Thread by @RadarAndStuff

So, fun fact: birds and insects show up on radar. Often. As in, pretty much every day. Can we visually delineate between migrating birds and insects on radar? Usually. Here is a bumbling threaded attempt to show some telltale signs of each using the last 24 hours of radar data.

First off, some caveats. We’ll focus on daytime and nighttime signals. At dawn and dusk the airspace basically turns into a veritable thunderdome. All kinds of things are taking flight and landing. Taxa are mixing together. It is chaos. We’ll mostly ignore dawn and dusk.

Except for this loop which shows the Oklahoma City roost of Purple Martins (birds, for you meteorologists listening in) taking flight at dawn. The other stuff is probably primarily insects taking flight into the morning boundary layer. Thunderdome. Enough said. No time for that.

Also, bats exist, show up on radar, and are neither birds nor insects. For areas with big bat pops, the nighttime airspace gets messy. Here’s a loop of bat colonies taking flight last night along the i-35 corridor in TX in the gaps between the rain showers. We won’t mess with TX.

Also also, we’ll just stick to large-scale, broad-front migration. Goofy things happen when you look at smaller scales. Like this case of spring raptor migration through Texas, where the birds are selectively soaring on thermal updraft streets. Not dealing with that either.

So with that out of the way, a good oversimplified rule of thumb is that large-scale migration signals on radar will be insects during the day and birds at night. This image from yesterday was from around 4:30pm local (Central) time. Most birds would have been firmly on land.

Reflectivity factor (the product we’re looking at) is proportional to the size of the animal to the 6th power, so big things (birds) typically dominate if they're present. Relatively few birds migrate over large regional spatial scales during the day, so bugs are usually visible.

Let’s focus on the Hastings, Nebraska radar site since it had some strong migration signals going on yesterday. It's the one circled in black, for those who forgot that Nebraska is a place.

Just kidding, Nebraska! You know we appreciate you!!

(not actually kidding)

Here’s an 8-hour loop from Hastings spanning sunset last night. You can see the diurnal migrants drop out of the airspace at dusk, immediately followed by the ascent of nocturnal migrants. This “changing of the guards” is a pretty common radar signature during migration seasons.

Most radar apps will have the option to view the Differential Reflectivity (ZDR) product. Insect migration will almost be entirely over 5dB (the very upper limits of the color scale), while birds will have a mix of values across much of the range.

ZDR from this same time span:

In fact, most of these dual-polarization radar products (which are related to the size and shape of stuff in the airspace) will have bird/insect differences. Here is the Differential Phase product loop:

And Correlation Coefficient, which is typically much higher for insects than birds. It's because insects are more friendly, in the electromagnetic signal-statistics sense. Don't quote me on that.

Back to the image from yesterday. Another hint of insect migration is the presence of “bow tie” features in the reflectivity factor product. Some of the radars have clearer bow ties than others, but most have the rough pattern.

These bow tie patterns are due to the elongated shape of most insects. When viewed from head- or tail-on aspect angles, insect have tiny cross-sections resulting in weak radar signals. When viewed from the side, insects have more reflecting area, resulting in stronger signals.

So these bow tie features also indicate the large-scale flight heading of the insect assemblage. Stronger bow tie signatures indicate greater directional cohesiveness in headings (more uniform flight behavior).

Zooming in on the Dodge City radar, the bow tie is clear in Reflectivity Factor, as well as the Differential Reflectivity product (which is related to the shape/aspect of the animals). The insects are generally aligned on a NNW to SSE axis.

The stuff in the lower right-hand corner of the image is precipitation. It turns out that weather radars are also good at detecting weather, which I think is a novel and promising application of the technology.

So those are a few things to look out for. Obviously other non-radar information (phenology, time of day, temperature) is extremely useful as well. Luckily there are a bunch of clever folks across the world working on these topics, and we're getting better every year.

While we're here, might as well add a shameless plug for a new project we've started working on in the OU @Aero_Eco group. This is SkyNet, a wind-oriented insect collection funnel designed to sample aerial insect assemblages. It still needs a mesh skin to cover its skeleton...

but once it is finished it'll be installed up here, around 200 feet above the OU campus, to investigate variability in species composition and phenology. The ultimate goal is to develop better radar classification algorithms for detecting and monitoring insect populations.

So if you want to come work on the SkyNet program to help make machines smarter, keep your eyes peeled for some grad position announcements from @Aero_Eco!