Python implementation of the Gill (1980) atmospheric model, with weak temperature gradient extension by Bretherton and Sobel (2003).
Adapted for Python by Rick Russotto based on Matlab code written by Adam Sobel and rewritten by Chris Bretherton.
The analytical model of the atmospheric circulation response to tropical heating described by Gill (1980) is a widely used theoretical tool of tropical meteorology. A Matlab version of the Gill model is available on Github (https://github.com/wy2136/GillModel) but no Python version, to our knowledge, has previously been made publicly available online. The version posted here also includes the WTG extension described in Bretherton and Sobel (2003).
Developed using Python 3.6.6. Imports NumPy version 1.15.0, Matplotlib version 2.2.2, and several functions from SciPy version 1.1.0.
Installation: clone this repository, or just download the "gill.py" file and type "import gill" in a Python script or session.
All of the computations and some basic plotting scripts are included in the "Gill.py" module. To run the model, one must first set up the structure of a heat source (or sink) using the "setupGillM_Gaussian()" function, which has arguments specifying the grid spacing and dimensions, half-widths, position and scaling of the heat source, and whether to subtract the zonal mean from the heat source (required for some applications). This function returns a dictionary which is passed to one of two computation functions: "GillComputations()" for the classical Gill model, or "WTG_Computations()" for the version with the weak temperature gradient approximation applied. The computation functions allow the specification of whether or not to run the model in an inviscid case (by setting "nodiss" = 1), and return a dict containing the steady state divergence, vorticity, zonal and meridional winds, and geopotential height perturbations arising from the specified heat source. Several functions are provided to plot results in a similar format to the Bretherton and Sobel (2003) paper.
Since the model is linear, multiple heat sources can be run individually and scaled relative to each other via the D0 parameter, and the responses can be added to obtain the combined response.
Included in the repository is a Jupyter notebook, which tests the Gill computation code by reproducing the figures in Bretherton and Sobel (2003). This notebook runs the example cases discussed in that paper, and saves the test figures in a "plots" directory. The notebook is also included as an execuatable Python script. These test cases provide examples of how to use the code and provide a way to make sure the code still produces correct results if parts of it are changed.
As an example, here is the first figure from the test suite:
Bretherton, C. S., and A. H. Sobel, 2003: The Gill model and the weak temperature gradient approximation. Journal of the Atmospheric Sciences, 60, 451–460, doi:10.1175/1520-0469(2003)060<0451:TGMATW>2.0.CO;2.
Gill, A., 1980: Some simple solutions for heat-induced tropical circulation. Quarterly Journal of the Royal Meteorological Society, 106, 447–462.