4-D Modeller

Lead Research Organisation: University of Bristol
Department Name: Geographical Sciences


The exponential growth in open-source data that is underway has been called the 4th industrial revolution. Data analytics is a huge market growth sector. Most of the data that are relevant to public good are open source. These data offer immense potential for tackling societal challenges and providing resources to end users but they present their own challenges on how to generate timely, actionable information from the vast, and evolving, datasets available. An example of this type of challenge is the wealth of Earth Observation data that are increasingly available through, for example, the EU Copernicus Programme. Terrabytes of data are transmitted daily. To extract useful information on, for example, evolving patterns of precipitation extremes, airborne pollution, ocean navigation hazards in ice infested waters or human and faunal heat stress usually requires specialist knowledge and specialist tools. This type of challenge: solving for multiple processes in four dimensions (4-D), i.e. in space and time, is common to an extremely broad class of problems ranging from Earth and atmospheric processes, epidemiology, disease transmission, human activities such as migration, crime, terrorism, consumption patterns, traffic flow, pollution modelling, infrastructure planning, population dynamics agricultural modelling, as well as non-human biology such as ecosystem dynamics. The aim of this Proof Of Concept proposal is to develop a flexible tool to tackle this class of problem, called 4-D Modeller. The tool will be fully scalable (i.e. able to address problems from local to global scale and datasets from thousands to >10^9 observations) and portable. It will be suitable for a range of computing platforms, will be fully documented, tested and include examples from environmental and epidemiological science. The concept builds on a decade of development and experience in tackling large scale 4-D challenges in geoscience.


10 25 50