GOES “operational” vs. “mesoscale” winds

April 24th, 2007 |

AWIPS GOES operational winds

An intense upper-tropospheric cutoff low was moving across the southern Rocky Mountain region on 24 April 2007, which was very evident on GOES-12 water vapor imagery (QuickTime animation). An AWIPS image of the 17 UTC operational “GOES high density winds” overlaid on a 17:10 UTC water vapor image (above) did not adequately resolve the closed circulation; in contrast, the CIMSS “GOES mesoscale winds” (or “Atmospheric Motion Vectors”) generated at 16:15 UTC (below) did show the closed circulation of the upper low over eastern Colorado. These mesoscale winds are part of a suite of products available on the Satellite-based Nowcasting and Aviation Application Program (SNAAP) site.

CIMSS mesoscale winds

The first tropical cyclone of the season?

April 20th, 2007 |

GOES-12 visible image

No…just a weak cyclonic vortex off the Atlantic coast of Florida that just happens to look like a tropical cyclone with an eye! GOES-12 visible imagery (above; Java animation) shows the cloud swirl as it developed what appeared to be an “eye” on 20 April 2007. While this vortex was responsible for some weak banded offshore rain features (radar reflectivity image), the GOES-12 10.7 µm IR brightness temperatures were still very warm (around 0º C), indicating a lack of deep convection. The vortex existed in a high-shear environment — around 90 knots on the CIMSS wind shear product (below) — which was not favorable for tropical cyclone development. In addition, the MODIS sea surface temperatures in that region were still below the 80º F (26.7º C) threshold generally considered necessary for tropical cyclone genesis, and GOES sounder total precipitable water values were only in the 20-30 mm (0.79-1.18 inch) range.


The history of this particular swirl is rather interesting. A Weather Channel Blog posting discussed a possible Mesoscale Convective Vortex (MCV) origin; if we follow satellite imagery back in time for a day or two, the development of this feature appears to have been tied to a fairly large mid/upper tropospheric cyclone that was just south of the Great Lakes on 18 April (QuickTime animations: GOES water vapor | GOES IR). A potential vorticity (PV) anomaly associated with the cyclone shows up as a local maximum of GOES Sounder total column ozone (light green to red enhancement on this QuickTime animation); a GFS model cross section through the region of the swirl at 18 UTC on 20 April (below) suggests that a PV “tail” lagging behind the main PV anomaly farther offshore may have played a role in helping to spin up the swirl.

GFS model cross section

Wildfire in southeastern Georgia

April 17th, 2007 |

MODIS + GOES shortwave IR, IR window images

What would eventually become the largest wildfire in Georgia history (the 472,000 acre Sweat Farm Road / Big Turnaround fire) began to burn in southeastern Georgia (in central Ware county, just south of Waycross, station identifier KAYS) during the afternoon hours on 16 April 2007. Much of that region had been experiencing moderate drought conditions, and high winds that day (gusting to 38 mph at Waycross) helped the fire to grow quickly, forcing some local residents to evacuate their homes. AWIPS images of the MODIS and GOES shortwave IR (3.7/3.9 µm) and IR window (10.7/11.0 µm) channels (above) revealed a significant “hot spot” associated with this fire complex at 03 UTC on 17 April (11 PM on 16 April, local time). Compared to the 4-km resolution GOES image, the 1-km resolution MODIS image offered a more accurate depiction of the shape of the active fire (reported to be 1 mile in width, and 12 miles in length), and also showed another smaller fire that was burning to the east of KAYS at that time. Note that portions of the larger fire were hot enough to saturate the shortwave IR detectors on both MODIS and GOES, yielding black pixels (“NO DATA” pixels) on the AWIPS imagery. McIDAS imagery of the fire hot spots depicted maximum brightness temperatures of 329.0 K (132.5 F) on GOES-12, 330.0 K (134.3 F) on MODIS channels 20/22, and 354.9 K (179.15 F) on MODIS channel 23. In addition, the fire was even hot enough to exhibit “warm pixels” on the IR window channel images (42º C on MODIS, 15º C on GOES). The automated CIMSS Wildfire ABBA product indicated saturated fire pixels over that area as early as 20:15 UTC on 16 April (4:15 PM local time).

It was not possible to locate the hottest fire pixel within the cluster of black “NO DATA” pixels on the AWIPS MODIS Band 20 (3.7 µm) IR images; however, one could use the MODIS Band 7 (2.1 µm) image (below) to pinpoint where the hottest portion of the fire was at that particular time (denoted by the small cluster of “bright” pixels in the upper right panel, near the center of Ware county in Georgia). Although the Band 7 imagery is primarily used for snow/ice detection, this near-IR channel (available during daytime only) will exhibit a signal where the hottest pixels are located — this is due to the fact that higher temperatures shift the peak of the Planck function to shorter wavelengths.


GOES-12 visible and 3.9µm IR imagesGOES-12 visible and 3.9 µm shortwave IR imagery (above; Java animation) showed a thick smoke plume drifting southeastward during the day on 17 April; the smoke reduced surface visibilities (as low as 1/2 mile at Jacksonville, Florida) and slowed traffic on area roadways. On McIDAS, the saturated fire pixels “wrapped around” and were displayed as very cold pixels (white enhancement) on the 3.9 µm IR imagery. The land surface beneath the thick smoke plume exhibited a slightly cooler 3.9 µm brightness temperature (lighter gray enhancement), due to a reduction in solar radiation arriving at the surface.MODIS true color imagery (below) indicated that the smoke from this fire had drifted as far as the northern Bahama Islands.MODIS true color image

Record-setting Nor’easter

April 16th, 2007 |

A record-setting Nor’easter storm system intensified over the eastern and northeastern US on 15 April / 16 April 2007 — this storm produced several tornadoes (including a fatal EF3 tornado near Mulberry, South Carolina), winds of 70-80 mph with a gust to 156 mph (Mount Washington, New Hampshire), waves to 33 feet (Buoy 44025), snowfall up to 23 inches (Locke, New York), 9.30 inches of rainfall (Riverdale, New Jersey), and widespread coastal flooding. QuickTime animations of GOES IR, water vapor, and visible channel imagery from AWIPS reveal the unusually large size of the cloud and water vapor fields associated with this powerful Nor’easter.



AWIPS GOES sounder ozone image

The GOES sounder-derived total column ozone product (above) depicted very high ozone values (400-450 Dobson units, red enhancement) along the western periphery of the storm. A southwest-to-northeast cross section through the storm using GFS model fields (below) show that the dynamic tropopause — assumed to be the 1.5 Potential Vorticity Units (PVU) surface — extended below the 600 hPa pressure level in both the high ozone feature (across northwestern Virginia) and also the core of the 500 hPa low (near Long Island, New York).
AWIPS GFS model cross section