SMCM-C

The Coastal version of the Stochastic Multcloud model

This is the python code of the coastal version of the stochastic multi-cloud-model (SMCM-C). A in depth describtion of the model can be found in the scientific journal article Coastal Tropical Convection in a Stochastic Modeling Framework.

The code should be running with both python versions 2 and 3.

Pre-Requisits

You need to have a working python 2 or 3 version installed on your computer to use the model. Additionaly you need to have the following external python modules installed.

• numpy
• netCDF4
• pandas
• pytz
• matplotlib and basemap for visualisation
• timezonefinder
• geocoder
• SALib

SALib and geocoder are only necessary if you want to run a Variance Based Sensitivity Test (SALib) or apply the model to realy world situations.

You can either install the above mentioned additional python packages with the os repository manager (apt, pacman, emerge, port) or via pip.

Running the Model

The Important Modules

The heart of the model contains two Classes:

SMCM (in the model module)	Coarsgraining (in the coarsgraining module)
Defines the transition-rates	Does the coarsgraining and nearest neighbor interaction

The Coarsgraining class inherits from SMCM so all you need to do in order to run the model is to create an instance of the Coarsgraning module. Take also a look at the attached ipython-notebook for details.

The Model Configuration

The model is configured by the text file constants.config. The configuration is saved in the following format:

key = value

Characters starting with '#' are ignored. You can add as many new keys and values as you like. The Config class form the configdir module will automatically detect them and pass the to the SMCM class where they will be made available. The following procedure can be seen as an example.

Suppose you want to add a new variable to the model. The easiest way to achieve that is to add a new entry into the constants.config file:

new_variable = value # A meaningful describtion of what this variable does

New_variable is then available to the SMCM and Coarsgraining class as a new class instance. If you created the Coarsgraining class with:

>>> CG = Coarsgraining('constants.config', 0.1, 0.4)

You can access new_variable via:

>>> CG.new_variable

Running an Example

A short example of how to run the cloud model is available in the test_run ipython-notebook

Running Experiments

The run_cloudmodel module contains a some methods to run various experiments, like a sensetivity study. To make use of all features of the modle it is highly recommended to have mpi4py module installed.