Sun, R., A. Subramanian, A. Miller, M. Mazloff, I. Hoteit and B. Cornuelle, 2018:
A regional coupled ocean–atmosphere modeling framework
(MITgcm–WRF) using ESMF/NUOPC: description and
preliminary results for the Red Sea
Geoscientific Model Development, sub judice.
A new regional coupled ocean–atmosphere model is developed to study air–sea feedbacks. The coupled model is
based on two open-source community model components: (1) MITgcm ocean model; (2) Weather Research and Forecasting
(WRF) atmosphere model. The coupling between these components is performed using ESMF (Earth System Modeling Framework)
and implemented according to National United Operational Prediction Capability (NUOPC) consortium. The regional
coupled model allows affordable simulation where oceanic mixed layer heat and momentum interact with atmospheric boundary
layer dynamics at mesoscale and higher resolution. This can capture the feedbacks which are otherwise not well-resolved
in coarse resolution global coupled models and are absent in regional uncoupled models. To test the regional coupled model,
we focus on a series of heat wave events that occurred on the eastern shore of the Red Sea region in June 2012 using a 30-day
simulation. The results obtained using the coupled model, along with those in forced uncoupled ocean or atmosphere model
simulations, are compared with observational and reanalysis data. All configurations of coupled and uncoupled models have
good skill in modeling variables of interest in the region. The coupled model shows improved skill in temperature and circulation
evaluation metrics. In addition, a scalability test is performed to investigate the parallelization of the coupled model.
The results indicate that the coupled model scales linearly for up to 128 CPUs and sublinearly for more processors. In the
coupled simulation, the ESMF/NUOPC interface also scales well and accounts for less than 10% of the total computational
resources compared with uncoupled models. Hence this newly developed regional model scales efficiently for a large number
of processors and can be applied for high-resolution coupled regional modeling studies.