Using the Scientific Python ecosystem to advance open radar science

Abstract:

The choice of a programming language or environment is rarely made with consideration of its benefits and disadvantages. Often it is something inherited from mentor or enforced by an institution. Python, developed as a “hobby” programming project, has seen increased migration of users from more traditional domain specific environments. This presentation charts our own journey in using the scientific python ecosystem, first as users and then as the developers of a community based toolkit for working with weather radar data, the Python ARM Radar Toolkit, Py-ART. We will highlight how a data model driven design approach can extend the usefulness and reusability of code and act as a bridge between amorphous mathematical algorithms and domain specific data. Finally we will showcase how Python and Py-ART can be used on clusters to tackle pleasantly parallel problems like deriving climatologies swiftly, painlessly and most importantly: reproducibly.