ccc-gistemp

HomePage: http://code.google.com/p/ccc-gistemp/

Author: Nick Barnes, David Jones

Download: https://pypi.python.org/packages/source/c/ccc-gistemp/ccc-gistemp-0.6.1.tar.gz

        CLEAR CLIMATE CODE GISTEMP README FOR RELEASE 0.6.1

Nick Barnes, Climate Code Foundation
David Jones, Climate Code Foundation

$Date: 2011-06-01 16:08:26 -0400 (Wed, 01 Jun 2011) $


CONTENTS

  1. Introduction
  2. Dependencies
  3. Installation
  4. Input Data
  5. Running
  6. Results
  7. Regression Testing
  A. References
  B. Document history
  C. Copyright and license

1. INTRODUCTION

This is release 0.6.1 of the Clear Climate Code GISTEMP project
(ccc-gistemp).

Clear Climate Code have reimplemented GISTEMP (the GISS surface
temperature analysis system), to make it clearer.  Work continues
towards making it more clear and more accessible.

ccc-gistemp release 0.6.1 is a release of ccc-gistemp version 0.6.
The purpose of version 0.6 is to make ccc-gistemp more useful by:
  - allowing more flexible choices of input data.  For example: GHCN v3;
    USHCN only; ocean only.
  - allowing a land mask to be used in the step where land and ocean
    data are combined.
  - making use on NumPy array.
  - adding a setup.py

In addition there are various bug fixes and improvements to clarity.

Changes since earlier releases are described in more detail in
release-notes.txt.

URLs for further information:

http://clearclimatecode.org/ Clear Climate Code website and blog.
http://ccc-gistemp.googlecode.com/ ccc-gistemp code repository.


2. DEPENDENCIES

You need Python and a machine that can run it, and a network
connection; there are no explicit operating system or CPU architecture
dependencies so "any" operating system or CPU should be okay.

Python comes in several versions.  We recommend Python 2.6 or Python
2.7, but ccc-gistemp should work on any version of Python from the
2.x branch (since 2.4).  It will not work with Python 3.x (which
python.org calls "shiny new thing").  Support for Python 2.4 (and
to some extent 2.5) remains fragile, and it has caused some problems
in the past.

The code should run on OS X, FreeBSD, Windows, and probably a variety of
other Unix-like operating systems.

A network connection is required to download the input files (which
need only be done once), and to display an optional graph from the
results.  If you use a proxy to access the internet then Python requires
that the "http_proxy" environment variable is set.  The proxy will need
to handle both HTTP and FTP requests (this seems to cause some trouble,
see "INPUT DATA" below for downloading data by hand).

Python may already be installed on your machine (for example, it comes
pre-installed on OS X), it may be possible to install it using your
operating system's package manager; for Windows you can download an
installer from http://www.python.org/download/ .  We recommend you use a
stable production release from the Python 2.x series (Python 3.x will
not work).


3. INSTALLATION

Unpack ccc-gistemp-0.6.1.tar.gz.


4. INPUT DATA

ccc-gistemp uses input data in the subdirectory input/.  This
input data includes large files (a few megabytes to a few dozen
megabytes) of temperature records from GHCN, USHCN, and sea surface
data, and small files of additional temperature records and station
tables from GISS.  ccc-gistemp includes code (tool/preflight.py)
to fetch this data from the originating organisations over the
internet.  It will not download a file if it is already present in
the input/ directory, so if you wish to run ccc-gistemp with updated
input data, you can delete the input/ directory before you start.

Downloading the input data is a common causes of problems.  Maintaining
the part of the code that does this (which has nothing to do with the
core GISTEMP algorithm) is a significant cost.  If the tools
we provide do not seem to download the input data correctly, you can
download the data "by hand" and install it in the input/ directory.  See
doc/input.txt for more details.


5. RUNNING

To run ccc-gistemp:

    python tool/run.py

That command runs steps 0 through 5.  To run only a single step or a shorter
sequence of steps, use the -s argument.  For instance:

    python tool/run.py -s 3         # Runs just step 3
    python tool/r