Stratocumulus to Stratiform and Everything Between (Warning: Long, Wonky)

I was going to write something about why an atmospheric scientist like me, who identifies primarily as a theorist and modeler, should go in the field --- and why I not only take every opportunity I can to do so, but bring as many students, postdocs and younger colleagues with me when I do. But my students Melanie and Zane have already written so articulately and persuasively on that theme that there is not that much left for me to say about it. So today I am going to write about science, and in particular some of what we observed on the NCAR GV flight I went on a week ago, August 12.

On that flight, we performed our B2 pattern, in which the plane makes a rectangular zigzag or "lawnmower" pattern from south to north (see the maps in this blog's first post), with the first east-west leg at 3N and the last around 11N up near our landing point at Liberia airport. At the southern end, one is close to the equatorial cold tongue, so the sea surface temperature is relatively cool. Then it warms up as one goes north. The sea surface temperature (SST) gradient, or spatial contrast between south and north, is sharper here in the eastern Pacific than just about anywhere else in the Tropics, and this is one of the most important motivations for doing OTREC here.

The sharp SST gradient has effects on the atmospheric circulation, and the clouds within it, that are central to some ongoing debates about the dynamics of the tropical atmosphere. Today, I'm not going to resolve those debates, though, nor even explain what they are. Instead, I'm going to show a few pictures to illustrate a beautiful example of something that I've long taught students about, but never observed so directly myself. Namely, the transition in cloud types over as the SST increases: from stratocumulus; to trade cumulus; to isolated deep convection, or "congestus" clouds; to organized deep convection. Each type of cloud is deeper than the one before, as the air at the surface becomes warmer, moister, and more buoyant, with increasing SST, and thus able to rise higher in the tropical atmosphere. That's about as much as I'm going to explain the physics; from here on out it's just pictures.

Below is a photo I took out the window from near 3N, just before we started the first west-east leg of our lawnmower pattern. That's a deck of low stratocumulus cloud, though one with some holes in it. The cloud top was less than 1 km above the ocean surface; I know this because of the temperature and humidity profiles taken by the dropsondes on this leg, one of which is shown below, and all of which were quite similar. To my surprise, the downward-pointing cloud radar on the plane couldn't see these clouds at all (not shown), but the radar engineer on board explained that this was not surprising to her. The water droplets that make up these kinds of clouds tend to be quite small, so they don't reflect the radar beam very strongly. Also, we were high up (around 42,000 feet, or nearly 13 km) and the clouds were quite low, so that large distance also makes the clouds harder for the radar to see.

Photo taken from the window of the NCAR GV at 7:21 local time (1321 UTC) on August 12, 2019. Position 3.1N, 86.0W.

Below is the first dropsonde trace of the flight, from near where the above photo was taken. If you know how to read a skew-T it shows you a well-mixed subcloud layer topped by a thin saturated layer (i.e., cloud) and a sharp inversion above it, a bit below the 925 hPa pressure level. This is a typical marine stratocumulus-topped boundary layer, entirely consistent with the photo.

Skew-T diagram from dropsonde released near the time and place of the photo shown above.

On the next leg - a little while later, a couple of degrees higher latitude (5.5N), and a bit warmer SST - we started to see rows of "trade" (that is, shallow) cumulus poking up through the stratocumulus. These were not isolated cumulus clouds, but rather organized in long streets. You can see that the stratocumulus is getting thinned out and breaking up. This behavior has been explained as a consequence of the deepening cumuli entraining more dry air from above the inversion, evaporating the cloud water.

Photo taken from the NCAR GV at 8:42 local time (1442 UTC) on August 12, 2019. Position 5.3N, 87.6W).

Just a little while later, along the same leg (so, still at 5.5N), the stratocumulus deck disappeared entirely, leaving only the shallow cumulus streets.

Photo taken from the NCAR GV at 8:51 local time (1451 UTC) on August 12, 2019. Position 5.3N, 88.7W).

