Tornado outbreak in Oklahoma and Kansas

May 10th, 2010 |
GOES-13 0.63 µm visible images

GOES-13 0.63 µm visible images

On yet another relatively rare SPC “High Risk” Convective Outlook day, a major outbreak of damaging tornadoes occurred across parts of Oklahoma and Kansas on 10 May 2010 (see: SPC storm reports | NWS Norman OK | NWS Witchita KS). McIDAS images of the GOES-13 0.63 µm visible channel data (above) showed the development of numerous long-track supercell thunderstorms across the southern Plains region — much of this development was along and ahead of an advancing dryline, where GOES-13 Sounder Derived Product Imagery (DPI) showed Lifted Index (LI) values in Oklahoma as low as -14º C 21 UTC (along with Convective Available Potential Energy (CAPE) values as high as 6000 J kg-1 at 19 UTC).

Also note the large hazy plume of blowing dust that was moving northeastward from the Texas panhandle across western and central Oklahoma, as well as a smaller area of blowing dust that moved southward out of southwestern Kansas and into the Oklahoma panhandle region later in the day — this was an indicator of the strong lower tropospheric wind fields that were present over much of the region on that day.

The corresponding set of GOES-13 10.7 µm IR images (below) revealed several “enhanced-v” storm top signatures and a number of pronounced cold overshooting tops — cloud top IR brightness temperatures became progressively colder into the late afternoon and early evening hours, with a number of cloud tops in the -70º C to -80º C range (black to white enhancement).

GOES-13 10.7 µm IR images

GOES-13 10.7 µm IR images

A 1-km resolution MODIS water vapor image (below) showed a very complex array of middle tropospheric wave structures, an indication that the very strong winds were interacting with the terrain of the region.

We will begin by focusing our attention on the severe storm that can be seen developing over northern Oklahoma, near the triple point intersection of the cold front, the dry line, and the warm front.

MODIS 6.7 µm water vapor image + surface fronts

MODIS 6.7 µm water vapor image + surface fronts

The GOES-13 satellite was placed into Super Rapid Scan Operations (SRSO) to support the VORTEX 2 field experiment — so satellite imagery was available at 1-minute intervals for short periods during the afternoon and early evening hours. GOES-13 SRSO visible images (below; also available as a QuickTime animation) showed some very interesting storm evolution and structure associated with a few of the earlier thunderstorms that developed across the Oklahoma/Kansas border region.

Of particular interest was the appearance of inflow feeder clouds that were seen to develop and become ingested into the southwest quadrant of one of the stronger storms during the 20:15 – 20:45 time period — and not long after these inflow feeder clouds were seen on the visible satellite imagery, this storm intensified and produced hail of 4.25 inches in diameter and a large tornado.

GOES-13 0.63 µm visible images (SRSO at 1-minute intervals)

GOES-13 0.63 µm visible images (SRSO at 1-minute intervals)

A comparison of the GOES-11, GOES-15, and GOES-13 visible images that were available during the 20:30 – 20:45 UTC period when these inflow feeder clouds were being ingested into the storm show the value of more frequent imaging, which allows the evolution of such features to be more clearly visualized and understood (below). The images are shown in the native projections of each GOES satellite.

GOES-11, GOES-15, and GOES-13 visible images during the 20:30-20:45 UTC period

GOES-11, GOES-15, and GOES-13 visible images during the 20:30-20:45 UTC period

A 1-km resolution MODIS 11.0 µm IR image at 19:49 UTC (below) revealed the presence of a well-defined cold/warm cloud top temperature couplet associated with a developing “enhanced-v” signature — both cloud top signatures can be indicators a potential for that storm to produce severe weather (a 20-30 minute lead time is typical for such satellite storm top signatures). In fact, this storm did later produce hail of 4.25 inches in diameter at around 20:10 UTC, and then a strong tornado around 20:33 UTC. Note that neither the cold/warm couplet nor the developing enhanced-v signature was apparent on the 4-km resolution GOES-13 10.7 µm IR imagery at that time.

MODIS 11.0 µm IR image + SPC storm reports

MODIS 11.0 µm IR image + SPC storm reports

============================================

POES AVHRR 12.0 µm IR image

POES AVHRR 12.0 µm IR image

Farther to the south, additional large areas of severe convection developed a couple of hours later over southwestern Oklahoma, which then moved eastward and produced large hail and tornadoes in the Oklahoma City and Norman areas. An AWIPS image of the AVHRR 12.0 µm IR channel data (above) showed the dramatic increase in areal coverage of cold cloud tops in southern and central Oklahoma at 22:32 UTC. A close-up view of that AVHRR IR image with an overlay of SPC storm reports (below) revealed a cluster of very cold cloud top brightness temperatures (as cold as -84º C, purple color enhancement) that was likely associated with the reports of tornadoes and hail of 4.00 and 3.75 inch in diameter in the Norman and Tinker Air Force Base areas.

