radar basics and estimating precipitation jon w. zeitler science and operations officer national...
TRANSCRIPT
Radar Basics and Estimating Precipitation Radar Basics and Estimating Precipitation
Jon W. ZeitlerJon W. Zeitler
Science and Operations OfficerNational Weather Service
Austin/San Antonio Forecast Office
Science and Operations OfficerNational Weather Service
Austin/San Antonio Forecast Office
Radar Beam BasicsRadar Beam Basics
As pulse volumes within the radar beam encounter targets, energy will be scattered in all directions. A very small portion of the intercepted energy will be backscattered toward the radar. The degree or amount of backscatter is determined by target:
size (radar cross section) shape (round, oblate, flat, etc.)
state (liquid, frozen, mixed, dry, wet) concentration (number of particles per unit volume)
We are concerned with two types of scattering, Rayleigh and non-Rayleigh. Rayleigh scattering occurs with targets whose diameter (D) is much smaller (D < /16) than the radar wavelength. The WSR-88D's wavelength is approximately 10.7 cm, so Rayleigh scattering occurs with targets whose diameters are less than or equal to about 7 mm or ~0.4 inch. Raindrops seldom exceed 7 mm so all liquid drops are Rayleigh scatters.
Potential problem: Nearly all hailstones are non-Rayleigh scatterers due to their larger diameters.
As pulse volumes within the radar beam encounter targets, energy will be scattered in all directions. A very small portion of the intercepted energy will be backscattered toward the radar. The degree or amount of backscatter is determined by target:
size (radar cross section) shape (round, oblate, flat, etc.)
state (liquid, frozen, mixed, dry, wet) concentration (number of particles per unit volume)
We are concerned with two types of scattering, Rayleigh and non-Rayleigh. Rayleigh scattering occurs with targets whose diameter (D) is much smaller (D < /16) than the radar wavelength. The WSR-88D's wavelength is approximately 10.7 cm, so Rayleigh scattering occurs with targets whose diameters are less than or equal to about 7 mm or ~0.4 inch. Raindrops seldom exceed 7 mm so all liquid drops are Rayleigh scatters.
Potential problem: Nearly all hailstones are non-Rayleigh scatterers due to their larger diameters.
Energy ScatteringEnergy Scattering
Probert-Jones Radar EquationProbert-Jones Radar Equation
Simplified Radar EquationSimplified Radar Equation
Since we technically don't know the drop-size distribution or physical makeup of all targets within a sample volume, radar meteorologists oftentimes refer to radar reflectivity as equivalent reflectivity, Ze.
The assumption is that all backscattered energy is coming from liquid targets whose diameters meet the Rayleigh approximation. Obviously, this assumption is invalid in those cases when large, water-coated hailstones are present in a sample volume. Hence, the term equivalent reflectivity instead of actual reflectivity is more valid.
Since we technically don't know the drop-size distribution or physical makeup of all targets within a sample volume, radar meteorologists oftentimes refer to radar reflectivity as equivalent reflectivity, Ze.
The assumption is that all backscattered energy is coming from liquid targets whose diameters meet the Rayleigh approximation. Obviously, this assumption is invalid in those cases when large, water-coated hailstones are present in a sample volume. Hence, the term equivalent reflectivity instead of actual reflectivity is more valid.
Equivalent Reflectivity (Ze)Equivalent Reflectivity (Ze)
(Equation 5) Reflectivity (Z) vs.Decibels of Reflrectivity (dBZ)
Reflectivity (Z) vs.Decibels of Reflrectivity (dBZ)
dBZ = 10log10ZdBZ = 10log10Z
Beam-FillingBeam-Filling
Sending vs. ListeningSending vs. Listening
Sending vs. ListeningSending vs. Listening
99.843% of the time the WSR-88D is listening for signal returns. 99.843% of the time the WSR-88D is listening for signal returns.
A low PRF is desirable for target range and power, while a high PRF is desirable for target velocity. The inability to satisfy both needs with a single PRF is known as the Doppler Dilemma. The Doppler Dilemma is addressed by the WSR-88D with algorithms.
