Community software for geoscientific analysis and visualization

PyNGL and PyNIO have currently only been released as beta versions. In FY2006, we plan to release our first official version of both Python packages and put forth a significant effort to announce the software to the Python community.

One of our major goals this year is to significantly enhance the file input and output capabilities of NCL and PyNIO to 1) include support for data formats like netCDF-4, HDF5, HDF-EOS5, GRIB2, 2) include data aggregation capabilities, and 3) include better support for large (>2 GB) datasets. Based on responses from the NCL survey and on our research efforts into other formats, we may add more data formats to the list.

As the capacities of computer systems and networks grow to handle larger datasets, research needs to be conducted into the scalability of NCL and PyNGL, especially in the area of analyzing and visualizing large datasets. Work has already been done to speed up single file I/O, but more work needs to be focused especially in the area of processing large amounts of data.

A major effort was completed in FY2004 to add the capability to contour non-uniform grids, which opened the doors for NCL and PyNGL to visualize a whole new suite of grids. Along these same lines, efforts are well underway to generate vectors and streamlines on these same grids, and plans are to provide an initial release this year.

Once the NCL/PyNGL/Vis5D+ survey is complete and the results are in, there will be a concerted effort to analyze the reports and update our project plans for this software.

The NCL, PyNGL, PyNIO, and Vis5D+ projects are a key ingredient in our larger strategy to develop generalized frameworks for geoscientific analysis and visualization. In FY2006, we plan to embrace the open source efforts by making NCL, PyNGL, and PyNIO fully open source. We believe this effort, along with our heavy focus on Python, will open the door for several potential collaborations and projects.

FY2005 Annual Report