Climate Informatics 2022 will host a hybrid hackathon during the conference in May. This year’s hackathon will focus on the topic of drought prediction. A brief description of the hackathon challenge is provided below. We will use ERA5 reanalysis data for this hackathon challenge. All participants will have access to AWS-based JupyterLab environments for development. More information on how to form teams and access the data and computing environment will be available the week before the hackathon. Participants (both in-person and virtual) should register for the conference by 6 May 2022.
Drought Prediction Challenge
Drought can lead to substantial environmental, societal, and economic impacts and can become more severe and prolonged regionally in a changing climate. Being able to accurately predict such events helps to mitigate and prepare. This year’s Climate Informatics Hackathon aims to forecast droughts in South Sudan and Ethiopia. Using the Standardized Precipitation Index (SPI) as a target measure, the challenge is to leverage the spatio-temporal modeling power of machine learning to anticipate those dangerous hazards far in advance to support mitigation efforts.