POES AVHRR 12.0 µm IR image + SPC storm reports + surface METAR data

POES AVHRR 12.0 µm IR image + SPC storm reports + surface METAR data

AWIPS images of the CIMSS experimental “Convective Initiation”, “Cloud Top Cooling Rate”, and “Overshooting Top” products are shown below; these products appeared to have some skill in providing some lead time to the development of this severe convection that later affected Oklahoma City and Norman. These CIMSS products are being provided to the Storm Prediction Center as part of the Hazardous Weather Testbed evaluation (a component of the GOES-R Proving Ground activities).

GOES IR + Convective Initiation + Cloud Top Cooling + Overshooting Top products

GOES IR + Convective Initiation + Cloud Top Cooling + Overshooting Top products

UPDATE: Added an N-AWIPS loop of Visible imagery with lightning and Convective Initiation, along with an N-AWIPS loop of Visible Imagery with Cloud-top cooling and warning polygons.

An undular bore and an oil slick off the Gulf Coast

May 8th, 2010 |
GOES-13 0.63 µm visible images

GOES-13 0.63 µm visible images

McIDAS images of GOES-13 0.63 µm visible channel data (above) revealed the formation of a packet of wave clouds over the northern Gulf of Mexico, associated with an undular bore moving southward ahead of an advancing cold frontal boundary on 08 May 2010.

The clouds eventually cleared out enough to reveal portions of the oil slick (which remained off the coast of Louisiana following from the explosion and sinking of the Deepwater Horizon offshore oil rig) on 250-meter resolution Red/Green/Blue (RGB) MODIS true color and false color images sourced from the SSEC MODIS Today site (below). Since the oil slick feature was once again located within the sun glint portion of the MODIS image swath, it was very easy to detect on the true color (created using bands 1/4/3 as the R/G/B channels) and false color (created using bands 7/2/1 as the R/G/B channels) imagery.

It appears as though a thin filament of the oil slick has recently been drawn westward (away from the core area of the oil slick near the source), and has been entrained into the sediment outflow region of the Mississippi River. You can follow the changes in appearance of the oil slick on this comparison of MODIS true color image from 21, 25, and 29 April and 01, 04, and 08 May.

MODIS true color and false color Red/Green/Blue (RGB) images

MODIS true color (bands 1/4/3) and false color (bands 7/2/1) images

The oil slick appears as a light gray feature on an AWIPS image of the MODIS 0.65 µm visible channel data, but shows up as a very warm (darker gray to black enhancement) area on the MODIS 3.7 µm “shortwave IR” image due to a high amount of sun glint reflection of solar radiation off the oil slick surface. However, note that the oil slick feature does not exhibit any sort of signature at all on the MODIS 11.0 µm “IR window” image (below).

MODIS 0.65 µm visible, 3.7 µm

MODIS 0.65 µm visible, 3.7 µm

===== 09 MAY UPDATE =====

MODIS true color (bands 1/4/3) and false color (bands 7/2/1) Red/Green/Blue (RGB) images

MODIS true color (bands 1/4/3) and false color (bands 7/2/1) Red/Green/Blue (RGB) images

The oil slick was once again a very obvious feature in the 250-meter resolution MODIS true color and false color images on 09 May 2010 (above). In this case, note the appearance of the light pink colored pixel near the center of the oil slick on the false color image — the near-IR Band 7 used in that particular RGB image is also sensitive to hot surfaces (for example, due to a fire), which would make such a feature show up as a light pink feature. Indeed, an AWIPS image of the MODIS 3.7 µm shortwave IR channel data (below) confirmed the presence of a relatively hot pixel (43.5º C, orange color enhancement), which could have been due to a small spot fire that was set in an attempt to burn off some of the surface oil.