A low PRF is desirable for target range and power, while a high PRF is desirable for target velocity. The inability to satisfy both needs with a single PRF is known as the Doppler Dilemma. The Doppler Dilemma is addressed by the WSR-88D with algorithms.
The Doppler DilemnaThe Doppler Dilemna
Range FoldingRange Folding
Subrefraction: dry adiabatic, moisture increases with height. In addition to underestimated echo heights, this phenomenon tends to reduce ground clutter in the lowest elevation cuts.
Superrefraction: temperature inversion. In addition to overestimated echo heights, increases ground clutter in the lowest elevation cuts and is the cause of what we normally refer to as anomalous propagation or AP echoes.
Subrefraction: dry adiabatic, moisture increases with height. In addition to underestimated echo heights, this phenomenon tends to reduce ground clutter in the lowest elevation cuts.
Superrefraction: temperature inversion. In addition to overestimated echo heights, increases ground clutter in the lowest elevation cuts and is the cause of what we normally refer to as anomalous propagation or AP echoes.
The Earth is Round!The Earth is Round!
Each pulse has a volume with dimensions of ~ 500 meters (~ 1500 meters) in length by ~ 1° wide in short pulse (long pulse) mode. This means that two targets along a radial must be at least 250 (750) meters apart for the radar to be able to distinguish and display them as two separate targets (i.e., more than H/2 range separation distance).
Each pulse has a volume with dimensions of ~ 500 meters (~ 1500 meters) in length by ~ 1° wide in short pulse (long pulse) mode. This means that two targets along a radial must be at least 250 (750) meters apart for the radar to be able to distinguish and display them as two separate targets (i.e., more than H/2 range separation distance).
Storms Too Close!Storms Too Close!
Storms or Bats?Storms or Bats?
Strategies to Fix ProblemsStrategies to Fix Problems
Drop Size DistributionDrop Size Distribution
Drop Size DistributionDrop Size Distribution
Rainfall RateRainfall Rate
Rainfall RateRainfall Rate
Rainfall RateRainfall Rate
R(Z) Relationships (Battan 1973)
BREAK!BREAK!
Sends and receives horizontal & vertical polarized radiation
Sends and receives horizontal & vertical polarized radiation
Image courtesy Terry SchuurImage courtesy Terry Schuur
What is Dual Polarimetric Radar?What is Dual Polarimetric Radar?
Hydrometeor:Hydrometeor:
• ShapeShape
• OrientationOrientation
• Dielectric constantDielectric constant
• Distribution of sizesDistribution of sizes
Polarimetric Variables Depend Polarimetric Variables Depend on Several Thingson Several Things
Polarimetric Variables Depend Polarimetric Variables Depend on Several Thingson Several Things
•Rainfall Estimation (vast improvement)Rainfall Estimation (vast improvement)•Bright Band Detection (vast improvement)Bright Band Detection (vast improvement)•Clutter Filtering/Data Quality Improvement Clutter Filtering/Data Quality Improvement (vast improvement)(vast improvement)•Rain/Snow Discrimination (vast improvement)Rain/Snow Discrimination (vast improvement)•Hail Detection (some improvement)Hail Detection (some improvement)•Updraft Location (some improvement)Updraft Location (some improvement)•Tornado Detection (some improvement)Tornado Detection (some improvement)
•Rainfall Estimation (vast improvement)Rainfall Estimation (vast improvement)•Bright Band Detection (vast improvement)Bright Band Detection (vast improvement)•Clutter Filtering/Data Quality Improvement Clutter Filtering/Data Quality Improvement (vast improvement)(vast improvement)•Rain/Snow Discrimination (vast improvement)Rain/Snow Discrimination (vast improvement)•Hail Detection (some improvement)Hail Detection (some improvement)•Updraft Location (some improvement)Updraft Location (some improvement)•Tornado Detection (some improvement)Tornado Detection (some improvement)
