TORNADIC STORMS - Western Oklahoma
9 October 2001
1-minute imagery from GOES-12
NSSL & CIMSS
to becoming operational, the newest in the series of US geostationary satellites,
GOES-12, has been in a 'science test' mode during September-October 2001.
The imager has improved resolution in the water vapor (6.5 micron) channel.
The imager was collecting rapid scan data during the formation of severe
thunderstorms in western Oklahoma. These storms produced several tornadoes,
one of which which hit Cordell, Oklahoma. The rapid scan data provided
images at 1 to 5 minute intervals.
In these close
up views, the visible images have been magnified 2 times. The loops are
centered on the moving frame of reference of the storm:
loop 1 : (2123-2242
loop 2 : (2245-2353
cover a larger region which includes most of the cluster of storms. They
are not magnified and are centered near the location of initial storm development.
loop 3 : (2123-2242 UTC)
loop 4 : (2245-2353 UTC)
loops contain infrared imagery from several channels:
channel (10.7 microns) loop 1 : (2123-2242
channel (10.7 microns) loop 2 : (2245-2353
vapor (6.5 microns) loop 1 : (2123-2242
vapor (6.5 microns) loop 2 : (2245-2353
IR (3.9 microns) loop 1 : (2123-2242 UTC)
IR (3.9 microns) loop 2 : (2245-2353 UTC)
window (13.3 microns) loop 1 : (2123-2242
window (13.3 microns) loop 2 : (2245-2353
- dirty window (10.7-13.3 microns) loop 1 : (2123-2242
- dirty window (10.7-13.3 microns) loop 2 : (2245-2353
- water vapor (10.7-6.5 microns) loop 1 : (2123-2242
- water vapor (10.7-6.5 microns) loop 2 : (2245-2353
interesting to note that the brightness temperature difference (10.7 -
6.5 microns) is positive over the thunderstorm
tops. Values range from 6-7
K near the back edge of the clouds to 1-2 K downwind of the overshooting
tops. The difference appears to have the opposite sign as would be expected
from previous studies using METEOSAT (Schmetz et al., 1997). The brightness
temperature in the water vapor band was often observed to be warmer (by
as much as 6-8 K) than in the window band above cold top convective clouds.
This observation was explained by stratospheric water vapor which emits
radiation at higher stratospheric temperatures than the emitted radiation
at cloud top. Their results suggested that the magnitude of the difference
in brightness temperature depends on the lapse rate above the tropopause.
Moreover, the greatest differences are expected when cloud tops are located
at the tropopause.
In light of the METEOSAT observations,
the possibility of a differential error in calibration between the two
channels should be considered. The GOES-12 measurements can be compared
to GOES-8 in Table 1. The (10.7 - 6.5 microns) difference is positive
for both GOES-12 and GOES-8, however, the magnitudes are about 2 K less
for GOES-8. If we were to consider a bias of several degrees and subtract
it from differences of both satellites, the results are physically similar
to those of Schmetz et al., 1997.
The greatest differences exist over the widespread
anvil (which presumably is near the tropopause height). They also occur
on a smaller scale just downwind of overshooting tops where additional
water vapor would be entrained into the stratosphere. The smallest differences
exist above some of the coldest (presumably, higher than the tropopause).
If, on the contrary, bias errors are not significant, than an alternate
hypothesis is in order.
Schmetz, J., S.A. Tjemkes, M. Gube,
L. van de Berg, 1997: Monitoring deep convection and convective overshooting
with METEOSAT. Adv. Space Res., Vol. 19, No. 3, pp433-441.
Note that to animate the full time period requires
the download of 53 images. It may take a while before the movie is set
up and ready to run, depending on network speed & the amount of memory
available to your computer. The loops use a Java
Applet written by Tom Whittaker at the Cooperative Institute
for Meteorological Satellite Studies ( CIMSS
) at the University of Wisconsin-Madison.
Choose from the entries in Table 1 to compare individual
bands and brightness temperature differences at indicated times (UTC).
GOES-12, GOES-8 comparisons are available at times indicated in bold/italics.