10.1 Radar meteorology

Principles of radar meteorology

Radar meteorology uses electromagnetic waves to detect and analyze atmospheric phenomena. These principles form the backbone of weather observation and forecasting, and understanding them is essential for interpreting radar data accurately.

Electromagnetic wave propagation

Electromagnetic waves travel at the speed of light in a vacuum, but atmospheric conditions alter their propagation through refraction and attenuation. Weather radars operate at wavelengths typically ranging from 1 to 10 cm, which correspond to three common bands:

• S-band (~10 cm): Best penetration through heavy precipitation, but lower spatial resolution. Used by the WSR-88D (NEXRAD) network.
• C-band (~5 cm): A middle ground between penetration and resolution. Common in European radar networks.
• X-band (~3 cm): Highest resolution, but attenuates significantly in heavy rain. Often used in portable or gap-filling radars.

The trade-off is straightforward: longer wavelengths see through rain better, while shorter wavelengths resolve finer details.

Radar equation fundamentals

The radar equation relates the power received back at the radar to the transmitted power and the properties of the target:

$P_r = \frac{P_t G^2 \lambda^2 \sigma}{(4\pi)^3 r^4}$

where:

• $P_r$ = received power
• $P_t$ = transmitted power
• $G$ = antenna gain
• $\lambda$ = wavelength
• $\sigma$ = radar cross-section of the target
• $r$ = range (distance) to the target

The $r^4$ term in the denominator is key: it accounts for the signal traveling out to the target and then back to the radar, with inverse-square losses in each direction. This means received power drops off very quickly with distance.

Pulse repetition frequency

Pulse repetition frequency (PRF) is how often the radar transmits pulses. Typical weather radar PRF values range from 300 to 1200 Hz.

PRF creates a fundamental trade-off:

• Higher PRF → more pulses per second → better velocity sampling, but the radar "listens" for a shorter time between pulses, which reduces the maximum unambiguous range (the farthest distance it can measure without confusion).
• Lower PRF → longer listening window → greater unambiguous range, but fewer velocity samples and a lower Nyquist velocity limit.

You can't optimize both range and velocity simultaneously with a single PRF. This is called the Doppler dilemma, and it drives many of the techniques discussed later (like dual-PRF methods).

Radar hardware components

Weather radar systems consist of several interconnected hardware elements. Each component plays a specific role in transmitting, receiving, and processing radar signals.

Transmitter and receiver systems

The transmitter generates high-power microwave pulses that probe the atmosphere. Two common power sources are:

• Klystron tubes: High power, stable frequency, but expensive and bulky. Preferred when precise frequency control matters (e.g., Doppler applications).
• Magnetron tubes: Cheaper and more compact, but less frequency-stable. Common in smaller or older radar systems.

The receiver detects the extremely weak return signals from atmospheric targets. Low-noise amplifiers boost the signal, and mixers convert the received microwave frequency down to a lower intermediate frequency that's easier to process digitally.

Antenna types and designs

• Parabolic dish antennas focus the radar beam into a narrow cone for directional transmission. The dish physically rotates to scan the atmosphere. Most operational weather radars use this design.
• Phased array antennas use many small radiating elements whose signals are combined electronically. By adjusting the phase of each element, the beam can be steered without moving parts, allowing much faster scanning.
• Antenna gain determines how tightly the beam is focused. Higher gain means better sensitivity and angular resolution.
• Dual-polarization antennas transmit and receive in both horizontal and vertical orientations simultaneously, enabling the polarimetric measurements covered below.

Signal processing units

Signal processors convert the raw analog return signal into usable digital data. Their main tasks include:

1. Analog-to-digital conversion of the received signal
2. Filtering to remove noise and ground clutter
3. Moment estimation: calculating reflectivity, radial velocity, and spectrum width from the processed returns
4. Generating radar products and formatting data for display and distribution

Radar reflectivity

Radar reflectivity measures how much transmitted power is scattered back to the radar. It provides information about the size, concentration, and type of hydrometeors (rain, snow, hail, etc.) in the atmosphere.

Rayleigh scattering theory

Rayleigh scattering applies when the radar wavelength is much larger than the target particle. For weather radar, this condition holds for most raindrops and ice crystals.

The critical result: scattering intensity is proportional to the sixth power of the particle diameter ($D^6$). This means a raindrop twice as large scatters $2^6 = 64$ times more power. A few large drops dominate the return signal far more than many small ones.

