Head in a Cloud

troposphere and stratosphere meet blogosphere

Head in a Cloud random header image

“The Role of Cloud-Top Brightness Temperatures in the Production of Heavy Precipitation”

March 7th, 2008 by miller · 3 Comments

“The operational meteorological community is increasingly realizing the important role of cloud microphysics in the production of heavy precipitation, especially snow(e.g., Roebber et al. 2003)”. This is a direct quote from the journal article “Cloud-Top Temperatures for Precipitating Winter Clouds” by Jay W. Hanna, David M. Schultz, & Antonio R. Irving, of which I will summarize. I chose this quote because it reveals why the authors wrote this particular paper. Understanding cloud microphysics can greatly increase operational meteorologist’s understanding of heavy precipitation events. This paper relates cloud-top brightness temperatures to observed surface precipitation types(snow, rain, freezing rain & sleet) in order to better understand the microphysical processes within clouds.


The study analyzed data containing 145,062 surface observations of snow, rain, freezing rain, & sleet with GOES(Geostationary Operational Environmental Satellite) derived cloud top brightness temperatures from the longwave infrared band(channel 4) centered over the surface observing station. This data was collected over two separate winter periods(Feb.-Mar. 2003 & Dec. 2003-Mar. 2004) for a total of six months of data. Cloud-top brightness temperatures from the GOES-8 spacecraft were used for the first winter period while GOES-12 spacecraft data was used for the second winter period. This change in instrumentation had an insignificant impact on the satellite-derived cloud-top brightness temperatures. In order to verify accuracy, radiosonde profile data of cloud-top temperatures was collected & analyzed for a 345 observation subset of the 145,062 original dataset. A comparison of the satellite-derived vs. radiosonde-derived cloud-top temperature data revealed that, for the most part, there were only small differences between the two types of data(most discrepancies were between -5 degrees & 10 degrees Celsius). However, some discrepancies were quite high(up to 60 degrees Celsius). The main reason for these large discrepancies appeared to result from the fact that the radiosonde detected cirrus clouds whereas the satellite could not detect these high thin clouds. Unfortunately, the wavelength range of the channel 4 longwave infrared band of which the satellite uses does a poor job of detecting these optically thin cirrus clouds. Still, the authors emphasize that most of the discrepancies were small.

Analyzation of the data in this study yielded some interesting results. Concerning snow, the distribution of cloud-top temperatures for clouds producing light snow revealed that the peak number of light snow episodes occurred near minus 16 degrees Celsius. However, all the other precipitation types(moderate & heavy snow, light/moderate/heavy rain, freezing rain, & sleet) had two peaks with one peak near minus 16 degrees Celsius & the other peak between minus 35 & minus 50 degrees Celsius. Light snow does not have a peak in the minus 35 to minus 50 degrees Celsius range, because light snow does not tend to form in the less stable environment that is represented by these colder temperatures. The reason that all forms of precipitation had a peak number of episodes near minus 16 degrees lies in the fact that this represents the temperature regime in which the maximum dendritic ice crystal growth occurs. Also, the distributions for snow decrease dramatically at temperatures higher than minus 15 degrees Celsius whereas the rain distributions decrease more gradually from minus 5 degrees Celsius to 0 degrees Celsius. The loss of the dendritic growth process is a major reason for the significant snow decrease. The more gentle rain decrease results from the warm-rain process. It is very interesting to note that the data revealed a pattern of higher rainfall intensity being associated with colder temperatures(minus 45 degrees Celsius for light rain, minus 47 degrees Celsius for moderate rain, & minus 50 degrees Celsius for heavy rain) whereas the most intense snow tended to occur at around minus 16 degrees Celsius. The high correlation between colder temperatures & higher intensity precipitation explains the rainfall patterns. The maximum dendritic ice crystal growth temperatures of around minus 16 degrees Celsius explains the snowfall patterns.

The analyzed data presented in this paper has contributed to a deeper understanding of the microphysical processes within clouds. According to the authors, previous research on the correlation between cloud-top temperatures & snowfall observations involving only snow-producing clouds has not been published. Hopefully, this research will encourage more study on the relationship between cloud-top brightness temperatures & observed surface precipitation as increasing knowledge on this subject can help operational meteorologists better understand the physical processes involved in heavy precipitation events.

Tags: clouds · MTR3440 · satellite · troposphere

3 responses so far ↓

  • 1 Harish // Mar 7, 2008 at 9:45 am

    Indeed an interesting post. However, a few things are not clear to me, probably author of this post can explain.
    What are physical reasons that cloud top temperature required for rainfall is colder than what required for snowfall?

    How do authors account for geography of a given place which may favor certain type of precipitation? What I mean to say is how do we know the correlation between cloud top temperature and precipitation type is due to some physical reason and not due to correlation in geographical variation?

  • 2 poker // Mar 31, 2008 at 12:20 pm

    How do authors account for geography of a given place which may favor certain type of precipitation? What I mean to say is how do we know the correlation between cloud top temperature and precipitation type is due to some physical reason and not due to correlation in geographical variation?

    yes, please answer!

  • 3 Sean Davis // Apr 4, 2008 at 2:09 pm

    Harish, poker,

    If I understand your question correctly, the answer is that one can look at a fixed geographical location and observe a correlation between cloud top temperature and precipitation properties.

Leave a Comment

*