README for GHCN Global Temperature Anomaly image re-projection

Abstract:
Six python scripts are contained in this project that re-project GHCN
Global Temperature Anomaly data onto a map (Hammer-Aitoff, equidistant cylindrical, 
or Robinson). The source images currently reside at:

Monthly Average Temperature Anom: ftp://ftp.nnvl.noaa.gov/View/ANOM/Images/Color/Monthly/
Annual Average Temperature Anom: ftp://ftp.nnvl.noaa.gov/View/ANOM/Images/Color/Yearly/



Python script listing:

ghcn_monthly_driver.py - Driver routine for monthly temperature anomaly images

ghcn_monthly_map.py – Map routine for monthly temperature anomaly images

ghcn_monthly_colorbar.py – Color bar routine for monthly temperature anomaly images

ghcn_yearly_driver.py - Driver routine for yearly temperature anomaly images

ghcn_yearly_map.py – Map routine for yearly temperature anomaly images

ghcn_yearly_colorbar.py – Color bar routine for yearly temperature anomaly images


Additional scripts in this project:
image_meta.py - A script to create metadata for the image stack
update_images.py – A simple script to update (or also grab) the monthly and yearly 
temperature anomaly source images



Program Usage:

Image production begins by executing ghcn_monthly_driver.py (ghcn_yearly_driver.py), which
will create a monthly (yearly) image of a desired size. The possible image sizes 
(in pixel width) include 620px, 1000px, 4096px (DIY), 1920px (HD, or HDSD). An example 
call to the monthly driver routine:

“path to python”  ghcn_monthly_driver.py “YYYYMM” “image size”

e.g., 
/usr/bin/python  ghcn_monthly_driver.py 201401 620 

The above call would produce an image of reprojected global monthly temperature anomaly 
data from NOAA View (http://www.nnvl.noaa.gov/view/).

Or an example call to the yearly driver routine routine:

“path to python”  ghcn_yearly_driver.py “YYYY” “image size”

e.g., 
/usr/bin/python  ghcn_yearly_driver.py 2013  620

The above call would produce an image of reprojected global annual temperature anomaly 
data from NOAA View (http://www.nnvl.noaa.gov/view/).

Valid “image size” values are:
620 (620px wide image)
1000 (1000px wide image)
DIY (4096px wide image)
HD (1920px wide image in high definition)
HDSD (1920px wide image suitable for either high definition or standard definition

Valid dates range from 188001 - present for monthly images (up to the previous month*)
	*release date tends to be mid to late month (18th of month) for monthly data
Or, 1880 through present (up to the previous year*)
	*release date tends to be in February



Software and Dependencies:

The scripts included here were built for Python 2.7. The following
modules must exist on the system for them to perform correctly:

matplotlib-1.1.1 (or higher)
PIL (python imaging library, 1.1.7 or higher, PIL is the Python Imaging Library)
mpl_toolkits.basemap
numpy
urllib
shapefile
sys
os
subprocess
time
matplotlib.font_manager (in order to utilize the true type font MS Trebuchet, which is not
	included as a "base" font in python*)
	*The usage of PIL and the font manager inside matplotlib in these scripts requires
	the ability to utilize true type fonts. As such for example on a linux box, one would
	need to have installed the freetype-dev packages. Specifically, libfreetype6-dev needs
	to be installed on ubuntu... 

Other files in this project:
colorbar and logo (.png and .eps) files included here are sized for each of the image 
sizes and are used by the various python scripts that produce the reprojected images



Notes on portability:

1) All image making scripts in the project need the path to Trebuchet MS true-type font.
e.g.,

	/usr/local/share/fonts/truetype/msttcorefonts/Trebuchet_MS.ttf

	*This path depends entirely on the location the sysadmin installs “add-on” fonts.

2) The two mapping routines (ghcn_monthly_map.py and ghcn_yearly_map.py) require the path
to the original images, e.g., 

	/GHCN_Global/Images/Monthly/Orig/