What the radar actually does
A network of about 160 radars run by three agencies, each one listening far more than it talks, and each measuring two things that are not what you think they are.
One machine, two measurements
A Weather Surveillance Radar, 1988, DopplerThe doppler weather radar the National Weather Service runs, roughly 160 of them across the country. When somebody says "the radar", this is the one they mean. transmits a pulse of microwave energy and then listens for what comes back. From the returned signal it derives two quantities, and confusing them is the commonest beginner error in the whole subject.
- Reflectivity
- How much energy came back. Big or numerous targets return more. It is the familiar green-to-red precipitation picture, and it tells you what is in the beam, not what is reaching the ground.
- Velocity
- How fast the targets in the beam are moving toward or away from the radar, from the frequency shift in the returned signal. Only motion along the beam. Lesson 6 is about what that restriction does to everything you think you are seeing.
The radio band around 10 cm wavelength that the WSR-88D uses. Long enough to see through heavy precipitation without being absorbed, at the cost of a wider beam than a shorter-wavelength radar would have., and why the wavelength is the design decision
The WSR-88D is an S-band radar, wavelength around 10 centimetres. That single choice explains a great deal about what the network is good and bad at.
- Long wavelengths penetrate heavy precipitation without being absorbed on the way in. A shorter-wavelength radar can be attenuated to blindness by the very storm you most want to see through, and S-band largely is not.
- Long wavelengths resolve less finely for a given dish size. The beam is about 0.95 degrees wide, which is the number lesson 2 is entirely about.
- Hail bigger than about the wavelength scatters strangely, which is why the hail estimates in lesson 8 lose skill exactly where you most want them.
The other radars you may encounter are different on purpose. Airport terminal Doppler radars are C-band, around 5 centimetres, with finer range resolution over a small area. Research radars go shorter still. None of that changes what a spotter does, and it explains why two radars can disagree.
Who owns the network
This one surprises people. The Next Generation Weather RadarThe programme that developed and deployed the radar network in the 1990s; WSR-88D is the radar itself. The two names are used almost interchangeably. network is not the National Weather Service's alone. It is operated jointly by three agencies: the NWS runs most of it, the Department of Defense runs sites largely tied to military installations, and the FAA runs a smaller number. Together it is about 160 operational radars, with the NWS share around 120.
The practical consequence for you is small but real. Your nearest radar may be a military or FAA site, its outages may be scheduled by somebody other than your forecast office, and the office may lose coverage for reasons nothing to do with the weather. Lesson 2 is about the holes in coverage that exist when everything is working.
Two ranges, not one
Which is why a storm can be visible on the reflectivity picture with no velocity data at all. It is not a malfunction, and it is not the app.
Where the network is going
Two things are worth knowing because you will hear both discussed and they are frequently confused.
- The Service Life Extension Program
- A refurbishment programme that ran from 2015 to 2024 and replaced the major components across the fleet: signal processor, transmitter, shelters, pedestal and backup generator. It is why a network designed in the 1980s is expected to remain operational into the middle 2030s.
- Phased array, and what comes next
- Research and demonstration, not deployment. A decision on the successor to NEXRAD is expected around 2030 with fielding much later. Anybody telling you phased array radar is in operational use across the country is a decade or more ahead of the facts.
A storm shows clearly on reflectivity at 350 km from the radar, but the velocity product shows nothing there. What is happening?
The two products are derived from the same returned pulse but need different things from it.Reflectivity is usable to roughly 460 km and velocity to roughly 300 km, because velocity needs unambiguous phase information that is lost at greater range. A storm visible on one and absent from the other at long range is the system working as designed.