PyRTM: Difference between revisions

This page documents the use of the Python Radiative Transfer Modelling wrappers and their associated programs, originally developed by Philip Schleihauf for the Queen's Applied Sustainability Research Group in 2012. Source code is available on GitHub.

The software was developed on Ubuntu Linux 12.04 and Solaris Unix 9. It might run on any unix-compatible environment (say, Windows with CygWin, etc.), but others have not been tested.

Required software for the wrappers includes:

• SMARTS 2.9.5 installed on the system PATH as smarts295
• SBdart 2.4 installed on the system PATH as sbdart
• Python 2.6 or 2.7. Support for Python 3 is experimental.
• Numpy
• dateutil

The wrappers should be functional with those basic requirements. Further packages to make your life easier are noted in the setup and reference documentation.

Setup

Fortran

Installing RTM Software on Linux

Compilers

• Installing _____

Radiative Transfer Modelling Applications

• Building SMARTS
• Building SBdart

Set Up the Environment

• Adding RTM executables to your PATH
• Install PyYaml

Python

Python 2.7

• Install Python 2.7 or Installing Python 2.7 on the HPCVL

Note that most unix installations come with python pre-installed. You can check if you have it by opening up a terminal and typing `python`. If it's installed you should see something like the following:

phil@ubuntu:~$ python
Python 2.7.3 (default, Aug  1 2012, 05:14:39) 
[GCC 4.6.3] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>

Note the first line reporting the version number. If you have a version lower than 2.6, the system just complains python: command not found, then you'll need to install python 2.7. Includes directions for installation without root privileges, and specific directions for setting up on the HPCVL.

Packages

• Setting up VirtualEnv (optional, but recommended)
• Setting up Numpy or Pylab

Numpy is required for the wrappers to work, and is recommended for computing time-series data on a remote cluster. Installing the Pylab environment will provide Numpy in addition to other useful features and is recommended for desktop use.

• Getting dateutil
• Getting fmm

The python package fmm is required to use the optimizer tools of PyRTM. PyRTM will try to install it for you if you use setuptools (ie. easy_install or pip).

PyRTM

• Installing PyRTM

Use

Overview and Concepts

• What PyRTM can and can't do
• Wrapper models: just super-powered python dictionaries
• Caching and lazy evaluation

Simple Modelling

• The RTM objects: method overview
  • irradiance
  • spectrum
• Model Settings
  • Pollution; example
• Clear-sky global horizontal irradiance modelling with SMARTS
• Plot a cloudy day's global direct spectral model with SBdart
• Iterate over different values of Angstrom's Coefficient and plot the relationship

Optimization

• The Optimizer object
• Optimize for cloud optical depth given an measured global irradiance

Time-Series

• blah blah blah

Reference

• Parameter support, defaults, and units table
• Outputs

Issues

License