A series of mid- to late-April MODIS true color images from the SSEC MODIS Today site (above) reveals that many of the lakes across northern Minnesota were still ice-covered as of 29 April 2008. According to the Minnesota State Climatology Office, the “ice-out” (or melting) of lakes across northern Minnesota this Spring season is delayed by about 2 weeks compared to historical normals, and this is the latest ice-out of many lakes in the region since 1996.
Severe convection along an advancing cold frontal boundary was responsible for several tornadoes and reports of damaging winds across southeastern Virginia on 28 April 2008 (SPC storm reports | Wakefield, Virginia NWS forecast office storm report). AWIPS images of the GOES-12 10.7 µm IR channel data (above) showed that cloud top brightness temperatures began to cool into the -55 to -65º C range (orange to red enhancement) after about 18:15 UTC (2:15 PM local time), but no classic severe storm signatures such an “enhanced-v” were noted. The strongest tornado moved through the Suffolk, Virginia area during the 20:05 – 20:15 UTC (4:05 – 4:15 PM local time) period, producing EF-3 damage (about 140 homes were either destroyed or damaged) and injuring over 200 people.
A comparison of 4-km resolution GOES-12 IR and 1-km resolution NOAA-18 IR images (below) revealed an area of cold cloud top brightness temperatures between Petersburg (KPTB) and Emporia (KEMV) in southeastern Virginia around 18:40-18:45 UTC (2:40-2:45 PM local time), close to the time that a tornado (denoted by “TOR” on the images) produced EF-1 damage near Lawrenceville, Virginia. Cloud top brightness temperatures were several degrees colder on the NOAA-18 IR image (-61º C, vs. -56º C on the GOES-12 IR image), but still rather unremarkable compared to the -70 to -80º C values that are often seen associated with severe convection in the Plains states. As discussed in the Jon Davies Severe Weather Notes blog, the atmosphere over southeastern Virginia on that day exhibited relatively weak instability (which was confined to low levels), which likely limited the resulting thunderstorm updraft intensities — stronger updrafts would probably have produced significantly colder satellite IR temperatures associated with the overshooting tops.
Hourly images of the MIMIC Total Precipitable Water (TPW) product from 28 April (below) showed that a plume of relatively high TPW (35-45 mm or 1.4-1.8 inches, light blue to green to yellow enhancement) originating over the Gulf of Mexico was streaming northeastward along the Eastern Seaboard during the day, helping to provide the necessary moisture for convective development along and ahead of the cold front that was moving eastward through the mid-Atlantic states.
AWIPS image combinations of the DMSP TPW + GOES-12 sounder TPW (below) also show the TPW plume, and include an overlay of the GOES-derived atmospheric motion vectors. These satellite winds showed that a large trough was centered over the southern Great Lakes and Ohio River Valley; in addition, a diffluent flow pattern within the upper troposphere (the 450-275 mb layers, green and cyan wind vectors) was evident over eastern Virginia. This diffluence aloft likely contributed to the creation of an environment that supported large-scale ascent favorable for severe thunderstorm development over that particular region during the afternoon hours on 28 April.
AWIPS images of the MODIS visible channel, 1.6 µm near-IR “snow/ice” channel, and Land Surface Temperature (LST) product (above) depicted a broad swath of heavy snow on the ground across parts of eastern South Dakota and western Minnesota on 26 April 2008. Snowfall amounts included 19.0 inches at Watertown, South Dakota (which set an all-time record for 24-hour snowfall accumulation there) and 15.5 inches at Brandon, Minnesota. On the MODIS visible image, the unfrozen lakes stand out as dark features against the bright white snow-covered ground; the snow cover appears very dark on the MODIS snow/ice image (since snow is a very strong absorber at the 1.6 µm wavelength), in contrast to supercooled water droplet clouds which appear as varying shades of white. Note how the MODIS Land Surface Temperature values within the snow swath were in the 30-40º F range (green colors), compared to much warmer LST values of 50-60º F (yellow to orange colors) over the bare ground regions on either side of the deep snow cover. The MODIS LST product gives an indication of the temperature of the actual land surface (or “skin temperature”), which can be several degrees different than the air temperatures measured in instrument shelters located about 5 feet above ground level.
A MODIS true color image from the SSEC MODIS Today site (below, viewed using Google Earth) further demonstrated the large contrast between the significant snow cover and the surrounding bare ground. According to the National Weather Service forecast office at Sioux Falls, a Trough of Warm Air Aloft (TROWAL) contributed to the heavy snow event (which forced Interstate 29 to be closed from Brookings – about 50 miles north of Sioux Falls – all the way to the North Dakota border).
Clusters of severe thunderstorms developed along the Texas / New Mexico border region during the late morning and afternoon hours on 23 April 2008. GOES-12 visible images (above) showed multiple low-level convective outflow boundaries that moved westward — and these outflow boundaries then played a role in subsequent convective development later in the day. The GOES-12 satellite had been placed into Rapid Scan Operations (RSO) mode, allowing images to be available as frequently as every 5-10 minutes during the daylight hours.
A comparison of 1-km resolution MODIS and 4-km resolution GOES-12 IR images (below) demonstrated the superior storm top temperature detection capability of the MODIS data — the largest storm over northwest Texas (located in the center of the image, between Abilene KABI and Lubbock KLBB) exhibited a well-defined “enhanced-v” signature on the MODIS imagery, in addition to a “warm trench” signature surrounding an area of overshooting tops just to the west of the enhanced-v. Note that the features on the GOES-12 IR image appear to be displaced a bit to the north/northwest compared to the MODIS IR image — this is a result of “parallax error” due to the large viewing angle of the GOES-12 satellite (which is positioned over the Equator at 75º W longitude).
These thunderstorms produced heavy rainfall, and were also responsible for a number of reports of tornadoes, large hail, and damaging winds (SPC storm reports). Isolated locations reported rainfall of 2 inches or more, creating several large swaths of wet ground which were very evident on MODIS IR and Land Surface temperature (LST) images (below) from the next day (24 April 2008). While the swaths of wet ground remained relatively cool during the following day (yellow to orange to light red colors), the surrounding dry ground areas were able to heat up much more quickly, exhibiting significantly warmer IR brightness temperature and LST values (darker red to black colors). In addition, a burn scar from a large fire that burned back in February 2008 could still be seen, with slightly warmer IR temperatures and LST showing up in the area of the burn scar (located just northwest of San Angelo, KSJT).