Assessment of short-range forecast atmosphere-ocean cross-covariances from the Met Office coupled NWP system
- 1School of Mathematical and Physical Sciences, University of Reading, Reading, United Kingdom (a.s.lawless@reading.ac.uk)
- 2National Centre of Earth Observation, Reading, United KIngdom
- 3Met Office, Exeter, United Kingdom
As part of the design of future coupled forecasting systems, operational centres such as the Met Office are starting to include interactions between the atmosphere and the ocean within the data assimilation system. This requires an improved understanding and representation of the correlations between short-range forecast errors in different variables. To understand the potential benefit of further coupling in the data assimilation scheme it is important to understand the significance of any cross-correlations between atmosphere and ocean short-range forecast errors as well as their temporal and spatial variability. In this work we examine atmosphere-ocean cross-covariances from an ensemble of the Met Office coupled NWP system for December 2019, with particular focus on short-range forecast errors that evolve at lead times up to 6 hours.
We find that significant correlations exist between atmosphere and ocean forecast errors on these timescales, and that these vary diurnally, from day to day, spatially and synoptically. Negative correlations between errors in sea-surface temperature (SST) and 10m wind correlations strengthen as the solar radiation varies from zero at night (local time) to a maximum insolation around midday (local time). In addition, there are significant variations in correlation intensities and structures in response to synoptic-timescale forcing. Significant positive correlations between SST and 10m wind errors appear in the western North Atlantic in early December and are associated with variations in low surface pressures and their associated high wind speeds, that advect cold, dry continental air eastward over the warmer Atlantic ocean. Negative correlations across the Indo-Pacific Warm Pool are instead associated with light wind conditions on these short timescales.
When we consider the spatial extent of cross-correlations, we find that in the Gulf Stream region positive correlations between wind speed and sub-surface ocean temperatures are generally vertically coherent down to a depth of about 100m, consistent with the mixing depth; however, in the tropical Indian and West Pacific oceans, negative correlations break down just below the surface layer. This is likely due to the presence of surface freshwater layers that form from heavy precipitation on the tropical oceans, manifested by the presence of salinity-stratified barrier layers within deeper isothermal layers that can effectively limit turbulent mixing of heat between the ocean surface and the deeper thermocline.
How to cite: Lawless, A., Valdivieso, M., Nichols, N., Lea, D., and Martin, M.: Assessment of short-range forecast atmosphere-ocean cross-covariances from the Met Office coupled NWP system, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-8030, https://doi.org/10.5194/egusphere-egu23-8030, 2023.