Spatio-temporal big data, we are drowning, help!

[npie762]

A team of engineers are trying to improve medical diagnoses in Uganda using high spatial and temporal resolution environmental factors and time-series analysis -- they've been working on this for nearly 20yrs. The observations that they want to incorporate number in the 10s of millions originating from thousands of files posted to an ftp site in a variety of file formats, the only person on the team with any coding skills is a collaborator at another institution who writes fortran code and no one is familiar with spatial or high spatial resolution data. They also want to "fuse" these data with other spatial data and do not understand how to combine spatial data of multiple raster spatial resolutions, projections, grid registrations, and other spatial data formats.

There are no manual techniques that would have any prayer of working for their application, therefore, the standard GIS software solutions (with which one team-member is somewhat familiar), are not an option to solve this problem. They have an example of a paper that appeared to do something similar and the methods section says little about how they overcame these big data challenges.

The research group has worked with the Penn State Maps Library in some capacity in the past and acknowledged Maps in a prior publication on their work.

[merama]

Problem: How to perform data analysis