Comment on “Historical trends in the jet streams” by Cristina L. Archer and Ken Caldeira


[1] Archer and Caldeira [2008] studied upper tropospheric wind trends over the Northern Hemisphere and Southern Hemisphere using reanalysis data for the satellite era 1979–2001. The authors presented maps showing regional and seasonal trend variations and concluded that, in general, jet streams have moved poleward and risen to lower pressures in both hemispheres and weakened in the Northern Hemisphere. Previous studies of the boreal winter record also detected poleward jet stream contraction but found that jet stream core speeds increased extensively over the midlatitudes [Strong and Davis, 2007], and showed that the leading pattern of jet stream pressure variability was an annular seesaw pattern linked to the Arctic Oscillation [Strong and Davis, 2006, 2008]. Furthermore, Strong and Davis [2007] reported that positive jet stream wind speed trends are robustly indicated by geopotential height trends in three data sets: the ECWMF record since 1958, the NCEP record since 1958, and the satellite-only portion of the NCEP record (1979–2007).

[2] Differences from Strong and Davis [2006, 2007, 2008] were not mentioned by Archer and Caldeira [2008], and stem largely from fundamentally different definitions for jet stream properties. Archer and Caldeira [2008] used columns of monthly mean data between 400 and 100 hPa to define jet stream properties as “mass and mass-flux weighted averages of wind speed, pressure, and latitude” (see their equations (1)–(3)). Strong and Davis [2006, 2007, 2008] defined jet stream and jet stream core properties using the six-hourly upper tropospheric surface of maximum wind (SMW) [Strong and Davis, 2005] together with a minimum wind speed threshold of 25.7 m/s (50 knots) as specified in the AMS Glossary of Meteorology [Glickman et al., 2000]. A minimum speed threshold is widely recognized as essential to jet stream identification [e.g., Reiter, 1963; Djuric, 1994; Glickman et al., 2000; Koch et al., 2006; Degirmendzic and Wibig, 2007], and Archer and Caldeira [2008] did not apply a minimum speed threshold.

[3] The SMW methods and results of Strong and Davis [2006, 2007, 2008] are consistent with the idea that a jet stream is a core of strong winds instantaneously exceeding a minimum threshold and concentrated within a narrow atmospheric stream whose orientation and position vary from day to day. Defined as such, jet streams are relevant to surface weather [e.g., Riehl et al., 1954], cyclogenesis [e.g., Carlson, 1994], hurricanes [e.g., McTaggart-Cowan et al., 2006], and clear air turbulence [e.g., Reiter and Nania, 1964]. Defined in time-filtered or time-averaged fields, so-called “mean jet streams” have been linked to storm tracks and baroclinic wave development [e.g., Blackmon et al., 1977]. It is less clear how mass weighted averages of monthly mean wind speed in the 400–100 hPa layer [Archer and Caldeira, 2008] relate to surface weather, cyclogenesis, hurricanes, and clear air turbulence.

[4] Quantifying properties of jet streams per se requires separate consideration of jet stream wind-speed-when-present and jet stream relative frequency as demonstrated in the works of Gallego et al. [2005], Koch et al. [2006], Strong and Davis [2006, 2007, 2008], and Degirmendzic and Wibig [2007]. Nonetheless, because Strong and Davis [2006, 2007, 2008] and Archer and Caldeira [2008] used markedly different measures of wind speed variability, their collective findings yield a more thorough description of upper tropospheric flow than either view alone. Most interestingly, the mass weighted monthly mean wind speed between 400–100 hPa generally weakened during boreal winter while the core speed and relative frequency of embedded jet streams per se generally increased. Understanding how such contrasting findings transcend issues of terminology, weighting, and averaging to touch on circulation and atmospheric dynamics may lead to insights concerning the interpretation of historical climate variability.