The severe thunderstorm that was producing a significant long-track tornado (approximately 75 miles long, and around 1 mile wide at times) across parts of northeastern Oklahoma and southwestern Missouri on 10 May 2008 exhibited a classic “enhanced-v” signature on both GOES-12 and NOAA-16 IR imagery. The coldest cloud top brightness temperature values in the region of the enhanced-v were -79 C on the 1-km resolution NOAA-16 AVHRR data, compared to -67 C on the 4-km resolution GOES-12 IR data. In addition, the visible imagery revealed a well-defined anvil plume spreading northeastward from the overshooting top region. Large hail (up to 2.75 inches in diameter) and wind gusts to 65 mph were reported around or just after the time of these images at Joplin, Missouri (station identifier KJLN). The long-track tornado produced EF-4 damage, and was responsible for at least a dozen fatalities in the Picher, Oklahoma and the Seneca/Racine, Missouri areas (SPC storm reports | Tulsa OK NWS summary | Springfield MO NWS summary).

View only this post Read Less

On the morning of 08 May 2008, we received the following email from David Zaff, Science and Operations Officer at the National Weather Service Forecast Office at Buffalo NY:

The NWS BUF office was marveling over the fine line starting from below the thumb of MI through Toronto and along the northern edge of Lake Ontario, but we had a hard time describing why it was there. A cold front had moved through overnight with widespread rain and this boundary appeared as the main synoptic event passed. It doesn’t appear to be tied to land/lake interactions or terrain. It’s clearly drier on one side and more moist on the other with a cap in place. There are also some neat gravity waves that pass into lake Ontario as the flow passes over the boundary when looping the vis earlier in the day.

Great question Dave, and a very interesting satellite imagery example! After some initial pondering, the satellite geeks at CIMSS came to the consensus that the sharp quasi-stationary cloud boundary seen on GOES-12 visible channel images (above, oriented WSW to ENE across lower Michigan,  southern Ontario, and northern Lake Ontario) was possibly due to an atmospheric hydraulic jump. A hydraulic jump can occur when fluid flow at a higher velocity discharges into a zone of lower velocity — a thinner layer of faster, laminar flow abruptly transitions to a deeper layer of slower, more turbulent flow.

But why would a hydraulic jump cause the sharp clearing line seen in the field of cumulus clouds? The answer might lie in the vertical structure of the atmosphere over that region; rawinsonde data plotted for Gaylord, Michigan, Detroit, Michigan and Buffalo, New York (below) revealed that a very strong subsidence inversion existed over lower Michigan that morning. The base of the temperature inversion (which was most dramatic at Detroit, yellow sounding plot) was around 900 hPa, with very dry air in place above that level — if turbulent flow associated with the hydraulic jump acted to rapidly entrain some of that very dry air to lower altitudes, such a marked cloud clearing could result.

GFS40 model fields (below) indicated that there was a weak post-frontal trough at 850 hPa, with a band of convergence just ahead of the trough axis that seemed to correspond to the sharp cloud boundary on the GOES-12 visible imagery. Such a band of convergence is consistent with the idea of “higher-velocity flow discharging into a zone of lower velocities” — the slowing of the flow would effectively create convergence.

A vertical cross section using initial condition (0-hour forecast) fields from the GFS40 model (below) showed the very low relative humidity values that existed above the low-level temperature inversion (over the northern 2/3 of the cross section area). Both the model ageostrophic vertical circulation and the omega fields suggested that there was a shallow region of boundary layer downward motion over  southern Ontario, approximately where the hydraulic jump and cloud clearing line were located around 12 UTC.

In addition to the hydraulic-jump-induced sharp cloud clearing line seen on GOES-12 visible imagery, there was also evidence of a downstream undular bore; parallel cloud bands were apparent on 250-meter resolution MODIS true color imagery (from the SSEC MODIS Today site) centered over western Lake Ontario (below), as well as farther to the west over Lower Michigan.

