- Interactive visible and 10.7 micron longwave IR comparison -
A low-level mesoscale vortex was observed over the western North Atlantic Ocean on 15 September 1998. NOAA GOES-8 visible imagery (above left) shows the vortex as a well-defined swirl of clouds centered around 31 North latitude, 73 West longitude. The cloud features within this vortex resemble spiral bands that are often seen with tropical cyclones. Animation shows that the vortex motion was to the west. GOES-8 10.7 micron longwave InfraRed (IR) imagery (above center) showed that the cloud tops associated with this mesoscale swirl were fairly warm (+10 to +20 C), indicating a lower-tropospheric feature. GOES-8 6.7 micron IR (water vapor) imagery (above right) showed no evidence of the low-level vortex, due to the fact that the weighting function for this IR channel peaks in the middle to upper troposphere (usually between 500 and 300 hPa).
The 12:00 UTC run of the ETA model on 15 September showed no hint of this feature at 1000 hPa or 850 hPa. In the upper troposphere (400 hPa), the ETA did reveal a cyclonic circulation located to the southeast of the vortex. This upper-level cyclone had remained nearly stationary near 29 North latitude, 70 West longitude during the preceeding day (14 September), as indicated by GOES-8 water vapor imagery. The low-level vortex did not appear as well defined in GOES-8 visible imagery on 14 September, but it could be seen drifting slowly northward along the 70 West longitude line that day (the clusters of strong convection that developed near 30 N / 70 W from late in the day 14 September to early in the day 15 September were likely due to low-level convergence being enhanced by the vortex, as upper-level diffluence was occurring within the deformation zone just northeast of the upper-level cyclone). On 15 September, the vortex began to move westward as it traversed from the eastern quadrant to the northern quadrant of the stationary upper-level cyclone (convection across that region was seen to dissipate after the departure of the low-level vortex).
The GOES-8 cloud tracked winds shown below were produced by the CIMSS Tropical Cyclones Team. On the left, low-resolution visible cloud tracked winds did not indicate any evidence of the vortex circulation, although a 20-knot wind vector was calculated near the western edge of the feature. These winds were calculated using lower resolution 4km visible imagery at 30 minute intervals. On the right are high resolution visible winds calculated using 1km resolution imagery at 15-minute intervals. These high-resolution winds did indeed reveal the cyclonic circulation of the low-level vortex, with wind speeds as high as 25 knots near the northern edge of the feature (wind speeds of 30 knots were calculated 3 hours later, at 15:45 UTC). No surface ship reports were seen near the vortex circulation, but ships located to the west and north reported wind speeds of only 5-10 knots (see visible image animation above).
This vortex persisted on 16 September, and moved inland along the coast of North Carolina on 17 September where it produced light rain and some isolated thunderstorm activity (the banded structure of the vortex was also evident on WSR-88D radar from Moorhead City, NC and Wilmington, NC). This feature was still handled poorly by the ETA model, judging from 12:00 UTC 1000 hPa wind streamline analyses on both days.