ARTMIP-Early Start Comparison of Atmospheric River Detection Tools:How Many Atmospheric Rivers Hit Northern California's Russian River Watershed?
Abstract
Atmospheric Rivers (ARs) are elongated plumes of enhanced water vapor transport usually associated with midlatitude cyclones, and are a key part of the global hydrologic cycle. Regionally, they provide an important contribution to annual precipitation and can also cause floods. Many different AR detection tools (ARDTs) have been developed and applied to various reanalysis products. However, the relative performance of these ARDTs has not yet been well evaluated. To this end, a community effort, organized by an ad-hoc planning committee, began developing an approach to perform such a comparison, called the Atmospheric River Tracking Method Intercomparison Project (ARTMIP). This presentation represents an early-start analysis that helps set the stage for ARTMIP. In this study, the performance of a diverse set of ARDTs and reanalysis products applied to a single location in northern California with a unique 12-year-long AR observatory is compared. The study quantifies the sensitivity of AR count, duration, and intensity to the choice of ARDT and renalaysis. The ARDTs compared here vary in use of different detection variables, thresholds, geometric shape requirements, Eulerian versus Lagrangian approaches, and the choice of reanalysis dataset. On average, 19±7 ARs were detected to make landfall in this region each year based on ARDTs applied to reanalysis datasets used in their original publication. The average annual count of AR landfalls is 19±4 based on ARDTs applied to the same reanalysis dataset. Applying a single ARDT to three different reanalyses with various spatial and temporal resolutions showed little performance sensitivity to the choice of reanalysis. The annual average AR count varied by a factor of two (10-25 ARs/year) based on the choice of ARDT. However, average duration and maximum intensity of ARs varied by less than 10 percent. Overall, fewer AR events are identified by ARDTs with more rigorous geometric criteria or higher intensity thresholds.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.A23H2938R
- Keywords:
-
- 3305 Climate change and variability;
- ATMOSPHERIC PROCESSESDE: 1620 Climate dynamics;
- GLOBAL CHANGEDE: 4301 Atmospheric;
- NATURAL HAZARDSDE: 4313 Extreme events;
- NATURAL HAZARDS