Rayleigh scattering breaks down when particles become comparable in size to the wavelength, such as large hail observed by shorter-wavelength radars. In those cases, Mie scattering theory is needed.

Reflectivity factor Z

The reflectivity factor $Z$ sums the sixth powers of all particle diameters in a unit volume of air:

$Z = \sum_{i=1}^{n} D_i^6$

Units are $\text{mm}^6/\text{m}^3$. Because $Z$ depends on $D^6$, it's extremely sensitive to the largest particles in the sample. This is why hail produces such high reflectivity values even when relatively few hailstones are present.

dBZ scale and interpretation

Because $Z$ spans many orders of magnitude, it's expressed on a logarithmic scale:

$dBZ = 10 \log_{10}(Z)$

where $Z$ is in $\text{mm}^6/\text{m}^3$.

Approximate interpretation:

Doppler radar principles

Doppler radar measures the motion of targets relative to the radar by exploiting the Doppler effect. This capability is essential for detecting wind patterns, storm rotation, and wind shear.

Doppler effect in meteorology

When a target moves relative to the radar, the frequency of the returned signal shifts:

$\Delta f = \frac{2v_r}{\lambda}$

where $\Delta f$ is the frequency shift, $v_r$ is the radial velocity of the target, and $\lambda$ is the radar wavelength. The factor of 2 accounts for the round trip (out and back).

• Motion toward the radar → positive frequency shift (higher frequency)
• Motion away from the radar → negative frequency shift (lower frequency)

Only the radial component of velocity (along the beam direction) is measured. A target moving perpendicular to the beam produces zero Doppler shift, even if it's moving fast.

Radial velocity measurements

Radial velocity is derived from the phase shift between successive returned pulses, not from measuring frequency directly. By convention:

• Positive values = motion toward the radar
• Negative values = motion away from the radar

This data reveals wind patterns, mesocyclone rotation in supercells, low-level wind shear near airports, and the overall motion of storms.

Nyquist velocity and aliasing

The Nyquist velocity is the maximum unambiguous velocity the radar can measure, set by the PRF and wavelength:

$v_{max} = \pm \frac{\lambda \cdot PRF}{4}$

Any true velocity exceeding $v_{max}$ gets "folded" or "aliased" into the wrong part of the velocity spectrum. For example, if $v_{max} = 25 \text{ m/s}$ and the true velocity is 30 m/s toward the radar, it will appear as 20 m/s away from the radar. Aliased velocities show up as abrupt, unrealistic color changes on velocity displays.

Dual-PRF techniques and dealiasing algorithms are used to extend the effective Nyquist velocity.

Dual-polarization radar

Dual-polarization (dual-pol) radar transmits and receives waves in both horizontal and vertical orientations. Because different hydrometeors have different shapes and orientations, the two polarizations interact with them differently. This provides a much richer picture than reflectivity alone.

Polarimetric variables

Four key variables come from dual-pol measurements:

• Differential reflectivity ($Z_{DR}$): The ratio of horizontal to vertical reflectivity (in dB). Positive values indicate particles wider than they are tall (like large raindrops flattened by air resistance). Values near zero suggest spherical or tumbling particles.
• Correlation coefficient ($\rho_{HV}$): How well the horizontal and vertical signals correlate. Values near 1.0 indicate uniform particle types. Values dropping below ~0.90 suggest a mix of particle types, melting, or non-meteorological targets (debris, birds).
• Specific differential phase ($K_{DP}$): The rate at which the phase difference between horizontal and vertical waves accumulates with range. Proportional to liquid water content and insensitive to attenuation, making it very useful for rain rate estimation.
• Linear depolarization ratio (LDR): Measures how much a target depolarizes the signal. Useful for identifying the melting layer and wet ice, but requires simultaneous transmit/receive capability.

Hydrometeor classification

Dual-pol radars combine multiple polarimetric variables to classify what type of particle the radar is seeing. Common categories include rain, dry snow, wet snow, hail, graupel, ice crystals, and mixed-phase precipitation.

Classification typically uses fuzzy logic algorithms that assign membership scores across categories based on the observed variable values. This improves understanding of precipitation processes and helps forecasters distinguish between, say, heavy rain and a rain/hail mixture.

