Changes in Jupiter's zonal velocity between 1979 and 2008

We show that the peak velocity of Jupiter's visible-cloud-level zonal winds near 24 degrees N (planetographic) increased from 2000 to 2008. This increase was the only change in the zonal velocity from 2000 to 2008 for latitudes between +/- 70 degrees that was statistically significant and not obviously associated with visible weather. We present the first automated retrieval of fast (similar to 130 m s(-1)) zonal velocities at 8 degrees N planetographic latitude, and show that some previous retrievals incorrectly found slower zonal winds because the eastward drift of the dark projections (associated with 5-mu m hot spots) fooled the retrieval algorithms. We determined the zonal velocity in 2000 from Cassini images from NASA's Planetary Data System using a global method similar to previous longitude-shifting correlation methods used by others, and a new local method based on the longitudinal average of the two-dimensional velocity field. We obtained global velocities from images acquired in May 2008 with the Wide Field Planetary Camera 2 (WFPC2) on the Hubble Space Telescope (HST). Longer-term variability of the zonal winds is based on comparisons with published velocities based on 1979 Voyager 2 and 1995-1998 HST images. Fluctuations in the zonal wind speeds on the order of 10 m s(-1) on timescales ranging from weeks to months were found in the 1979 Voyager 2 and the 1995-1998 HST velocities. In data separated by 10 h, we find that the east-west velocity uncertainty due to longitudinal fluctuations are nearly 10 m s(-1), so velocity fluctuations of 10 m s(-1) may occur on timescales that are even smaller than 10 h. Fluctuations across such a wide range of timescales limit the accuracy of zonal wind measurements. The concept of an average zonal velocity may be ill-posed, and defining a temporal mean zonal velocity as the average of several zonal velocity fields spanning months or years may not be physically meaningful. At 8 degrees N, we use our global method to find peak zonal velocities of similar to 110 m s(-1) in 2000 and similar to 130 m s(-1) in 2008. Zonal velocities from 2000 Cassini data produced by our local and global methods agree everywhere, except in the vicinity of 8 degrees N. There, the local algorithm shows that the east-west velocity has large variations in longitude; vast regions exceed similar to 140 m s(-1). Our global algorithm, and all of the velocity-extraction algorithms used in previously-published studies, found the east-west drift velocities of the visible dark projections, rather than the true zonal velocity at the visible-cloud level. Therefore, the apparent increase in zonal winds between 2000 and 2008 at 8 degrees N is not a true change in zonal velocity. At 7.3 degrees N, the Galileo probe found zonal velocities of 170 m s(-1) at the 3-bar level. If the true zonal velocity at the visible-cloud level at this latitude is similar to 140 m s(-1) rather than similar to 105 m s(-1), then the vertical zonal wind shear is much less than the currently accepted value. Published by Elsevier Inc.