Applications of Dual Applications of Dual Polarization RadarPolarization Radar
Applications of Dual Applications of Dual Polarization RadarPolarization Radar
Backscattering:Backscattering:ZZhh - reflectivity factor for horizontal polarization - reflectivity factor for horizontal polarization
ZZDRDR - differential reflectivity - differential reflectivity
||ρρhvhv(0)| - co-polar correlation coefficient(0)| - co-polar correlation coefficient
Propagation - forward scattering:Propagation - forward scattering:ΦΦDPDP - differential phase - differential phase
KKDPDP - specific differential phase (range derivative of - specific differential phase (range derivative of
ΦΦDPDP))
Backscattering:Backscattering:ZZhh - reflectivity factor for horizontal polarization - reflectivity factor for horizontal polarization
ZZDRDR - differential reflectivity - differential reflectivity
||ρρhvhv(0)| - co-polar correlation coefficient(0)| - co-polar correlation coefficient
Propagation - forward scattering:Propagation - forward scattering:ΦΦDPDP - differential phase - differential phase
KKDPDP - specific differential phase (range derivative of - specific differential phase (range derivative of
ΦΦDPDP))
Polarimetric VariablesPolarimetric VariablesPolarimetric VariablesPolarimetric Variables
Shapes of Large Drops in Equilibrium Shapes of Large Drops in Equilibrium
Differential Reflectivity (ZDR)
• Definition: the ratio of the power returns from the horizontal and vertical polarizations
• Units: decibels (dB)
vv
hhDR Z
ZZ 10log10
Simple ZDR Calculation for a Sample of Raindrop Sizes
Simple ZDR Calculation for a Sample of Raindrop Sizes
What does ZDR Mean?
• ZDR > 0 Horizontally-oriented mean profile
• ZDR < 0 Vertically-oriented mean profile
• ZDR ~ 0 Near-spherical mean profile
• ZDR > 0 Horizontally-oriented mean profile
• ZDR < 0 Vertically-oriented mean profile
• ZDR ~ 0 Near-spherical mean profile
Eh
Ev
Eh
Ev
Eh
Ev
-4-4 -3.5-3.5 -3-3 -2.5-2.5 -2-2 -1.5-1.5 -1-1 -0.5-0.5 00 0.50.5 11 1.51.5 22 2.52.5 33 3.53.5 44 4.54.5 55 5.55.5 66
Small (Spherical) <<< RAIN >>> Large (Oblate)
Dry <<< GRAUPEL >>> Wet
Dry (Prolate) <<<<< HAIL >>>>> Melting (Oblate)
Aggregated/Low-Density <<< CRYSTALS >>> Pristine/Well-Oriented
Dry <<< SNOW >>> Wet
GROUND CLUTTER / ANOMALOUS PROPAGATION
BIOLOGICAL SCATTERERS
DEBRIS
CHAFF
Differential Reflectivity (ZDR)Differential Reflectivity (ZDR)
1.1. median liquid drop sizemedian liquid drop size (ZDR↑,median drop diameter↑)
2.2. hail shaftshail shafts (ZDR ~ 0dB or negative coincident with high Zh)
3. areas of large rain drops or liquid-large rain drops or liquid-coated icecoated ice (ZDR ~3-6 dB)
4.4. convective updraftsconvective updrafts (ZDR ~1-5 dB) above 0oC level
5. tornado debris ball
ZDR is a Good Indicator of:ZDR is a Good Indicator of:
•Values are biased towards the larger hydrometeors (D6 dependence)•Tumbling/Random orientation will bias toward 0 ZDR
•Can be noisy if:-Low / Insufficient sampling (low SNR)
- Reduced correlation coefficient (CC)
•Values are biased towards the larger hydrometeors (D6 dependence)•Tumbling/Random orientation will bias toward 0 ZDR
•Can be noisy if:-Low / Insufficient sampling (low SNR)
- Reduced correlation coefficient (CC)
ZDR Limitations (Gotchas)ZDR Limitations (Gotchas)
May 9th tornadic supercell: Intense
ZDR Column
0oC level in-cloud ~17 kft
ρhv
Affected by:
• Hydrometeor types, phases, shapes,
orientations
• Presence of large hail
Correlation Coefficient (ρ hv): A correlation between the reflected horizontal and vertical power returns. It is a good indicator of regions where there is a mixture of precipitation types, such as rain and snow.