Quantitative precipitation estimation

Dual-pol data significantly improves rainfall rate estimates compared to reflectivity alone. Three common approaches:

• R(Z): Traditional reflectivity-only relationship. Simple but sensitive to drop size distribution variability and hail contamination.
• R(Z, $Z_{DR}$): Adds differential reflectivity to account for drop size, reducing errors from varying drop size distributions.
• R($K_{DP}$): Uses specific differential phase, which is largely immune to attenuation, hail contamination, and partial beam blockage. Most accurate in heavy rain.

Operational systems often blend these approaches depending on conditions.

Weather radar products

Radar data is processed into various products that forecasters and researchers use to visualize and analyze atmospheric conditions.

Base reflectivity and velocity

Base reflectivity displays echo intensity on a color-coded map at a single elevation angle. Multiple elevation angles are available, giving you slices through the atmosphere at different heights. This is the most fundamental radar product for assessing precipitation location and intensity.

Base velocity shows the radial motion of targets at a given elevation angle. Colors typically indicate motion toward (green) or away from (red) the radar. Velocity data is essential for identifying rotation, convergence, and wind shear.

Composite reflectivity

Composite reflectivity takes data from all elevation scans and displays the maximum reflectivity value found in each vertical column. This gives a quick overview of where the most intense precipitation exists across the entire radar volume, regardless of altitude. It's useful for rapid assessment of storm intensity but can sometimes overstate surface precipitation if high reflectivity exists only aloft.

Storm relative motion

Storm relative motion (SRM) subtracts the storm's own movement vector from the base velocity data. This isolates the winds within the storm relative to its motion.

SRM is particularly valuable for detecting:

• Mesocyclones: Persistent rotation signatures in supercell thunderstorms
• Tornado vortex signatures: Tight couplets of inbound/outbound velocities
• Inflow/outflow regions: Where air is feeding into or exiting the storm

Without removing storm motion, these internal circulation features can be masked by the storm's overall translation.

Radar data interpretation

Interpreting radar data requires combining knowledge of radar physics with atmospheric dynamics. Pattern recognition is a core skill for forecasters.

Precipitation patterns

• Stratiform precipitation appears as widespread, relatively uniform reflectivity. A bright band (a thin layer of enhanced reflectivity) often appears at the melting level, where snowflakes begin to melt and briefly become coated in water, dramatically increasing their radar cross-section.
• Convective precipitation shows up as isolated, intense cells with sharp reflectivity gradients. These are associated with thunderstorms and strong vertical motions.
• Banded structures often indicate frontal boundaries, deformation zones, or lake-effect snow bands.

Severe storm signatures

Several radar signatures are associated with severe weather:

• Hook echo: A curved appendage on the rear flank of a supercell, often associated with tornado development.
• Bounded weak echo region (BWER): An area of reduced reflectivity bounded above by strong echoes, indicating a powerful updraft lofting precipitation upward.
• V-notch (or enhanced-V): A V-shaped notch in the storm top, visible in satellite or upper-level radar data, indicating strong divergence aloft from an intense updraft.
• Three-body scatter spike (TBSS): A spike of reflectivity extending radially away from a high-reflectivity core, caused by energy bouncing between large hailstones and the ground. A strong indicator of large hail.

Non-meteorological echoes

Not everything on radar is weather:

• Ground clutter: Stationary, high-reflectivity returns near the radar from terrain, buildings, or towers. Usually filtered out, but can leak through.
• Anomalous propagation (AP): Temperature inversions or ducting bend the radar beam toward the ground, producing false echoes at long range.
• Biological echoes: Birds, insects, and bats are often visible in clear-air mode, especially during migration events or at dawn/dusk.
• Chaff: Thin metallic strips released as radar countermeasures. Produces distinctive, slowly dispersing echo patterns.

Dual-pol variables (especially $\rho_{HV}$) are very effective at distinguishing meteorological from non-meteorological echoes.

Advanced radar techniques

Newer technologies and methods continue to push the capabilities of radar meteorology in terms of resolution, coverage, and data quality.

Phased array technology

Phased array radars steer their beams electronically rather than by rotating a dish. This allows:

• Full volume scans in under 60 seconds (compared to ~5 minutes for conventional radars)
• Adaptive scanning that focuses on rapidly evolving storms while still monitoring the broader area
• Simultaneous multi-function use, such as weather surveillance and aircraft tracking on the same system

The faster update rate is especially valuable for tracking tornadoes, microbursts, and other phenomena that evolve on timescales shorter than conventional scan intervals.