 

 

A dropsonde skew-T from around this time shows a boundary layer that is now no longer well mixed. The top half is weakly stable, with potential temperature increasing gradually with height, and water vapor mixing ratio decreases with height. The inversion is weaker and located at higher altitude than in the previous skew-T. This one is indicative of a trade cumulus boundary layer, and again consistent with the photos.

Skew-T plot from a dropsonde launched from near where the just above images were taken.

A 5-minute reflectivity "curtain" from the cloud radar shows a few of these trade cumuli as the plane passed over them. You can see that they top out a little above 5000 feet, or around 1.5 km, which is a little above the inversion in the skew-T above.

As we went further north, and passed over still warmer SST, we started to see a few cumulus congestus - isolated deep convective clouds - forming to the north. Here's a photo of one of them, below. The sun was getting higher in the sky, and my phone's camera was not so good at coping with that so it looks a little overexposed, but you can see the cloud sticking up a lot higher than its neighbors, right in the middle of the image.

Photo taken from the NCAR GV at 9:14 local time (1542 UTC) on August 12, 2019. Position 6.4N, 87.4W, looking north).

Below is a skew-T from a dropsonde that, as best as I can figure, was launched near to the cloud in the above image. (When I took that photo we were at 6.4N; the next west-east leg was at 7.5N, and that's where the below sonde was dropped. I'm not great at estimating distances from the plane but around 1 degree, or 110 km, seems like a plausible guess for how far away that cloud was. Could have been less. Anyway, it was a bit to the north, over warmer SST.)  The inversion is gone, and the atmosphere is pretty humid -- though not saturated -- up to about the freezing level, which is typically as high as one expects congestus to get.

Skew-T from dropsonde launched somewhere in the vicinity of the congestus cloud shown in the image just above.

By the time the plane got to around that latitude, 7.5N, the radar was seeing little showery cumuli below us that were deeper than the ones down at 5.5N. Not as deep as the congestus I shot in the photo above, but that was an isolated cloud and we'd have to have been lucky to fly right over it.

As we got to still higher latitude (8.7N), we got into real organized deep convection. I don't have good photos from this time. The funny thing is that although the deep convection is what we really came to see, one typically doesn't get good pictures when one is in it, or anywhere near it, as the outflow just makes everything grey and one can't see very far. But here is a radar reflectivity curtain beginning at 8.7N, 86.6W and continuing east. You can see that there was a lot of much deeper cloud. Some of the echo reaches the ground, indicating light rain. Where there is no echo at the ground, it was probably actually raining harder; the cloud radar can't see very far through rain, so it gets attenuated and just doesn't see down there, and the high reflectivity values with a sharp cutoff to black below are indicative of that.

Reflectivity curtain from the HIAPER Cloud Radar beginning at 10:25 local time (1625 UTC) on August 12, 2019. Plane location at start of curtain 8.7N, 86.6W.

Here's a combined visual and infrared satellite image showing the plane position at the at time. Now we're flying into real deep convection that is organized on the mesoscale - that is, the angry red blob, indicating thick, high, cold cloud tops, with lightning flashes shown by the purple crosses. The large horizontal extent indicates that the strong convective updrafts have generated a large stratiform anvil - not to be confused with the stratocumulus we saw at the start! "Strat" means "layer" and both cover large areas, but the stratiform anvil is thick and high and rainy, while the stratocumulus at lower latitude were thin and low and produced little rain.

Combined visual and infrared satellite image showing NCAR GV position at 10:25 local time (1625 UTC). Purple crosses are lightning flashes.

Skew-T from dropsonde launched at 1629 UTC on August 12, 2019.

To finish, let's look at the big picture. Here's a visible image from during the flight that includes the box bounded by the equator and 10N and 80W and 90W. If you read everything above closely, and stare at this image long enough, you can see the cloud changes I've been describing. I'm going to be using this example in my classes for years to come!

PS: In the meantime, over the last few days we have had a developing easterly wave offshore to the northwest, and both the GV and the NOAA P-3 flying repeatedly to watch as it slowly makes its way towards being a tropical cyclone. There will surely be posts about that before long...