The NOAA Center for Artificial Intelligence (NCAI) is creating a clearinghouse of open-science resources focused on artificial intelligence (AI) and machine learning (ML). This initiative aims to serve both NOAA and the public by utilizing standardized development environments to enhance accessibility and collaboration.
Learning Journey Library and Concept
NCAI’s Learning Journey library offers a collection of curated, community-generated interactive learning materials designed to enhance understanding of environmental data science. These short instructional modules cater to varying user needs and skill levels, aiming to expand AI/ML knowledge and improve access to NOAA data, in line with the key recommendations from NCAI’s 2021 Training Gap Assessment.
At the core of each learning journey are interactive notebooks (e.g., Jupyter and RMarkdown). Each journey consists of a series of notebooks that detail prerequisites and expected learning outcomes. This modular approach enables individuals to engage based on their own experience and knowledge, fostering a more personalized learning experience.
Echofish
This multi-notebook Learning Journey showcases how to process Simrad EK60 (18-710 kHz, split beam) echosounder data in the cloud using echopype. It explores the application of frequency differencing for detecting marine organisms and demonstrates how to convert Zarr file format to GeoJSON for use with GeoPandas.
Access Echofish learning journey via NCAI GitHub offsite link
Rip Currents
This notebook explores binomial (two classes) classification modeling, using an application example of the prediction of hazardous rip currents in San Diego, California, USA.
Access Rip Current learning journey via NCAI GitHub offsite link
Tropical Cyclone PRecipitation, Infrared, Microwave, and Environmental Dataset (TC PRIMED)
This learning journey provides a step-by-step guide for using TC PRIMED. TC PRIMED aims to improve disparate data sources by consolidating forecast products into a single repository. Centered around satellite passive microwave observations of tropical cyclones, TC PRIMED contains over 197,000 overpasses of 2,300 global tropical cyclones from 1998 to 2021. It is an AI-ready dataset tailored for tropical cyclone applications.
Contributed by the Cooperative Institute for Research in the Atmosphere offsite link (CIRA). See also: TC PRIMED project offsite link; how to use these files offsite link; dataset via NOAA Open Data Dissemination offsite link
Access TC PRIMED learning journey via NCAI GitHub offsite link
MagNet: Model the Geomagnetic Field
This learning journey provides an example of how to develop and evaluate a deep learning model for estimating/predicting the disturbance-storm-time (Dst) index, a key space weather storm indicator. It is a derived outcome of the 2020 NOAA prize competition, MagNet: Model the Geomagnetic Field, contributed by the Cooperative Institute for Research in Environmental Sciences offsite link (CIRES).
Access MagNet learning journey via NCAI GitHub offsite link
How to Contribute - Resources
Access Contributor Resources learning journey via NCAI GitHub offsite link