| 英文摘要: | The airline industry closely monitors the midlatitude jet stream for short-term planning of flight paths and arrival times. In addition to passenger safety and on-time metrics, this is due to the acute sensitivity of airline profits to fuel cost. US carriers spent US$47 billion on jet fuel in 2011, compared with a total industry operating revenue of US$192 billion. Beyond the timescale of synoptic weather, the El Niño/Southern Oscillation (ENSO), Arctic Oscillation (AO) and other modes of variability modulate the strength and position of the Aleutian low and Pacific high on interannual timescales, which influence the tendency of the exit region of the midlatitude Pacific jet stream to extend, retract and meander poleward and equatorward1, 2, 3. The impact of global aviation on climate change has been studied for decades owing to the radiative forcing of emitted greenhouse gases, contrails and other effects4, 5. The impact of climate variability on air travel, however, has only recently come into focus, primarily in terms of turbulence6, 7. Shifting attention to flight durations, here we show that 88% of the interannual variance in domestic flight times between Hawaii and the continental US is explained by a linear combination of ENSO and the AO. Further, we extend our analysis to CMIP5 model projections to explore potential feedbacks between anthropogenic climate change and air travel. The northeastern subtropical Pacific between Hawaii and the continental US is a major corridor of long-distance commercial air travel. The AirTime statistic (wheels-up to wheels-down) for roughly 250,000 flights between Honolulu (HNL) and Los Angeles (LAX), San Francisco (SFO), and Seattle–Tacoma (SEA) from 1995 to 2013 by four major carriers (United Airlines (UA), American Airlines (AA), Delta Airlines (DL), and Hawaiian Airlines (HA)) are analysed and compared with observed8 daily zonal winds at roughly cruising altitude (300 mb; Fig. 1a). To isolate the signal associated with atmospheric variability (as opposed to systematic changes in traffic, technology or policy), rather than analysing flight times in one direction or the other, the difference between westbound and eastbound flight times (ΔT) of each route is computed. There is substantial seasonal-to-interannual variability (~1 h) in monthly smoothed records of ΔT, which is remarkably consistent across different routes and carriers (Fig. 1b). For example, the ΔT records for the HNL–LAX route exhibit correlations of 0.91 (DL versus HA) to 0.95 (UA versus DL). ΔT records for a given route are also significantly correlated with other routes for the same carrier; the HNL–LAX route is correlated 0.86 with HNL–SFO, and HNL–SFO is correlated 0.65 with HNL–SEA. Moreover, correlations are very high between these three routes and other routes that extend well onto the continent: HNL–LAX is correlated 0.81 with HNL–DEN (Denver), 0.82 with HNL–DFW (Dallas–Fort Worth), 0.75 with HNL–ORD (Chicago O’Hare), and 0.73 with HNL–ATL (Atlanta).
Affiliations
-
Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, USA
- Kristopher B. Karnauskas &
- Jeffrey P. Donnelly
-
Massachusetts Institute of Technology/Woods Hole Oceanographic Institution Joint Program in Oceanography, Cambridge, Massachusetts 02139, USA
-
University of Wisconsin, Madison, Wisconsin 53715, USA
|
