A sampling of TC PRIMED products from Typhoon Maria
at 1013 UTC on 9 July 2018 in the western North Pacific, where
a) is GPROF,
b) is GPM DPR precipitation rate,
c) GPM DPR reflectivity,
d) 36.6 GHz, e) 89 GHz, and f) infrared from Himawari-8.
Welcome to the home of TC PRIMED!
Why TC PRIMED?
Tropical cyclone forecasters rely heavily on a multitude of forecast products from various sources
to gain insight into the state of the tropical cyclone and issue forecast guidance.
These forecast products include, but are not limited to
Observations from low-Earth-orbiting satellites, with passive microwave sensors that can peer
through clouds to provide forecasters with a glimpse at the convective and precipitation
structure of tropical cyclones
While these products are invaluable for short-term forecasting applications, climate researchers
would also benefit from having a consistent, long-term collection of the products to conduct
multidecadal studies of tropical cyclones. However, the widely disparate data sources of these
products hinder the development of new tropical cyclone forecast products and new tropical cyclone
research.
What is TC PRIMED?
The NOAA/CSU Tropical Cyclone PRecipitation, Infrared, Microwave,
and Environmental Dataset (TC PRIMED) aims to ameliorate the issue of disparate data
sources by consolidating forecast products from various data sources into an analysis- and artificial intelligence (AI)-ready format. TC PRIMED
centers data around satellite passive microwave observations of tropical cyclones.
The development team has compiled and released the final version of the TC PRIMED files for the Northern Indian Ocean and Western North Pacific storms in the 2024 tropical cyclone season. The files are up on the AWS S3 bucket and the NCEI repository.
Additionally, the team has updated a variable in the environmental files and included changes to two 2023 Northern Indian Ocean storms. For more details, see our "resolved issues" section on the products page.
These updates to TC PRIMED are possible thanks to the Department of Defense Joint Typhoon Warning Center decks publicly available and funding from the United States Department of the Navy Office of Naval Research, which enables continued development and additions of new tropical cyclone seasons.
The development team has compiled and released the final version of the TC PRIMED files for the North Atlantic and eastern North Pacific storms in the 2024 tropical cyclone season. The files are up on the AWS S3 bucket and the NCEI repository. Therefore, the preliminary version for these storms is no longer publicly available.
In addition to the 2024 North Atlantic and eastern North Pacific storms, the National Hurricane Center and Central Pacific Hurricane Center released an updated final best-track information for EP052022 (Hurricane Danielle). TC PRIMED now includes these updates.
These updates to TC PRIMED are possible thanks to the National Hurricane Center and Central Pacific Hurricane Center decks publicly available and funding from the United States Department of the Navy Office of Naval Research, which enables continued development and additions of new tropical cyclone seasons.
The GPM-API package can now obtain, process, and visualize data from the Tropical Cyclone PRecipitation, Infrared, Microwave and Environmental Dataset (TC PRIMED).
GPM-API (https://pypi.org/project/gpm-api/) is a Python interface to download, read, process, and visualize products from the GPM data archive. GPM-API was developed independently of TC PRIMED and has recently included utilities to interface with TC PRIMED data on the Amazon Web Service (AWS) Simple Storage Service (S3) bucket.
The TC PRIMED development team finalized the dataset documentation for v01r01. This documentation outlines changes since v01r00 and provides more detail about the team compiles TC PRIMED and calculates metrics.
Users can find the dataset documentation on NODD, the official NCEI archive, and by following this link to the publications page.
NOAA highlighted TC PRIMED in the 2024 NOAA Science Report. TC PRIMED is one example of how NOAA and partners are working to reduce impacts from hazardous weather through artificial intelligence ready datasets and machine learning.
The development team has updated several storms to capture post-season changes to tropical cyclone best track files from the National Hurricane Center, Central Pacific Hurricane Center, and Joint Typhoon Warning Center.
What are the changes and the impact of these changes?
Different aspects of the best track storm metadata can change. These include the period over which the best track represents a storm, the storm center position and intensity. In these updates, center position changed between 0.1 and 0.8 degrees with a 15 knot maximum change in intensity.
The development team has compiled and released the final version of the TC PRIMED files for the Southern Hemisphere storms in the 2023 tropical cyclone season. The files are up on the AWS S3 bucket and should be up on the NCEI repository soon. Therefore, the preliminary version for the Southern Hemisphere storms in the 2023 tropical cyclone season will no longer be publicly available.
These updates to TC PRIMED are possible thanks to the Joint Typhoon Warning Center for making their Southern Hemisphere decks publicly available and funding from the United States Department of the Navy Office of Naval Research, which enables continued development and additions of new tropical cyclone seasons.
The TC PRIMED team recently realized that a variable in the TC PRIMED files was miscalculated. That variable is the storm_heading in the overpass_storm_metadata group and it involves the conversion of the zonal and meridional components of storm motion to direction.
The team has fixed that variable and have uploaded the fixed TC PRIMED files onto the AWS S3 bucket. The fixed files are currently being uploaded to the NCEI repository.
The Python package facilitates easier access to TC PRIMED data on the AWS S3 bucket by allowing users to query and retrieve data based on the a) tropical cyclone season, basin and/or number, b) TC PRIMED file type (i.e., particular satellite, sensor, or TC PRIMED environmental files), c) dates, and/or d) version type (final or preliminary). Users can look at examples of its application on its GitHub repository (https://github.com/CSU-CIRA/tcprimedapi/).
