Monday November 14 was my first full day in the Maldives. The night I arrived, it started raining right on cue, just about the time I got off the plane. There were at least a couple hours of heavy rain that evening, and even some thunder (which is relatively unusual here). The next morning, there was again heavy rain for a couple of hours. It caused a lot of the streets to turn into deep puddles that lasted through the day, and completely soaked S-Polka radar chief scientist Bob Houze on his bike ride to work. Here's what it looked like on the SMART-R radar in the thick of it, about 10 AM local time (5 UTC):
You can see that the echoes are not widespread at all, but pretty concentrated right around the radar. In the infrared satellite image - which basically sees high, thick clouds, similar to outgoing longwave radiation which we have discussed in previous posts - there is no trace of anything at Gan. (The tiny white speck you see if you look close is not a cloud, it's there in every image and is apparently some permanent feature.) Our rain here was invisible from space in the IR.
Later, though, a line of cells straddling the equator, slanting NW/SE, did develop just west of Gan and move towards the east. Here it is on the IR, at 6:40 pm local time:
This line continued to develop more later. Maybe this line was related to our earlier shower, or maybe the coincidence in time was incidental; from the radar it seemed that the cells on top of us were not the same as those that turned into the line, though they were close in both space and time.
This episode shows how localized tropical convective rain can be. What amazed me about it was how long the rain lasted here at one spot, while still remaining invisible from space. Systems are usually moving, so if they last a while at one spot, that usually means they are big, and then one would think they'd show up on satellite images. This one went on for a couple of hours while staying out of the satellites' view; I suppose it was just moving slowly.
It's a cliche in our field to say that weather varies on a wide range of scales. In the field, one can gain a new appreciation of the truth behind the cliche by observing many of the scales at once.
This is a really interesting piece, thanks for the work on the blog.
ReplyDeleteWhat does this mean for precip datasets based on satellites (i.e. TRMM), i.e. how many localized events like this are potentially missing?
Keep up the good work guys!
That's a good question. I'm not sure I can answer it authoritatively, but the TRMM precipitation radar (PR) has resolution comparable to a ground-based one and I would expect it could have seen our system had it been passing over at the time (the biggest inherent limitation of TRMM in my book is coverage - it doesn't see a large area at once and doesn't repeat any given point even once per day). But that's the PR; most other satellite data sets (and even many of the precip products produced by TRMM, which has many different ones using different types of measurements, not all of which come from the TRMM satellite) are based to some degree on infrared and they would miss it entirely. Of course in an area-integrated sense, they're still not missing much, because storms that rain hard on a very small area for a couple hours still aren't producing a lot of rain overall in the big scheme of things. But I can't make that statement quantitative.
ReplyDeleteGreat post! This entry, and the one that showed pictures of clouds made me search for some pictures of clouds from back home (Puerto Rico). From memory, and from the pictures, I recall that many strong showers from towering clouds that had a rather small base. They develop quickly and were either loners or would cluster around another, much larger cloud. They rained hard, but if you were driving you could get away from the rain in like a minute or so. I've never had the curiosity to see if the IR images actually captured them. They were not uncommon during the summer months though.
ReplyDeleteGreat blog an great stuff going on at Gan!