To improve the access to clear rapidly-generated biodiversity data at a low cost, B-Cubed is packaging known methods together into standardised workflows. Due to their adaptability, these workflows can be run by anyone for any region of interest and can be updated according to the latest advances in data, methods and models.
D4.1 Documentation on modelled data cubesB-Cubed is creating exemplar repeatable workflows for the creation of data cubes, which enable the faster aggregation of biodiversity data from heterogeneous sources. These will act as templates and building blocks for others to use and adapt to their own needs.
D4.2 Deep learningModern deep-learning techniques have great potential for modelling species distributions and/or habitat distributions. B-Cubed is training and evaluating such models on very large sets of occurrence cubes and associated environmental tensors to discover long-term spatiotemporal dependencies in species distribution models and to investigate the explanatory power of such long-term dependency patterns.
D5.5 Indicators softwareB-Cubed’s automated workflows aim to increase the application of standardised modelling to many more species and locations. The workflows for calculating indicators from biodiversity data cubes are to be provided as functions, merged into R packages.