The TC PRIMED team rolls out minor release v01r01 on the NOAA Open Data Dissemination.
The new release includes the following changes:
Extends the dataset back to and up to , excluding the SH basin
Expands the number of storms from around 2,300 to 3,552
Increases number of overpasses from around 197,000 to 242,613
Adds a preliminary version for the current season (i.e., 2024)
The TC PRIMED team would like to thank the United States Department of the Navy Office of Naval Research for providing the support for this release.
The TC PRIMED team participated in the first NOAA Center for Artificial Intelligence/NESDIS Common Cloud Framework (NCCF) Show and Tell.
The NOAA Center for Artificial Intelligence recently released an initial set of interactive open science Jupyter notebooks that highlight NOAA mission area data and machine learning use cases in the form of a Learning Journey Library. The Learning Journey Library increases the economic value and societal benefit of NOAA data and promotes artificial intelligence proficiency for Earth science machine learning applications. Participants got to interact with these Learning Journeys in the NCCF and receive instruction from subject-matter experts. The tropical cyclone case used the Tropical Cyclone PRecipitation, Infrared, and Environmental Dataset based Learning Journey and specifically focused on using these data to develop an artificial neural network to make a precipitation retrieval algorithm for tropical cyclone scenes.
TC PRIMED highlighted in NESDIS Impacts Briefings for Artificial Intelligence.
NESDIS released a new report called the NESDIS Impacts Briefings. The briefings tell the stories behind NOAA satellites and information and feature work from NOAA satellite missions and environmental data. In the Artificial Intelligence (AI) briefing, NESDIS highlighted the Tropical Cyclone Precipitation, Infrared, Microwave, and Environmental Dataset (TC PRIMED; Razin et al. 2023).
TC PRIMED listed as the NOAA Office of the Chief Information Officer dataset contribution along with the NOAA Center for AI Learning Journey Library to the multi-agency National AI Research Resource (NAIRR) Pilot. The Office of Science and Technology Policy and National Science Foundation led NAIRR effort seeks to democratize the AI research and development through aggregating multi-agency AI resources and efforts.
See Naufal Razin's interview with the American Meteorological Society (AMS) for a profile in the Bulletin of the AMS and short write up corresponding to the TC PRIMED BAMS article in the issue print and digital edition.
The TC PRIMED Learning Journey Jupyter notebooks are now available as part of the NOAA Center for Artificial Intelligence (NCAI) Learning Journey Library.
The NCAI Learning Journey Library seeks to enable internal and external NOAA data users to understand AI-ready datasets and explore machine learning in the context of the NOAA mission space. The TC PRIMED Learning Journey, which includes a set of six Python-based Jupyter Notebooks, walks users through understanding TC PRIMED’s data structure, plotting TC PRIMED data, performing an analysis using TC PRIMED, applying a machine learning approach to TC PRIMED, and preprocessing TC PRIMED data for more complex machine learning approaches.
The American Meteorological Society publishes the TC PRIMED development team's dataset manuscript in the November issue of the Bulletin of the American Meteorological Society!!!
Razin, M. N., C. J. Slocum, J. A. Knaff, P. J. Brown, and M. M. Bell, : Tropical Cyclone Precipitation, Infrared, Microwave, and Environmental Dataset (TC PRIMED). Bull. Amer. Meteor. Soc., 104, E1980–E1998, https://doi.org/10.1175/BAMS-D-21-0052.1.
For the Learning Journey discussed during the demo at the workshop, see our products page.
TC PRIMED for the 1998 to 2021 seasons is now publicly available through the NOAA Open-Data Dissemination (NODD) program on Amazon Web Services (AWS) through the support of the NOAA Center for Artificial Intelligence. For details on how to access the data from the AWS s3 bucket, see our products page. TC PRIMED will also be available through NCEI shortly. Through NCEI, more fine-grain, storm-level search functionality will be available.
For details on the structure of the data on NODD, see our products page.
Thanks to support from the Office of Naval Research, TC PRIMED has been extended to include the 2020 and 2021 seasons. With this extension, the dataset now has an additional 23,000 overpasses of over 200 additional tropical cyclones.
CIRA and STAR finalize the TC PRIMED submission agreement with NCEI. This marks the start of data
ingest for TC PRIMED at NCEI.
The TC PRIMED large-scale, storm-centric environmental has been published in the Journal of Climate.
For details, see the citations section of the TC PRIMED publications
page.
The TC PRIMED manuscripts describing details for the entire dataset and the large-scale,
storm-centric environmental diagnostics have been accepted and posted to the Early Online Release.
In addition, NCEI has minted the TC PRIMED DOI. For details, see the citations section of the TC
PRIMED publications page. Data availability is pending ingest at
NCEI.
Naufal Razin presents on "AI-Readiness of the Tropical Cyclone Precipitation, Infrared, Microwave,
and Environmental Dataset" at the Research
Showcase at the Earth Science Information Partners (ESIP) July 2022 Meeting. The talk
discusses an early evaluation of TC PRIMED against the draft ESIP Checklist to Examine AI-readiness for
Open Environmental Datasets.
TC PRIMED fosters new research to better
understand tropical cyclones and new forecast product development, facilitates increased use of
satellite and diagnostics data, and promotes collaboration between