ρhv Usage
• Identify hail growth regions in deep moist convection (mixtures of hydrometeors)
• Reduce ground clutter/AP contamination (ρhv very low in these areas)
• Identify giant hail ???
ρhv
Correlation Coefficient Correlation Coefficient ((hvhv))
Reflectivity (ZReflectivity (Zhh))
SNOW~0.85-1.00
CLUTTER~0.5-0.85
CHAFF~0.2-0.5
Giant Hail, Protuberances, Mie Scattering: min Giant Hail, Protuberances, Mie Scattering: min ρρhvhv
ρhv Minimum…in Theory
Differential Phase Shift (ΦDP)
• Definition: the difference in the phase shift between the horizontally and vertically polarized waves
• Units: degrees (o)
VHDP
DP = h – v (h, v ≥ 0) [deg]
The difference in phase between the horizontally-and vertically-polarized pulses at a given range along the propagation path.
- Independent of partial beam blockage, attenuation, absolute radar calibration,
system noise
Differential Phase Shift DP
What Affects Differential Phase?
Forward Propagation has its Advantages
• Immune to partial (< 40%) beam blockage, attenuation, calibration, presence of hail
Gradients Most ImportantGradients Most Important
Specific Differential Phase Shift (KDP)
• Definition: range derivative of the differential phase shift
• Units: degrees per kilometer (o/km)
12
12
2)()(
rrrr
K DPDP
DP
• Provides a good estimate of liquid water in a rain/hail mixture
• Indicates the onset of melting
Specific Differential Phase (KDP): A comparison of the returned phase difference between the horizontal and vertical pulses. This phase difference is caused by the difference in the number of wave cycles (or wavelengths) along the propagation path for horizontal and
vertically polarized waves. This is the range derivative of DP,
typically calculated in 1-5 km increments along the radial.
Specific Differential Phase: KDP
Specific Differential Phase Shift (KDP)
*** Non-meteorological values not shown here because they are removed anywhere CC < 0.90 (or 0.85) ****** Non-meteorological values not shown here because they are removed anywhere CC < 0.90 (or 0.85) ***
-0.5-0.5 00 0.50.5 11 1.51.5 22 2.52.5 33 44 55
Small <<< RAIN >>> Large
Dry <<< GRAUPEL >>> Wet
Dry (Prolate) <<<<< HAIL >>>>> Melting (Oblate)
Dry/Aggregated <<< CRYSTALS >>> Pristine/Well-Oriented
Dry <<< SNOW >>> Wet
Kdp Usage
• To isolate the presence of rain from hail R(Z, Zdr, Kdp) much better than R(Z) Most sensitive to amount of liquid water
• To locate regions of drop shedding, “Kdp columns”• Drops are shed from melting or growing
hailstones near the updraft, forming a Kdp column
• To distinguish between snow/rain• Kdp in wet, heavy snow is almost always larger at
a fixed value of Zh than that observed for rain
KDP Limitations (Gotchas)• KDP values set to “No Data” at CC <
0.90, or 0.85)• Sensitive to non-uniform beam filling• Unreliable at far ranges
• KDP Smoothing techinque:KDP Smoothing techinque:
1. < 40 dBZ, KDP computed at each gate from 12 adjacent gates either side (6.25 km)
2. > 40 dBZ, KDP computed at each gate from 4 adjacent gates either side (2.25 km) to preserve heavy cores
1. < 40 dBZ, KDP computed at each gate from 12 adjacent gates either side (6.25 km)
2. > 40 dBZ, KDP computed at each gate from 4 adjacent gates either side (2.25 km) to preserve heavy cores
Compare Z and KDP fields at each gate
Marginally Severe Supercell
Beam Height ~ 4600 ft AGLBeam Height ~ 4600 ft AGL
ZZZDRZDR
ρHVρHVHCAHCA
5.25” diameter hail5.25” diameter hail
14 May 2003
Correlation Coefficient (CC)
• Definition: how similarly the horizontally and vertically polarized backscattered energy are behaving within a resolution volume for Rayleigh scattering
• Units: none (0-1.00)
2/122/12
*
)0(vvhh
hhvv
HV
SS
SS
ThinkThink Spectrum Width for HydrometeorsSpectrum Width for HydrometeorsTMTMThinkThink Spectrum Width for HydrometeorsSpectrum Width for HydrometeorsTMTM
Sij = An element of the backscatter matrix
Correlation Coefficient Values
•0.96 ≤ CC ≤ 1 Small hydrometeor diversity*
•0.80 ≤ CC < 0.96 Large hydrometeor diversity*
•CC < 0.70 Non-hydrometeors present
•0.96 ≤ CC ≤ 1 Small hydrometeor diversity*
•0.80 ≤ CC < 0.96 Large hydrometeor diversity*
•CC < 0.70 Non-hydrometeors present
* Types, sizes, shapes, orientations, etc.* Types, sizes, shapes, orientations, etc.
