Sensitivity of NMME Seasonal Predictions to Ocean Eddy Resolving Coupled Models

The research proposed here is based on the hypothesis that the presence of oceanic
mesoscale features, that is fronts and eddies, significant modify local air-sea coupling, which in
turn affects the local representation of the predictable large scale climatic features. For example,
preliminary results from retrospective forecast experiments with a global ocean eddy resolving
coupled prediction system indicate that rainfall forecasts along the west coast of North America
are significantly affected by eddy activity in the Kuroshio region, and comparisons with the
current North American Multi-Model Ensemble (NMME) forecasts indicate that this effect is
absent in the current operational system.

The proposed research leverages current efforts in the NMME project and the newly
developing SubX project to test the above hypothesis. Specifically, we propose to repeat the
NMME retrospective predictions with a version of CCSM4 that utilizes significantly higher
resolution in the ocean (and ice) component model (i.e., 0.1 degree vs. 1 degree) and increased
atmospheric component model resolution (0.5 degree vs. 1 degree). The initialization strategy
follows the approach currently in use for both NMME and SubX; and the preliminary
retrospective forecasts provide proof of concept, and indicate that this approach could be used
for the operational NMME.

In addition to implementing the NMME retrospective forecast protocol, the proposed
work includes a detailed large-scale forecast quality assessment using both deterministic and
probabilistic measures. While the proposed research focuses on one particular model, we will
examine how the inclusion of this model affects the multi-model forecast quality. The analysis of
retrospective forecasts will extend beyond monthly and seasonal means to consider extremes and
the shifts in the weather statistics within seasonal and monthly time-scales (and even week 3-4).
Following on some of the preliminary results presented here, we critically examine how the
resolved oceanic mesoscale features change the air-sea coupling, and the local and remote
forecast evolution and quality.

The experience and expertise of the research team is ideally suited to implement the
proposed research, and the project is relevant to the MAPP CTB program. In particular, the call
for proposals seeks efforts that will examine “improving multi-model prediction system such as
the North American Multi-Model Ensemble by testing and demonstrating the utility of new or
higher resolution models…” The PI is the science lead for both the NMME and SubX projects,
and fully understands the issues and challenges in integrating a new model into the operational
NMME. In terms of “readiness levels,” the project is currently at RL5 (the final level of
development) and upon completion of the project we anticipate being at RL7 (well into the
demonstration phase).