Millions of people are affected by natural disasters every year. International organisations such as ECMWF and the Copernicus Emergency Management Services are providing free historical and forecast datasets to improve our understanding of the processes leading to natural disasters as well as provide timely information to allow first responders to act quickly and efficiently at the onset of a dangerous event. Beside availability of data, reproducible workflows are important to share lessons learnt and ensure best practice.

In this short course we aim to showcase how ECMWF, as data providers, envisage the use and post-processing of their open datasets.
During the session we will mention pros and cons of different approaches and demo a walkthrough to: 1) retrieve wildfire reanalysis and forecast data using ECMWF web services (including the experimental retrieval system based on the RASDAMAN array database), 2) assemble multi-layer datasets to explore seasonality and spatial characteristics of historical spatial data, 3) compute country - specific danger levels, 4) effectively visualise forecasts to support practitioners with examples from recent events, 5) calculate impacts on population and affected land cover, 6) produce interactive plots and maps to communicate findings.

This is a hand-on course that will employ the use of bash scripts for data retrieval and the statistical programming language R for processing and visualisation. We invite attendees to bring their laptop as there will be plenty of opportunities to experiment with data and follow along the case studies.

To register to this short-course, please fill in this online form: https://goo.gl/forms/9cVAHUWYq83Ji0py1