Correlation Coefficient (CC)
Non-Meteorological
Regime
Meteorological Regime
Overlap
0.20.2 0.30.3 0.40.4 0.50.5 0.60.6 0.70.7 0.80.8 0.850.85 0.90.9 0.910.91 0.920.92 0.930.93 0.940.94 0.950.95 0.960.96 0.970.97 0.980.98 0.990.99 11
Large <<< RAIN >>> SmallLarge <<< RAIN >>> Small
Wet <<< GRAUPEL >>> DryWet <<< GRAUPEL >>> Dry
Wet / Large <<<<< HAIL >>>>> Dry / SmallWet / Large <<<<< HAIL >>>>> Dry / Small
CRYSTALSCRYSTALS
<<Melting Layer>> Wet <<< SNOW >>> Dry<<Melting Layer>> Wet <<< SNOW >>> Dry
GROUND CLUTTER / ANOMALOUS PROPAGATIONGROUND CLUTTER / ANOMALOUS PROPAGATION
BIOLOGICAL BIOLOGICAL SCATTERERSSCATTERERS
DEBRISDEBRIS
CHAFFCHAFF
What is CC Used for?
• Not-met targets (LOW CC < 0.70)
– Best discriminator
• Melting layer detection (Ring of reduced CC ~ 0.80 – 0.95)
• Giant hail? (LOW CC < 0.70 in the midst of high Z/Low ZDR)
Marginally Severe Supercell
What about the rest?All > 0.97
What about the rest?All > 0.97
InsectsInsectsPrecipPrecip
CC Limitations (Gotchas)
• High error in low signal-to-noise ratios (SNR)
• If low, errors increase in other dual-pol variables
One hour point measurements: Radar estimates vs. gages
R(Z) R(Z, KDP, ZDR)
Polarimetric Rainfall Algorithm vs. Conventional
Polarimetric Rainfall Algorithm vs. Conventional
Bias of radar areal rainfall estimates
Spring hail cases
Cold season stratiform rain
QPE Process in a NutshellStep 1
1. Hybrid scan the variables into Polar, 1 degree azimuth, 250 m bins
Hybrid Hydroclass
QPE Process in a Nutshell
2. Apply an instantaneous Rate: R(Z), R(KDP), and R(Z,ZDR)
But which one is accepted?
ZZR714.0
017.0)(
)(0.44)(882.0
KDPsignKDPR KDP
ZDRZZDRZR67.1770.0
0142.0),(
QPE Process in a Nutshell
3. Assign a variation of 1 of those 3 rates to each bin based on HCA precip type
Based on 43 events (179 hrs) of radar rainfall data
comparisons to a dense network of rain gauges in C. OK
Based on 43 events (179 hrs) of radar rainfall data
comparisons to a dense network of rain gauges in C. OK
Rate Designation TableR (mm/hr) Conditions Echo
Classes
Notcomputed
Nonmeteorological echo (Ground Clutter or Unknown) is classified GC ,UK
0 Classification is No Echo or Biological NE, BI
R(Z, ZDR) Light/Moderate Rain is classified RA
R(Z, ZDR) Heavy Rain or Big Drops are classified HR, BD
R(KDP) Rain/Hail is classified and echo is below the top of the melting layer RH
0.8*R(Z) Rain/Hail is classified and echo is above the top of the melting layer RH
0.8*R(Z) Graupel is classified GR
0.6*R(Z) Wet Snow is classified WS
R(Z) Dry Snow is classified and echo is in or below the top of the melting layer
DS
2.8*R(Z) Dry Snow classified and is echo above the top of the melting layer DS
2.8*R(Z) Ice Crystals are classified IC
QPE Output (all produced via hybrid scan)
• 4bit, 250 m Hybrid-scan Hydro Class• 8bit, 250 m Rate• 4 bit, 250 m 1hr Accum• 4 bit & 8bit versions of 250 m STP Accum (G-R
bias applied)• 8 bit, 250 m no G-R bias applied STP• 8 bit, 250 m User Selectable (will be used for any
and all accumulation time periods)• 8 bit, 250 m 1hr and STP Difference field vs.
