Sunday, May 16, 2010

There is a bird in my radar!

During the past few weeks, several of you have asked essentially the same question:

" It's been completely dry but the weather radar last night showed lots of echos, suggesting it rained all night. What is going on ?"

Well, I can give the you the answer: Birds!

Weather radar can see more than raindrops. It can see the mountains, but that signal is generally removed successfully (terrain clutter) since the mountains generally don't move. Weather radar can see other objects in the air and the amount of return generally increases by the sixth power of the object's diameter. (Doubling the size of an object increases the amount of the radar signal scattered back increases by 64 times). Since a bird is much bigger than a raindrop, you can imagine that it would provide a good return.

Another point. The weather radar used by the National Weather Service has two modes: clear-air and precipitation. Clear air mode is much more sensitive and is used when it is not precipitating to get some information on the winds. In this mode, discontinuities in the atmosphere (e.g., where density changes rapidly) and bugs (which get blown about by the wind) can show on the radar (within tens of km of the radar site) to provide some useful information.

But now we get to the birds. During dry periods the radar is on the hypersensitive clear air mode and during the night (particularly during migration periods in spring and fall) a whole lot of birds are up there. According to my birder friends and a few articles I have read on the subject, a number of birds (like songbirds) like to migrate at night, typically flying into the bottom 10,000 feet of the atmosphere. This time of the year they are flying north and in the autumn to the south. The amazing thing is this migration is really tied to the clock...after sunset, the radar is filled with echo, remains all night, and like magic disappears after sunrise. (Keep these numbers in mind: on May 15th, sunset was at 8:40 PM (3:40 UTC/GMT), sunrise was at 5:32 AM (12:32 UTC).

OK, you want to see if for yourself? Here is the radar (Camano Island) at 8:57 PM on May 14th, right after sunset. Not much echo, just ground clutter from terrain and some close in returns near the radar.

Here is the radar image less than an hour later, when it is starting to get dark 9:36 PM). Echo is starting to fill the domain. And it is NOT raining.

And here is the situation at 11:04 PM. Much of the domain visible by the radar has echo. The birds are everywhere!

At 5:06 AM the next day, there is lots of echo still:

But at 6:04 AM (remember sunrise was at 5:32 AM), the echo (and the birds) are mostly gone!

And at 8:02 AM, much less still:

Now want to see something neat? The NWS radar is a Doppler radar and thus can measure the speed of the targets towards or away from the radar. We can check out the velocities of the birds! Here is is:

Greens indicate velocities toward the radar (in knots) and yellows/oranges away. The radar location is in the middle. The birds are migrating north, roughly at 20 kts! (a would have been a problem if it had been the other way).

Ornithologists know about the value of the weather radar and use this technology to track migrations. Occasionally, there are weekend weather people on TV that make the mistake of misidentifying birds for precipitation...but you won't.


smokejumper said...

You need to have a speech on how to avoid being bird pooped on while biking. Use the birdar!

Spencer said...

If radar reflection is proportionate to the size of the object to the sixth, shouldn't planes dominate radar images? Why don't we observe a line of fast-moving blips a mile apart heading in towards seatac?

Upupaepops said...

I am confused as to the date of these images. December? sunset and sunrise times don't jive

I am thrilled to see this article. I never thought about how birds are filtered.

Camano Island is right in the center of shorebird and wintering flocks of snow geese and swans. Must generate some amazing echos.

Cliff Mass Weather Blog said...

Sorry..there was a typo..December should be May!...cliff

Michael said...

I saw lots of birds at dusk last night , made me believe you!

Thanks for clearing it up Cliff

tom897 said...

Fun and interesting post. One technical caveat. Radar reflection is proportional to the sixth power of a raindrop's diameter d.

However, that rule only applies when the dimensions of the reflecting object are small relative to the radar wavelength---the reflector is in the Rayleigh region.

I think many weather radars operate at a wavelength of about 10 cm (4 inches). So, reflections from big birds and little birds would not follow the d-to-the-sixth rule. d-to-the-sixth does explain most of the difference between radar reflectivity of raindrops and birds.


garyLambda said...

What is interesting is that until radar was first used along the gulf Coast ornithologists had no idea how many birds were making the non stop flight from Mexico to Texas. Including hummingbirds which during that one hop lose 1/3 of their body weight.

davidotazu said...

It is wrong to feed the public clear-air mode radar images when they are interested only in precipitation data. I understand that potentially the reason these radars are switched to clear-air mode is to utilize such expensive equipment for some useful tasks during dry weather. However, the solution would be to conduct sweeps of the sky alternating between clear-air and precipitation mode. Although the temporal resolution would suffer a bit (by half), the consistency of data would be preserved for users (weather stations, phone apps, etc.) who only want to know about precipitation.