INTERPRETATION
|
Regional and Mesoscale Meteorology Team Cooperative Institute for Research in the Atmosphere (CIRA) Colorado State University Fort Collins, Colorado |
May 17, 1999
|
|
This topic focuses on 2 types of satellite cloud composites/climatologies
using GOES imagery: Visible cloud frequency composites and 10.7 µm
cloud frequency composites. Both types of composites have advantages
and disadvantages and will be referred to here in example imagery over
Costa Rica.
Figure 1 |
Figure 2 |
The visible cloud frequency
composite is very desirable because the high resolution of the imagery
allows for detailed representation of information related to geographic
features such as cloud development in mountainous regions or cloud development
along land/water boundaries. Cloud frequency images can be made for
various time frames, but here the focus is on particular hours over 1 month
periods. For the desired hour, a cloud free image is obtained and
used as a "background" image. The series of images for that same
hour over the month are compared pixel by pixel with the "background" image.
If the pixel in the image in significantly brighter than the pixel in the
background image, it is designated as cloud. The number of cloudy
pixels for a particular hour over the entire month is summed and then represented
as a frequency. Figure 1 is an example
of a 1 km visible cloud frequency centered over Costa Rica for February,
1999 at 20:15 UTC.
February constitutes one of the months of the dry season.
In Costa Rica, this translates to dry on the Pacific side of Costa Rica
and less wet on the Atlantic side of Costa Rica. During June, in
what is considered a wet month for the whole of Costa Rica, it is not uncommon
to have near 100% cloud frequency during the afternoon over most of the
country and significant convection and cloud cover after dark. In
these instances, it is helpful to use 10.7 µm imagery to develop
cloud frequency composites based on cloud top temperatures colder than
a determined value. In Costa Rica, rain falls from low level stratus
as well as deep convective storms. A preliminary cloud frequency
of the 10.7 µm imagery was derived by identifying pixels with temperatures
colder than 0 C as cloud. Figure 3 shows
an example for June 1998 at 20:15 UTC.
When compared with the visible cloud frequency for February
at the same time, overall, the region is cloudier in June than in February.
One would expect a higher cloud frequency in many regions of Costa Rica
in June than in February. This is not particularly evident from these
images for two reasons: 1) The temperature threshold of 0 C is not including
the low stratus that are warmer than 0 K, and 2) June 1998 was less cloudy
than normal due to the influence of El Nino. You can see from Figure
4 that June 1997 at 20:15 UTC (non
El Nino year) had a higher cloud frequency than in 1998.
There are many possiblities for tayloring the development
satellite climatologies. If you haven't checked out the Tallahassee
Summer Sea Breeze Climatology example of satellite regime composites, please
do so at the following address:
|
