The NEXRAD Dual-Polarization Upgrade, the Hundred and Sixty Million Dollar Radar Overhaul That Changed What Pilots See in the Cockpit, and the New Data Hidden Behind a Basic Reflectivity Return

The $160 million dual-polarization upgrade to the U.S. NEXRAD network, completed between 2011 and 2013, fundamentally changed what ground-based weather radar can measure. For the first time, the system can reliably distinguish rain from hail, identify tornado debris signatures in real time, and filter non-meteorological clutter from the data underlying every weather product pilots brief from.

What Is NEXRAD and How Does It Work?

NEXRAD stands for Next Generation Radar. The hardware at each site is designated WSR-88D - Weather Surveillance Radar, 1988, Doppler. The network spans the continental United States, Alaska, Hawaii, Puerto Rico, and overseas installations at American military and government facilities, totaling 160 systems. The National Weather Service operates the majority; the Air Force and FAA operate the rest.

Each installation houses a dish antenna approximately 28 feet in diameter inside a fiberglass radome. The transmitter fires pulses at S-band - a 10-centimeter wavelength - chosen specifically because it penetrates heavy rain without being absorbed or scattered, allowing the beam to reach through a storm and return data from the far side. The system completes a full volume scan every 4 to 6 minutes, sweeping multiple elevation angles and building a three-dimensional precipitation picture out to roughly 230 nautical miles.

What Single-Polarization Radar Could and Couldn’t Do

From the late 1980s through the early 2000s, all 160 systems worked identically. Each fired a single horizontally polarized pulse and measured the energy that scattered back. That return produced two products: reflectivity (measured in dBZ, indicating precipitation intensity and location) and Doppler velocity (the actual motion of precipitation).

Doppler velocity was genuinely transformative. It’s how meteorologists identify mesocyclone rotation before a tornado forms and detect low-level wind shear near airports. Reflectivity gave operators location and rough intensity - a cell at 60 dBZ indicated a severe storm.

But reflectivity only measures how much energy bounced back. It tells you nothing about the shape of what’s doing the bouncing.

Why Drop Shape Changes Everything

Raindrops are not spherical. Aerodynamic drag flattens them as they fall - small drops remain nearly round, medium drops resemble M&Ms, and the largest drops flatten into a shape closer to a hamburger bun, significantly wider than tall.

Hailstones tumble as they fall. Their irregular shapes rotate and wobble, presenting no consistent orientation to a radar beam. The return from a hailstone is essentially random in its polarization.

Snowflakes, dry snow, wet snow, aggregated clumps, hexagonal ice crystals, and dendrites each produce distinct radar signatures. Add insects, birds, military chaff, dust, and tornado-lofted debris - a single-polarization radar sees all of it as reflectivity. You could make probabilistic judgments, but not definitive ones.

How Dual-Polarization Works

The upgraded radars transmit two pulses simultaneously: one horizontally polarized, one vertically polarized. The antenna fires both and receives both returns independently. Three new measurement products result.

Differential Reflectivity (Zdr) is the ratio of horizontal to vertical return. A large, flattened raindrop produces a high positive Zdr - it scatters far more energy horizontally than vertically. A tumbling hailstone, with no consistent orientation, produces a Zdr near zero. A sphere produces a Zdr near zero. A column-shaped ice crystal can produce a negative Zdr. That single ratio separates categories of precipitation that were previously indistinguishable.

Correlation Coefficient (CC, or rho-hv) measures the consistency of returns between the horizontal and vertical channels from pulse to pulse. Uniform rainfall produces very consistent returns - the correlation is typically above 0.97. Mixed precipitation (hail mixed with rain, ice pellets mixed with sleet) drops the correlation toward 0.90 or below. Non-meteorological targets like birds or tornado debris can drop it below 0.80, sometimes much lower.

Specific Differential Phase (Kdp) measures phase shift rather than amplitude. It provides an accurate rainfall rate estimate even in heavy precipitation, where the standard reflectivity relationship saturates and becomes unreliable.

The Hydrometeor Classification Algorithm

These three measurements feed the Hydrometeor Classification Algorithm developed by the National Weather Service. The algorithm takes Zdr, CC, Kdp, and basic reflectivity, runs them through a logic system built on decades of field research, and assigns each radar pixel a precipitation type: heavy rain, light rain, rain-hail mix, large hail, graupel, wet snow, dry snow, ice crystals, biological targets, or non-meteorological echoes.

The algorithm isn’t perfect - nothing in meteorology is. But the improvement over probabilistic guessing is significant, and in specific applications it’s transformative.

What Dual-Polarization Means for Hail Detection

Under single-polarization, forecasters could infer likely hail from high reflectivity. Under dual-polarization, the hail signature is specific: low Zdr (hailstones tumble with no consistent orientation), elevated reflectivity (hailstones are large), and lower correlation coefficient (mixed hail sizes create chaotic returns). Three converging signals are more reliable than one.