MODIS 0.65 µm visible and 3.7 µm shortwave IR images

MODIS 0.65 µm visible and 3.7 µm shortwave IR images

===== 10 MAY UPDATE =====

MODIS true color (bands 1/4/3) and false color (bands 7/2/1) images

MODIS true color (bands 1/4/3) and false color (bands 7/2/1) images

The oil slick was once again a prominent feature in the 250-meter resolution MODIS true color and false color images on 10 May 2010 (above). However, note that the appearance of the oil slick was a bit different than what was seen on the previous days when sun glint was helping to illuminate the feature: on this particular day, the “brighter” portion of the oil slick appeared to be surrounded by a very dark signature. It is not entirely clear what this “dark signature” was on the MODIS true color imagery — but one idea is that it could have been due to oil that was still sub-surface (and had not yet come to the surface where it could then help to reflect light back up toward the satellite in the sun glint region of the overpass swath).

Another significant volcanic ash plume from Eyjafjallajökull in Iceland

May 6th, 2010 |
Meteosat-9 SEVIRI volcanic ash retreival products

Meteosat-9 SEVIRI volcanic ash retrieval products

The Icelandic volcano Eyjafjallajökull (which started to become active again in late March 2010) continued to remain active into early May, with another significant plume being observed on 06 May 2010. EUMETSAT Meteosat-9 SEVIRI volcanic ash retrieval products (above) showed a plume streaming southeastward from Iceland, with the maximum ash cloud height reaching 17.27 km. These volcanic ash retrieval products provide a demonstration of the type of products that will be available with the ABI instrument on the GOES-R satellite — they are available in near-realtime on the CIMSS GOES-R Proving Ground site.

A Terra MODIS Red/Green/Blue (RGB) image (using bands 01/04/03) shows the brown ash plume curving southeastward and then southward over the eastern Atlantic Ocean (below).

Terra MODIS RGB image (using bands 01/04/03)

Terra MODIS RGB image (using bands 01/04/03)

As a result of this most recent volcanic eruption, some airports in Scotland, Northern Ireland and the Irish Republic were closed on 06 May.

===== 07 MAY UPDATE =====

A Terra MODIS RGB image using bands 01/04/03 (below) showed a very long and narrow volcanic plume emanating from Eyjafjallajökull on 07 May.

Terra MODIS RGB image (using bands 01/04/03)

Terra MODIS RGB image (using bands 01/04/03)

Using MODIS imagery to diagnose areas of light winds over water

May 5th, 2010 |
MODIS 0.65 visible image + MODIS Sea Surface Temperature product + NAM surface winds

MODIS 0.65 µm visible image + MODIS Sea Surface Temperature product + NAM surface winds

AWIPS images of the MODIS 0.65 µm visible channel data (above) revealed a few dark features within the otherwise bright sun glint region just off the coast of Florida and Georgia on 05 May 2010. The corresponding MODIS Sea Surface Temperature (SST) product indicated that SST values were also quite warm within these dark visible image features (SST values as high as 79º F, darker red colors) — in fact, just as warm as those seen a bit farther to the east within the Gulf Stream. In contrast, to the north of the dark/warm features, the MODIS SST values were only around 69-70º F (yellow to green colors).

This type of visible image “dark signature” over water illuminated by sun glint has been observed before — for example,  over Lake Michigan — and is recognized as a signature of regions of very light or calm winds. The surface wind field from the NAM20 model (cyan wind barbs) did show a west-to-east oriented axis of calm winds just to the south of the dark/warm signature noted on the MODIS imagery. In this case, the MODIS imagery could have been used to identify the slight southward bias of the NAM20 model wind fields at that time.

MODIS 0.65 µm visible image + NAM surface winds + HPC surface frontal analysis

MODIS 0.65 µm visible image + NAM surface winds + HPC surface frontal analysis

The Hydometeorological Prediction Center (HPC) surface frontal analysis for 15 UTC is overlaid on the MODIS visible image (above).  Judging from the location of the dark features on the visible image, it would appear the the weak cold front had not yet advanced as far southeastward as the area of light/calm winds seen just after 16 UTC .

MODIS true color image (visualized using Google Earth)

MODIS true color image (visualized using Google Earth)

A closer view using a 250-meter resolution MODIS true color Red/Green/Blue (RGB) image  — created using MODIS bands 1/4/3 as the R/G/B channels — acquired from the SSEC MODIS Today site (above, visualized using Google Earth) is annotated to point out the subtle yet potentially important clues as to the exact location of the leading edge of the cold frontal boundary at the time of the Terra satellite overpass (approximately 16:18 UTC). This example also suggests that frontal boundaries do not necessarily resemble the nice smoothly-drawn features we see on weather maps, but instead can be complex in their shape due to various interactions with terrain, coastlines, etc.