I am pleased to announce the Earth Observation branch of Thredds Data
Server 4.2. This is a branch of Thredds Data Server 4.2 with
enhancements to support Earth Observation (satellite imagery).
https://github.com/bencaradocdavies/tds42
https://github.com/bencaradocdavies/tds42-lib
https://github.com/bencaradocdavies/ncwms-tds42
In each repo, the default earthobs branch contains the new version; the
master branch tracks the upstream subversion (manually, via git svn).
This layout should make it very easy to obtain diffs, using github or
offline tools.
Continuous integration:
https://cgsrv8.arrc.csiro.au/jenkins/job/tds42-earthobs/
Most recent build:
https://cgsrv1.arrc.csiro.au/swrepo/tds42-earthobs/
Features:
- ncWMS scanline support for NetCDF4 / HDF5, necessary for large data files.
- ncWMS support for large source grids (with more than 2**31-1 points).
* For example, WMS for a 647 GB NetCDF 4 file with a single ubyte
variable on a 461276x1505407 grid.
- Support for false colour WMS layers created from NetCDF variables.
For details, see the README (autogenerated by github):
https://github.com/bencaradocdavies/tds42
Demonstration services:
A couple of TB of test Landsat data from 2005 with 8-bit bands:
http://siss2.anu.edu.au/thredds/catalog/landsat/2005/catalog.html
For example, Laamu Atoll, Maldives:
http://siss2.anu.edu.au/thredds/godiva2/godiva2.html?server=http://siss2.anu.edu.au/thredds/wms/landsat/2005/02/LS7_ETM_OTH_P51_GLS2005_145_059_20050201.bx.nc
- "False Colour (R=7,G=4,B=1)" is a 252 colour reduced-depth false
colour using the original ncWMS colour mapping.
- "False Colour (Red=Band7, Green=Band4, Blue=Band1)" uses the new
full-depth (8-bit per channel) false colour.
Here are a few test Landsat NBAR scenes:
http://siss2.anu.edu.au/thredds/catalog/landsat/nbar/catalog.html
For example, Adelaide, Australia:
http://siss2.anu.edu.au/thredds/godiva2/godiva2.html?server=http://siss2.anu.edu.au/thredds/wms/landsat/nbar/LS5_TM_NBAR_P54_GANBAR04_097_084_20070623.nc
The NBAR products are 16-bit bands. This test configuration shows
several different gamma choices; all are full-depth false colour.
I do not at this time have any resourcing to submit these changes to
trunk, but I am keen to hear any community interest.
Thanks again to Ethan Davis for providing buildTds.xml, and thanks to
Jon Blower for ncWMS!
Kind regards,
--
Ben Caradoc-Davies <Ben.Caradoc-Davies@xxxxxxxx>
Software Engineer
CSIRO Earth Science and Resource Engineering
Australian Resources Research Centre
