Identifying 2-meter temperature forecasts of opportunity over North America

Melissa Breeden

CIRES and NOAA Physical Sciences Laboratory

Tuesday, Dec 07, 2021, 2:00 pm

JOIN THE WEBINAR:

Abstract

Skillful subseasonal (weeks 3-8 lead times) 2-meter temperature forecasts over North America are highly desirable, but at present forecasts have generally low skill in operational dynamical models. Given the chaotic nature of the atmosphere, at subseasonal lead times only a subset of forecasts is expected to be skillful, motivating the search for so-called ‘forecasts of opportunity’. Forecasts of opportunity are related to low-frequency phenomena, such as the El Niño-Southern Oscillation, that can impart long-lasting teleconnections to the extratropics. Given seasonal variations in the storm track and low-frequency phenomena, it is likely that forecasts of opportunity differ between seasons, though this has not been extensively investigated. Prior research has demonstrated that during boreal winter, an empirical-dynamical model, namely a linear inverse model (LIM), produces north Pacific 500-hPa geopotential height forecasts comparable to the ECMWF IFS and NCEP CFSv2 forecast models. Moreover, the LIM could identify forecasts of opportunity in both its own forecasts and in the other models.

In this study, we use LIMs consisting of tropical outgoing longwave radiation (OLR) and Northern Hemisphere streamfunction from the Japanese 55-year Reanalysis dataset to examine subseasonal North American 2-meter temperature predictability. We consider how predictability evolves during the three phases of the spring transition of the north Pacific jet (winter, spring, and summer), revealing that LIM forecast skill evolves substantially during the three phases, with a distinct skill minimum during spring. While the LIMs’ 2-meter temperature skill is somewhat low on average (similar to operational dynamical models), we are able to identify, in all three phases, 10% of forecasts that are highly skillful. For these forecasts of opportunity, we find that skill is high in different locations in each phase, with spring skill spatially confined to the west coast. Skill during these forecasts of opportunity indicates a strong contribution from the tropical OLR, consistent with prior studies pointing to tropical convection as a source of extratropical subseasonal skill.