Multi-radar multi-sensor systems

MRMS (Multi-Radar Multi-Sensor) is the operational example in the U.S. These systems:

• Mosaic data from overlapping radars to fill coverage gaps and reduce range-dependent biases
• Integrate non-radar data sources: satellite imagery, surface observations, lightning detection networks, and numerical model output
• Produce seamless, high-resolution products over large geographic areas
• Provide automated severe weather and precipitation algorithms that no single radar could match

Radar data assimilation

Data assimilation incorporates radar observations into numerical weather prediction (NWP) models to improve forecast initial conditions. This is especially impactful for short-range (0–6 hour) forecasts of convective weather.

Common techniques include:

• 3D-Var and 4D-Var: Variational methods that find the model state best fitting both the prior forecast and the observations
• Ensemble Kalman filter (EnKF): Uses an ensemble of model runs to estimate background error covariances, making it well-suited for flow-dependent errors

Challenges remain significant: radar data has non-Gaussian error distributions, the relationship between radar observables and model variables is nonlinear, and the sheer volume of radar data requires careful thinning or superobbing strategies.

Limitations and challenges

Every radar measurement comes with caveats. Understanding these limitations is just as important as understanding the data itself.

Beam blockage and attenuation

• Beam blockage occurs when terrain, buildings, or other obstacles partially or completely obstruct the radar beam. Affected areas show artificially low reflectivity or no data at all.
• Attenuation is the loss of signal strength as the beam passes through heavy precipitation. This is a bigger problem at shorter wavelengths (X-band and C-band) than at S-band.
• Mitigation strategies include dual-pol attenuation correction (using $K_{DP}$), multi-radar compositing, and careful siting of radar installations.

Range folding and velocity ambiguity

• Range folding happens when echoes from beyond the maximum unambiguous range are incorrectly placed at a closer range. These "second-trip echoes" can contaminate the display, especially in widespread convective environments.
• Velocity ambiguity (aliasing) occurs when true velocities exceed the Nyquist limit, as discussed earlier.
• Both problems stem from the Doppler dilemma. Dual-PRF techniques, staggered PRT (pulse repetition time) methods, and post-processing algorithms help resolve these issues but don't eliminate them entirely.

Radar coverage gaps

The Earth's curvature means that even with a network of radars, the beam overshoots low-level phenomena at long range. At 200 km from the radar, the lowest beam may be over 2 km above the surface.

• Mountainous terrain compounds this with beam blockage.
• Gaps can cause missed detections of shallow precipitation, low-level rotation, and boundary-layer wind shear.
• Gap-filling radars (often smaller X-band systems), terminal Doppler weather radars at airports, and integration with surface observation networks help address these limitations.

Applications in meteorology

Radar data supports a wide range of operational and research applications across meteorology and hydrology.

Nowcasting and severe weather detection

Short-term forecasting (0–6 hours), or nowcasting, relies heavily on radar. Automated algorithms detect signatures of severe weather in near real-time:

• Mesocyclone detection algorithms flag persistent rotation
• Tornado vortex signatures trigger tornado warnings
• Hail detection algorithms use reflectivity profiles and dual-pol data
• Warning decision support systems integrate these detections to help forecasters issue timely alerts

Dual-pol has notably improved the detection of tornado debris (via low $\rho_{HV}$ in a tornado debris signature) and discrimination between rain and hail.

Rainfall estimation and hydrology

Radar provides spatially continuous rainfall estimates at high temporal resolution (every 5–10 minutes), which rain gauges alone cannot match. Applications include:

• Flash flood forecasting and monitoring
• Input to hydrologic models for river stage prediction
• Water resource management and reservoir operations
• Merging radar estimates with gauge data for bias correction produces the most accurate rainfall fields

Wind field analysis

Doppler velocity data reveals atmospheric motion across a range of scales:

• Airport weather: Detection of wind shear, microbursts, and gust fronts critical for aviation safety
• VAD (Velocity Azimuth Display): A technique that derives vertical wind profiles from a single radar by analyzing velocity data at a constant range across all azimuths
• Dual-Doppler analysis: When two radars observe the same storm from different angles, their radial velocities can be combined to retrieve the full 3D wind field within the storm. This technique is a cornerstone of severe storm research.