83rd Annual

Monday, 10 February 2003: 4:30 PM
Storm track variability as seen from aircraft and ship observations
Edmund K. M. Chang, SUNY, Stony Brook, NY
Upper tropospheric eddy variance/covariance statistics are commonly used to indicate the intensity of storm track activity. Recent publications have suggested that variance/covariance statistics are much more sensitive to model biases and changes in observational network and observation quality. In this paper, high-pass eddy variance computed from NCEP/NCAR reanalysis data are compared with similar statistics computed directly from unassimilated aircraft and ship observations to verify the storm track variability seen in the reanalysis data. We need to analyze aircraft and ship observations instead of rawinsonde observations because the storm track peaks are located over the oceans where rawinsonde observations are nearly absent.

First, aircraft observations is used to verify the seasonal cycle of storm track activity in the North Pacific. Our results confirm previous findings that the Pacific storm track activity exhibit a minimum in mid- winter and maxima in fall and spring, and that the mid-winter Pacific storm track is significantly stronger during the early 1990s than during the early 1980s.

Next, aircraft observations is combined with rawinsonde observations from weather ships to confirm recent findings that the Northern Hemisphere winter storm track activity has experienced a significant increasing trend during the second half of the twentieth century. Preliminary results suggest that the unassimilated observations are consistent with a secular increase in storm track intensity, albeit at a slower pace than that suggested by the NCEP/NCAR reanalysis data. Several issues, such as possible changes in data quality and coverage over the years, remain to be tackled.

Currently, we are compiling surface ship reports in an attempt to compute high-pass variance in surface pressure as another indicator of storm track activity. This data set should have experienced less changes in terms of quantity and quality compared with aircraft data, and is thus more suitable for diagnosing climate signals. Results based on analyses of this data set will be discussed at the conference.

Supplementary URL: