| D1 | Report on the generation of the project data cube | Explains the scope and goal of the data cube, provides the specifications (spatio-temporal resolution, gridding, size, chunk size), lists EO and non-EO data, their characteristics and pre-processing applied, describes training datasets and provides data cube access details and licenses. | July 2022 |
| D2 | Report on analytics, validated and trustworthy Deep Learning architectures for seasonal fire hazard forecasting in Europe | Provides state of the art in seasonal fire prediction, analytics and fusion methodology and results on top of data cube, DL modeling using GNNs, scientific validation protocol and data used, explainability insights, class activation maps, perturbation analysis, feature ranking methodologies and results, causal relationships, details and specification of the final qualified and trustworthy prediction model. | December 2022 |
| D3 | Report on prototype system for season fire hazard forecasting in Europe with Explainability tools | Provides system requirements, specifications and architecture, describes automatic data ingestion chains and backend technical approach, web-GIS online system, user experience and customization functionality, approach and implementation of prediction model explainability tools. | March 2023 |
