Data assimilation greatly improves the ability of ocean models to produce accurate forecasts and meaningful hindcasts. The marine data assimilation group is responsible for the maintenance and improvement of the data assimilation component of Met Office ocean models.

Additionally, the marine data assimilation group is responsible for the upkeep and development of both the OSTIA system and the GMPE. OSTIA is a SST analysis product that is used for a wide range of applications, including numerical weather prediction and climate monitoring. GMPE is a SST product derived from many international SST analyses, including OSTIA.

Temperature and salinity profiles, SST, SSH, and sea ice concentration observations are all regularly assimilated into Met Office FOAM systems. Other ocean products assimilate a subset of these data types with, for instance, assimilation in OSTIA only using SST and sea ice. Much of the data comes from satellites with  Jason and Envisat programmes providing SSH measurements, SST observations coming from  GHRSST and the  EUMETSAT  OSI SAF  providing sea ice concentration data. A wide variety of surface and profile in-situ measurements, including from the  Argo programme, are also used. All data are quality controlled and bias corrected before assimilation.

At present assimilation within FOAM systems in the deep ocean and OSTIA uses an optimal interpolation type method which is currently being adapted for use with NEMO model configurations for the shelf seas. Work is in progress to update our systems to use the NEMOVAR variational assimilation scheme.

Key aims

• To improve the accuracy of ocean model forecasts by the use of data assimilation. 
• To continually improve our assimilation methods. 
• To continually improve the representation and processing of errors in both models and observations. 
• To provide an accurate analysis of daily SST and sea ice using the OSTIA system. 
• To ensure operational ingestion of ocean data into Met Office FOAM systems. 
• To aid in the construction of data assimilation systems for coupled models and seamless prediction.

Current projects

• Upgrading our assimilation scheme to the NEMOVAR variational method. 
• Implementing assimilation into the NEMO model for the shelf seas. 
• Generating SST and sea ice reanalyses using OSTIA. 
• Including analysis of lake temperature in the OSTIA system. 
• Improving sea ice concentration assimilation in the NEMO/ CICE coupled ocean and sea-ice model. 
• The group is involved in the  MyOcean project, particularly through the SST TAC, but also as part of the global and north-west European shelf MFCs.