consideration of the above questions (in the light of talks and posters) and report back.
Stephen English, Laura Ferranti, Richard Forbes, Christian Grams, David Lavers, Linus Magnusson, Mark Rodwell, Irina Sandu, Heini Wernli