Radar echo tops vs. satellite cloud tops: what is ideal?

There has been a lot of discussion over the past couple of weeks on what the term 'tops' means in aviation speak. Radar-derived echo tops or satellite-derived cloud tops? By in the large, the current consensus is echo tops. So, why is this? Check out the example below for the answer...

20150818 1345 UTC cloud top heights (top left), echo tops (top right), IR imagery (bottom left) and visible imagery (bottom right) at the AWIPS-2 CAWS support deask

Cloud tops this morning with the associated convection over the Midwest were upwards of FL350 feet in the most intense of the convection. Echo tops showed tops of FL300. This difference is expected as echo tops would typically be below the top of the cloud. It is in this extra 5000 feet where the cloud top vs. echo top discussion comes. Generally, the rule of thumb is not to fly above anything higher that FL350 and the echo tops in this case gave the green light. However, cloud tops would not have. Losing an extra 5000 feet of airspace to work with would have a significant impact on aviation operations. Furthermore, pilots are typically fairly knowledgable on the clouds they should or should not fly through just by visual inspection from the cockpit. It is for these reasons that those in the traffic flow world tend to go by echo tops.

Should they really be? That question is debated. However, there are times when the echo tops can be vastly lower than the cloud tops, but for very good reason. The two images below highlight an cell in south central Iowa that captures this.

20150818 1345 UTC cloud top heights (AWIPS-2 display)

20150818 1345 UTC echo tops (AWIPS-2 display)

When sampled, cloud tops estimated FL380+ at the highest. However, echo tops were only FL250 - FL300. Why such a huge difference? An overshooting top. The GOES-R overshooting top detection product below confirms this.

20150818 1345 UTC overshooting top detection, visible, and ASDI flight tracks

Most of the time the 'bubbly' texture in the visible imagery can help identify the overshooting top as it did fairly well in this example. ASDI flight tracks show that most traffic diverted around this area of convection. However, it is not always so easy to tell. It is in cases like these that additional information on the growth or decay of the echoes, such as the cloud top cooling may be useful. Being able to identify if the an echo top is associated with a cell that is growing or decaying may provide insight on why there is such a big difference, and whether aircraft need to be aware that a particular area of echo tops will be increasing rapidly even if they are currently low.