 |
NRL Monterey, Marine Meteorology Division
|
|---|
|
NRL Monterey, Marine Meteorology Division
|
|---|
| Convective Cloud Top |
|---|
) |
| Cloud top height gives the heights of convective and other upper-level
clouds in a color-coded format. It is based on a combination of NOGAPS
and geostationary satellite data. It is especially useful for analysis of
thunderstorms and other weather systems that produce high cloud tops. To
determine cloud top we first take the infrared brightness temperature from
each pixel of the geostationary satellite image. The temperature at this
pixel is compared to a NOGAPS vertical temperature profile that is valid
for the same time as the image. Interpolation fills in between missing
NOGAPS levels. When a temperature match is found, that NOGAPS level is
assumed to be the cloud top height. For example, if the brightness
temperature is -40 C, the NOGAPS temperature profile is searched for -40
C. If the height of the NOGAPS -40 C temperature is 30,000 feet, then
that level becomes the cloud top at that pixel in the cloud top image. For clear skies or cloud tops below 15,000 feet, the cloud top image shows black, meaning no cloud top is determined. The product omits low clouds because of the difficulty in determining their heights correctly. The absence of clouds at low levels should not be taken to mean that clouds do not exist there. The accuracy of the product depends on various factors, including the optical depth of the cloud in question, i.e. the thickness of the cloud. But it should generally be assumed to be about plus or minus 5000 feet. |
| The cloud top image is similar in appearance to an IR image, but has different information content. An IR image shows cloud height only indirectly: colder pixels usually represent higher clouds. On an IR image, a thunderstorm looks "cold" (or bright white using a black and white color table), and one can assume that the cloud tops of the storm are indeed "high" in a qualitative sense. However, the cloud top product gives the height explicitly. It is especially useful for aviators on long flights. It can be used to estimate the tops of convection that pilots will likely encounter en-route. |
| The cloud top product cannot give cloud tops for clouds lower than 15,000
feet. This is because the assumptions of the product often fail below
this altitude. The "temperature matching" technique, which assumes a
roughly linear decrease of temperature with height, fails at low-levels
where inversions are common. Second, land and sea surfaces often have
nearly the same satellite temperature as low clouds, especially at night.
Thus, the product may type relatively warm land pixels as low-level
clouds. To avoid these problems, our algorithm simply does not allow
clouds below 15,000 feet. The user should not believe, however, that the
absence of cloud tops at low levels in the product means that there are no
clouds there. The second problem concerns thin clouds. Thin clouds may have warmer brightness temperatures than the actual physical temperature of the clouds. This effect occurs because the satellite is "seeing through" the thin clouds to warmer clouds or to the warmer surface below. Thus, a thin cloud tends to have heights that are too low, because the temperature matching technique matches them with a NOGAPS temperature that is higher (lower height) than the physical cloud temperature. |
| Cloud Top Image | Cross Section from San Francisco to New York |
|---|---|
) |
) |
| The image on the left shows the cloud top product, with color-coded heights for the United States. Blues over Nebraska indicate tops from about 25-30 thousand feet. Over eastern Nebraska and Iowa there are isolated regions of tops near 55,000 feet associated with thunderstorms. The image on the right shows a cross-section of cloud tops from San Francisco to New York. Notice the thunderstorms tops of near 55,000 feet (blue spikes) over eastern Nebraska and Iowa. The estimation of thunderstorm tops is one of the primary purposes of this product. | |
| Cloud Top Image | Infrared Image (basis of Cloud Top Image) |
|---|---|
|
) |
| The image on the left is the same cloud top example as was shown above. The image on the right is the infrared image that it is based upon. The brighter whites on the infrared image indirectly indicate high cloud tops, since colder cloud temperatures occur at higher altitudes. However, no direct quantitative information about cloud top heights is available from the infrared image alone, prompting the need for a separate cloud top image. Notice that the cloud top image does not show clouds below 15,000 feet, as off the Southern and Baja California coasts, and off the East Coast. These clouds do appear in the infrared image; however, the cloud top algorithm omits them. (See limitations above for more explanation). The cloud heights shown for thin cirrus clouds are usually too low. For example, look at the cirrus in the infrared image (right) north of the Great Lakes. The blues and purples (left image) suggest middle clouds, but the true tops are higher. (See limitations above for more explanation.) | |
Author: Tom Lee Last Updated: Tue Dec 3 07:15:36 2002 Produced by: The Composer (Ver: 1.1.1 ) |
|