There is also a phenomenon called the three-body scatter spike. When large hail exists at altitude in a storm, some radar energy hits the hail, bounces to the ground, reflects back up to the hail, and returns to the radar from a direction that makes the system report a second echo downrange from the actual storm. In single-polarization data, this artifact was difficult to interpret consistently. In dual-polarization data, the spike is unmistakable - the CC in that secondary echo drops near zero - making it one of the most reliable indicators of large hail at altitude, sometimes visible before the hail reaches the surface.

For pilots on instrument flight plans above the freezing level, large hail at cruise altitude is an airframe threat regardless of what the surface weather looks like.

The Tornado Debris Signature

When a tornado is on the ground throwing debris into the air, that debris appears in the radar as a region of very high reflectivity combined with correlation coefficient values that can drop below 0.70. Pieces of buildings, vehicles, trees, and farm equipment are random in size, shape, and orientation - the CC can drop near zero in the strongest cases.

Before dual-polarization, meteorologists had to infer from the rotation signature alone whether a tornado was actually on the ground. After the upgrade, when a low-CC return appears collocated with a strong mesocyclone, the inference becomes a confirmation. The National Weather Service updated its warning language to allow confirmed tornado warnings when a debris signature is present, replacing the previous “tornado possible” threshold. Warning lead times improved; false alarm rates dropped.

How Clutter Filtering Affects What Pilots See

During heavy spring bird migration across the central United States, single-polarization radar produced rings of false echoes spreading out from radar sites - biological returns that no precipitation could explain. The Hydrometeor Classification Algorithm flags those returns as biological targets. The precipitation products flowing out of the network are cleaner than they were a decade ago.

This matters directly to flight planning. A cleaner base product means more accurate SIGMETs, more accurate convective outlooks, and better Terminal Aerodrome Forecasts (TAFs). The forecasters writing those products are working from better raw data than they had in 2010.

What Pilots Actually See in the Cockpit

Most pilots never see the dual-polarization products directly. When you open an EFB and look at the NEXRAD composite, you’re looking at reflectivity - a single variable, color-coded by intensity. The differential reflectivity, correlation coefficient, and specific differential phase are not displayed in most cockpit applications.

The improvement is real, but it shows up indirectly: better clutter filtering, more accurate underlying data, and more reliable forecast products. The interface between improved radar data and the pilot position hasn’t kept pace with the improvement in the data itself.

The full dual-pol product suite is publicly available. The National Weather Service’s radar viewer at weather.gov displays all dual-polarization products. Radarscope, widely used by storm spotters and meteorologists, provides full access to every dual-pol product for every NEXRAD site. Time spent with those products on the ground - examining a thunderstorm, switching between reflectivity and correlation coefficient, watching how an actively growing cell’s updraft tower appears as a low-correlation chaos signature inside a strong return - changes how you interpret what you see in the cockpit.

The Research Behind the Upgrade

The science predates the upgrade by decades. The National Severe Storms Laboratory in Norman, Oklahoma - part of NOAA’s National Weather Service - had been operating experimental polarimetric radars since the 1980s. Dave Zrnic and Alexander Ryzhkov at that laboratory produced foundational research alongside a cohort of scientists who spent careers on this problem. The experimental prototype radar at the Norman site, designated KOUN, served as the testbed for validating operational algorithms before the nationwide rollout.

The gap between the research proving dual-polarization worked and its operational implementation across the full network was approximately 20 years - a product of federal procurement cycles, infrastructure complexity, budget processes, and the engineering challenge of retrofitting existing hardware at 160 sites.

What Comes Next: SENSR and Phased Array Radar

NOAA is developing the next generation of ground-based weather radar under a program called the Spectrum Efficient National Surveillance Radar (SENSR). It’s built around phased array technology - electronically steered beams that eliminate the need for a physically rotating dish, enabling update rates far faster than the current 4-to-6-minute scan cycle. In rapidly evolving convection, that time window matters: a storm can jump from 60 dBZ to 70 dBZ in minutes.

SENSR remains years from operational deployment. The dual-polarization network is what exists now, and it represents a genuine, documented improvement in the quality of data underlying every weather product pilots use.

Key Takeaways

• The NEXRAD dual-polarization upgrade, completed 2011–2013 at a cost of $160 million, added vertical polarization to all 160 U.S. weather radar sites, changing the fundamental physics of what the radar measures.
• Three new products - differential reflectivity (Zdr), correlation coefficient (CC), and specific differential phase (Kdp) - allow the system to distinguish rain from hail, identify tornado debris, and filter biological clutter.
• A low CC collocated with a mesocyclone is now a confirmed tornado indicator; the National Weather Service updated its warning language accordingly, improving lead times and reducing false alarms.
• Most pilots see only reflectivity in EFB applications; the full dual-pol suite is available at weather.gov and in Radarscope.
• The next-generation replacement, SENSR, will use phased array technology for faster scan rates but remains years from operational deployment.