https://github.com/space-physics/msise00
NRL MSISE-00 atmospheric model-- in Python and Matlab
https://github.com/space-physics/msise00
build-on-run msis
Last synced: 6 months ago
JSON representation
NRL MSISE-00 atmospheric model-- in Python and Matlab
- Host: GitHub
- URL: https://github.com/space-physics/msise00
- Owner: space-physics
- License: mit
- Created: 2015-03-27T06:53:11.000Z (over 10 years ago)
- Default Branch: main
- Last Pushed: 2023-12-27T19:22:47.000Z (almost 2 years ago)
- Last Synced: 2024-05-02T03:09:45.762Z (over 1 year ago)
- Topics: build-on-run, msis
- Language: Fortran
- Homepage: https://ccmc.gsfc.nasa.gov/pub/modelweb/atmospheric/msis/
- Size: 4.39 MB
- Stars: 46
- Watchers: 9
- Forks: 21
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
# MSISE-00 in Python and Matlab
[](https://zenodo.org/badge/latestdoi/32971905)
[](https://github.com/space-physics/msise00/actions/workflows/ci.yml)
[](https://github.com/space-physics/msise00/actions/workflows/ci-matlab.yml)
[](http://pepy.tech/project/msise00)
NRL MSISE-00 atmospheric model for Python.
Also accessible via Matlab.
Valid from altitude z = 0..1000 km.
The plot immediately below shows a slice at 200km on a world-wide grid.
The yellow ball represents the sun footprint on Earth.
This plot is from [Matlab](./src/msise00/tests/test_msise00_matlab.m) calling MSISE00:
## Units
* Temperature: degrees Kelvin [K]
* Density: particles per cubic meter [m^-3]
* Mass density: kilograms per cubic meter [kg m^-3]
## Install
This process is used for the Python and [Matlab](#matlab) access to MSIS.
Any Fortran compiler should work.
Fortran compiler may be installed by
* MacOS / Homebrew: `brew install gcc`
* Linux: `apt install gfortran`
* [Windows](https://www.scivision.dev/install-msys2-windows)
Get MSISE00 and install Python package:
```sh
pip install msise00
```
or for the latest development code
```sh
git clone https://github.com/space-physics/msise00
pip install -e msise00
```
optionally check that MSIS is working via:
```sh
pytest msise00
```
This Python module uses our build-on-run technique.
The first time you use this Python module, you will see messages from the Meson build system.
## Examples
MSISE00 can be used from the command line, as an imported module, and even from Matlab.
### import module
```python
import msise00
from datetime import datetime
atmos = msise00.run(time=datetime(2013, 3, 31, 12), altkm=150., glat=65., glon=-148.)
```
atmos is an
[xarray.Dataset](http://xarray.pydata.org/en/stable/generated/xarray.Dataset.html)
containing all the simulation output values.
`atmos` is 4-D: (time, altitude, lat, lon), and indexed like `atmos['N2']`
### Command Line
Write NetCDF4 output (HDF5 compatible) with command line argument `-w filename.nc`.
Simple examples under [Examples/](./Examples)
* Altitude Profile at a single time:
```sh
msise00 -t 2017-08-21T20:48 -c 40 -90
```
* Alt. profile time-range with hourly time step:
```sh
msise00 -t 2017-08-21 2017-08-22 -c 40 -90
```
* Grid: time,lat,lon: generates the plots in the README:
```sh
msise00 -t 2016-09-20 2016-09-21
```
* single time lat/lon can be plotted:
```sh
msise00 -t 2017-01-02T12
```
### Matlab
Matlab use of MSISE00 is demonstrated in
[Examples](./Examples/)
MSISE00 will **automatically compile** "build on run" on first run from Matlab as with Python.
Optionally, verify Matlab is working by from the top `msise00/` directory in Terminal
```sh
buildtool test
```
### Fortran
Those wishing to use MSIS from a plain Fortran program can do from the top-level msise00/ directory:
```sh
cmake -B build
cmake --build build
```
optionally:
```sh
ctest --test-dir build
```
## Reference
* Original fortran [code](https://ccmc.gsfc.nasa.gov/pub/modelweb/atmospheric/msis/)
* 1200+ citations 2002 [paper](http://onlinelibrary.wiley.com/doi/10.1029/2002JA009430/pdf)