Another question that remains is: what role (if any) did topography play in the formation and/or maintenance of such a hydraulic jump? An AWIPS topography image (below) shows that there is some slightly higher terrain over  southern Ontario, but the general orientation of the topography does not match that of the sharp clearing line seen on satellite imagery.

View only this post Read Less

Chile is one of the most volcanically active countries on Earth. The latest volcano to erupt is Chaiten, which had previously lain dormant for at least 1000 years. Chaiten is at approximately 42 degrees South latitude, 72 degrees west longitude, close to Golfo de Ancud. A series of eruptions, starting on May 2, has prompted the evacuation of Chaiten, a provincial capital with a population of 4000.

GOES-10 captured the plume of an eruption that started near 12 UTC on 6 May. Volcanic ash does not have an emissivity of 1; that is, it does not emit as a blackbody. The emissivity at 10.7 microns is smaller than the emissivity at 12 microns. The smaller signal received at 10.7 microns (relative to the assumed blackbody) is interpreted as a cooler emitting surface. If the blackbody temperatures at 10.7 and 12.0 microns are compared, then, values at 12.0 microns are warmer. A channel difference can be used to highlight the horizontal extent of the volcanic ash. In the loop shown, the bluest pixels correspond to a blackbody temperature difference of nearly 10 K. That is, the 12 micron blackbody temperature is 10 K warmer than the 11 micron blackbody temperature. The remnants of an older eruption are also noted near the Atlantic Coast of Argentina.

A sequence of GOES-10 imager and sounder IR difference products during the 02-08 May period (below) shows evidence of plumes from multiple eruptions of the Chaiten Volcano. The GOES-10 Imager can provide nearly continuous (15 minute) coverage of the evolving ash cloud, while the GOES-10 Sounder can provide details on the upper-level SO2 plumes once every four hours. The former is derived utilizing the 11.0 micrometer and 12.0 micrometer bands from the Imager. SO2 plumes are revealed by differencing the 7.4 micrometer and 13.3 micrometer bands from the Sounder.

Click here to see an animated gif every from every four hours — that is, each hour for when sounder data are available.

An AVHRR false color image from 05 May (below, viewed using Google Earth) revealed a long plume from the Chaiten volcano, which stretched eastward across Argentina and then southeastward over the South Atlantic Ocean.

View only this post Read Less

AWIPS images of the GOES-12 6.5 µm “water vapor” channel (above) showed the development of late-season winter storm that produced heavy snowfall and blizzard conditions (with wind gusts to 64 mph or 29 m s-1 at Rapid City SD) across the Black Hills region and surrounding portions of eastern Wyoming, southeastern Montana, and western South Dakota during the 30 April through 02 May 2008 period. A closer view using GOES-12 10.7 µm IR images reveals that there were a number of cloud-to-ground lightning strikes during various phases of the storm’s development, as large convective elements formed and intensified upwind of the Black Hills.

A MODIS true color image (above, viewed using Google Earth) shows the areal extent of the snow cover on 03 May 2008, as well as the locations of the heaviest total snowfall amounts: 54.5 inches (138 cm) at Lead SD, and 24.0 inches (61 cm) at Sundance WY.

A comparison of MODIS true color and false color images from 03 May (above) demonstrates how the false color imagery (which uses the MODIS 2.1 µm near-IR channel) can differentiate between snow cover (which, along with ice crystal clouds and ice-covered lakes, appear as cyan-colored features) and supercooled water droplet clouds (which appear as shades of white).

Consecutive daily MODIS false color images from 02, 03, and 04 May (below) show that the snow cover was melting rapidly under the influence of the strong May sun. The 54.5 inches of snow that fell at Lead SD had a liquid equivalent of 4.64 inches (12 cm), so the rapid snowmelt led to a quick rise of creeks and streams that caused some flooding problems.

View only this post Read Less