Closing the heat budget of the SOC climatology through spatially dependent inverse analysis parameter adjustment. COAPEC Project - Balancing the Atlantic Heat and Freshwater Budgets, Report No. 3

Previous research aimed at closing the heat budget of the Southampton Oceanography Centre (SOC) air-sea flux climatology through the method of inverse analysis is extended to include spatially dependent parameter adjustments and error specification. In earlier analyses, a balanced solution was achieved using globally fixed parameter adjustments, primarily an increase of 19% to the latent heat flux and a reduction of 6% to the shortwave flux. In the new method, the global ocean is divided into various sub-regions in order to allow the parameter adjustments to vary spatially. With this approach a balanced version of the SOC climatology is obtained that requires smaller adjustments, in the range 2-12%, to the latent heat flux than previously but larger changes to the shortwave, up to 18%, depending on region. In addition to enabling direct spatial dependence of the parameter adjustments we have also explored the possibility of making the parameter error spatially dependent both by sub-region and through a dependency on observation density. The various solutions obtained have been evaluated both through the large scale implied ocean heat transport and local comparisons with research buoy measurements. Some improvement is found in the level of agreement of the heat transport with the applied constraints but the buoy comparisons reveal similar problems to those obtained in our previous research. Further, the larger adjustment to the shortwave flux with the new solutions leads to significant differences with respect to satellite based estimates of this component of the flux. We conclude that the earlier solution in which the latent heat flux is increased by 19% is in better agreement with independent estimates than the new spatially dependent solutions. It is thus our preferred means of closing the SOC climatology heat budget imbalance through inverse analysis.