Details of Award
NERC Reference : NE/M006433/1
Beyond the code: developing software craftsmanship for environmental scientists
Training Grant Award
- Lead Supervisor:
- Professor AM Walker, University of Leeds, School of Earth and Environment
- Grant held at:
- University of Leeds, School of Earth and Environment
- Science Area:
- Freshwater
- Marine
- Terrestrial
- Atmospheric
- Earth
- Overall Classification:
- Atmospheric
- ENRIs:
- Biodiversity
- Environmental Risks and Hazards
- Global Change
- Natural Resource Management
- Pollution and Waste
- Science Topics:
- Software Eng. Methods & Tools
- Software Engineering
- Environmental Informatics
- Science & Eng. using HPC
- High Performance Computing
- Abstract:
- Every day postgraduate and early career researchers (PGRs and ECRs), and their future employers, request training in software development. These skills underpin the realisation of scientific models and the creation of data processing workflows. With the explosion in size and availability of massive data sets, this pressure for training is acute in the environmental sciences. Furthermore, there is a particularly chronic shortage of people with a deep domain specific knowledge and the ability to develop and improve quantitative numerical models on which so much of modern science relies. Increasingly commonly the problem is not that environmental scientists cannot write code. It is instead that they often do not develop the craftsmanship needed to work with others to develop models and data analysis pipelines that are robust, sustainable and reusable. These are the very attributes that that scientific models and methods should exhibit. We propose to develop and deliver a training programme that will overcome this and teach environmental scientists to build, use, validate, and share software well. We will build on the successful 'Software Carpentry' model shown in a series of studies (and by our own experience of running previous courses) to be highly effective. This blended learning approach combines live coding, peer-programming, and demonstrations by instructors with intensive hands-on programming exercises supported by a group of 'helpers' (peers of the learners with additional programming experience - helpers are often recruited from previous cohorts of learners). The first two days of the proposed workshops will cover the core skills needed to undertake best practice for scientific computing. We will weave specific material of relevance to environmental scientists into this core curriculum with a particular focus on: (1) handling, analysing and plotting geospatial data, (2) dealing with and modelling time-series data, and (3) programmatically accessing environmental data from remote repositories and live data sources. This will allow us to teach much of the core material in a manner which can be directly used by environmental scientists, be they seismologists, meteorologists or geographers. The second half of each workshop will focus on the use of these key skills in a more realistic scenario with attendees working together in four small groups to develop exemplar applications to allow the processing and modelling of environmental data. By using ideas from the core curriculum, and working alongside our helpers, we expect attendees to be able to produce a project that will extract data from a remote data resource, process and model this data, and visualise the results as maps or time series. For example, the remote data resource may be earthquake locations and seismological data, the processing may be refining the source mechanism, and visualisation could show the type of earthquakes across a geographical region. Alternatively, Met Office data could be used to identify regions drought in the UK in particular time periods. We will engage with attendees prior to the workshop to help identify groups with similar interests who will work together on a task of mutual interest. As well as reinforcing the key lessons from the workshop, this exercise will allow attendees to practice the soft skills needed for collaborative model development that is common in modern science.
- NERC Reference:
- NE/M006433/1
- Grant Stage:
- Completed
- Scheme:
- Doctoral Training
- Grant Status:
- Closed
- Programme:
- Advanced Training
This training grant award has a total value of £36,196
FDAB - Financial Details (Award breakdown by headings)
| Total - Other Costs |
|---|
| £36,196 |
If you need further help, please read the user guide.