Representation of Tropical Cyclone Precipitation in Global Reanalysis Datasets

• Other Authors: Jones, Evan (author), Wing, Allison (professor co-directing thesis), Parfitt, Rhys (professor co-directing thesis), Hart, Robert E. (Robert Edward), 1972- (committee member), Florida State University (degree granting institution), College of Arts and Sciences (degree granting college), Department of Earth, Ocean and Atmospheric Science (degree granting department)
• Format: Master Thesis
• Language: English
• Published: Florida State University 2020
• Subjects: Meteorology; Pacific; Merra; North Atlantic
• Online Access: https://diginole.lib.fsu.edu/islandora/object/fsu%3A776946/datastream/TN/view/Representation%20of%20Tropical%20Cyclone%20Precipitation%20in%20Global%20Reanalysis%20Datasets.jpg

Each year, tropical cyclones (TCs) impact communities around the world by producing rainfall with devastating damage, loss of life and can contribute a non-trivial amount to climatological annual mean rainfall. The representation of TC precipitation in datasets such as reanalyses is thus crucial for both forecasting purposes and climate projections. This study quantifies the spread in TC precipitation across eight different reanalysis datasets: CFSR, ERA-20C, ERA-40, ERA5, ERA- Interim, JRA-55, MERRA-2 and NOAA-20C. TC precipitation is assigned using two TC tracking methods, manual tracking and objective tracking, via a fixed 500-km radius from each TC center. Climatological TC precipitation and TC precipitation fraction assigned using TempestExtremes is generally lower than that assigned using IBTrACS. This is due to the fewer number of TCs tracked in TempestExtremes, position differences between the best-track and reanalyses, as well as the underestimation of reanalysis TC intensity. Both methods capture the same general spatial patterns of TC precipitation and TC precipitation fraction globally. Compared to TRMM, assignment from both methods provides less annual average TC precipitation and TC precipitation fraction over most areas, with the exception of land and some ocean basins. In most basins, with the exception of the West Pacific Ocean, and most reanalyses, with the exception of MERRA-2, the systematic bias toward IBTrACS-derived TC precipitation is greatly minimized by considering TC precipitation normalized by TC density. The relative spread in TC precipitation across reanalyses is larger for TC precipitation derived from TempestExtremes than from IBTrACS, where the East Pacific Ocean has a higher relative spread than the WPAC with both tracking methods. Partitioned by Saffir-Simpson best-track intensity, the largest relative spread across reanalyses in TC precipitation is from TCs of major hurricane strength. A comparison of IBTrACS-derived extratropical phase TC precipitation in the North Atlantic ...