MPAS-SW-TL-AD

NOTE: This library is still under active development, any suggestions and comments are greatly appreciated. It has been extensively tested with idealized and real atmosphere (ERA5 500 hPa) simulations.

---

Prerequisites

• Python, numpy, scipy, pandas (Anaconda recommended)
• F90Nml
• netCDF4

This package includes the MPAS-Shallow Water dynamics under the Python-Fortran structure and its tangent linear and adjoint components [1].

---

Installation

Run the following commands to compile the package:

cd src
make

Successful compilation should have a module_sw_mpas.so or module_sw_mpas.*.so generated for future Python imports.

---

Demo Simulation

The example python scripts for the nonlinear forward (run-fwd.py), tangent linear (run-tlm.py), and adjoint (run-adj.py) calculations can be found under the demo directory. If the scripts are to be executed under directories other than the demo, simply modify the two lines at the top of each script

import sys
sys.path.append('../src/')

to the path of the source directory.

---

[1]: Tian, X. (2020). Evolutions of Errors in the Global Multiresolution Model for Prediction Across Scales - Shallow Water (MPAS-SW). Q. J. Royal Meteorol. Soc., https://doi.org/10.1002/qj.3923

[2]: Tian, X. and K. Ide (2021). Hurricane Predictability Analysis with Singular Vectors in the Multiresolution Global Shallow Water Model, J. Atmos. Sci., https://doi.org/10.1175/JAS-D-20-0238.1