Tuesday, 8 January 2019
Hall 4 (Phoenix Convention Center - West and North Buildings)
Literature supports the notion that rainfall occurring over regions of open ocean are distributed more similarly to rainfall observed at atoll locations within those regions than to rainfall observed at non-atoll locations. This notion is consistent with the differences in potential thermodynamic and orographic effects characteristic of the two categories. Differences in historical trends in the distributions of extreme rainfall events, between atoll and non-atoll sites, can reasonably be expected. This study examines these differences with respect to accumulation periods ranging from 1 to 30 days. Data are drawn from the Pacific Rainfall (PACRAIN) database, a searchable repository of in-situ observations initiated and maintained by the University of Oklahoma and supported by a research grant from the NOAA Climate Program Office/Climate Observation Division. Only stations with daily records are included in the study. Stations lacking complete records for all years in the study period for at least one calendar month are excluded. Extreme rainfall benchmarks are quantified in terms of event rankings relative to the datasets and in terms of probabilities inferred from kernel density estimators. Changes over time are shown via linear regression of benchmark values and by differences in first and last decade statistics. Differences in trends between non-atoll stations are also discussed in the context of host island size and elevation, station elevation, and geographic region.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner