Cyclone Center: A Crowd Sourcing Application of the HURSAT-B1 Data Record

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 7 January 2015
Christopher C. Hennon, Univ. of North Carolina, Asheville, NC; and K. R. Knapp, C. J. Schreck III, J. P. Kossin, and S. E. Stevens

Handout (345.5 kB)

The Hurricane Satellite (HURSAT) data set is a 32-year (1978-2009) record of geostationary satellite data centered on tropical cyclones. The data consists of calibrated observations from the International Satellite Cloud Climatology Project (ISCCP) B1 data set, available in 3-hourly increments at 8 km resolution. The continuity, homogeneity and global scope of HURSAT B1 make the data set a useful tool for investigating decadal- and climatic-scale changes in global tropical cyclones.

One unique project that uses HURSAT B1 data is Cyclone Center. The goal of Cyclone Center is to resolve inconsistencies in the global tropical cyclone intensity record that arise from sparse observational data and divergent analysis methods at respective tropical cyclone centers. Because the HURSAT B1 data contain nearly 300,000 images, Cyclone Center expedites the analysis process by using crowd sourcing. Non-experts from around the world perform "classifications" of HURSAT B1 brightness temperature data by answering simple questions on cyclonecenter.org. Citizen scientist responses are collected, quality controlled, and then used to determine a consistent global tropical cyclone intensity record.

Results based on the first year of Cyclone Center classifications will be presented. We will show that: 1) The HURSAT B1 data set is a useful tool for diagnosing tropical cyclone intensity, despite its relatively coarse horizontal resolution, 2) crowd sourcing by non-experts produces results comparable to an objective computer scheme used on HURSAT B1, and 3) our approach can resolve gross differences in tropical cyclone intensity. Work is underway to adopt a more comprehensive intensity analysis scheme, which we expect will further improve the performance of the approach. Classifications on cyclonecenter.org continue and participation is open to anyone.

Supplementary URL: http://cyclonecewnter.org