Legacy
• Typical Radar sampling limitations (snow at 2000 ft AGL may not be snow at the surface)
• Verification
• “Fuzzy” Logic and cross over between types
• Differentiating between light rain and dry snow in weak echoes
Melting layer detection can help
• Typical Radar sampling limitations (snow at 2000 ft AGL may not be snow at the surface)
• Verification
• “Fuzzy” Logic and cross over between types
• Differentiating between light rain and dry snow in weak echoes
Melting layer detection can help
Hydrometeor Classification Algorithm Hydrometeor Classification Algorithm ChallengesChallenges
Hydrometeor Classification Algorithm Hydrometeor Classification Algorithm ChallengesChallenges
Melting Layer Detection
• Mixed phase hydrometeors: Easy detection for dual-pol!– Z typically increases
– ZDR and KDP definitely increase
– Coexistence of ice and water will reduce the correlation coefficient (CC ~0.95-0.85)
• Precipitation echoes – stratiform or convective regions – with high SNR
• Middle tilts (4°-10° elevation angles)
• Limitation: Overshoot precip
• “Project” results to other tilts in time and space
• Precipitation echoes – stratiform or convective regions – with high SNR
• Middle tilts (4°-10° elevation angles)
• Limitation: Overshoot precip
• “Project” results to other tilts in time and space
Melting Layer Detection Algorithm Melting Layer Detection Algorithm MethodologyMethodology
Melting Layer Detection Algorithm Melting Layer Detection Algorithm MethodologyMethodology
ML Product in AWIPS
Hail Detection• Dual-Pol Hail Signature
– High Z (> 45 dBZ)– Low ZDR (-0.5 to 1 dB), Low KDP (-0.5 to
1 o/km) if dry or mostly dry– Reduced CC (0.85 to 0.95)
• Limitations– Size detection?– Hail signatures may get diluted by
• Rain mixing with hail• Far range
Rain/Snow DiscriminationRAINRAIN SNOWSNOW
ZZ < 45 dBZ< 45 dBZ < 45 dBZ< 45 dBZ
ZZDRDR 0 to 2 dB0 to 2 dB -0.5 to 6 dB-0.5 to 6 dB
KKDPDP 0 to 0.6 deg/km0 to 0.6 deg/km -0.6 to 1 deg/km-0.6 to 1 deg/km
CCCC >0.95>0.95 >0.95 (can be less if >0.95 (can be less if wet)wet)
If the variables overlap so much, how can polarimetric radar discriminate between rain and snow???
Rain/Snow Discrimination: It’s all in trends with height
• Rain– Polarimetric signatures (ZDR and KDP) have a direct
dependence on Z– ZDR and KDP do not typically increase with height– Almost always a pronounced melting layer above rain
• Snow– Polarimetric signatures (ZDR and KDP) do not have
dependence on Z– ZDR and KDP typically increase with height– Differences between “warm” and “cold” snow
• “Cold” snow has higher polarimetric variables than “warm” snow
Warm vs. Cold vs. Wet Snow
• Temperature determines this– < -5oC = “Cold”– ~+1oC > T > -5oC = “Warm”– > +1oC = “Wet”
Crystals (plates, columns, needles)Crystals (plates, columns, needles)
Aggregate Crystals (Dry)Aggregate Crystals (Dry)
Aggregate Crystals (Wet)Aggregate Crystals (Wet)
Surface. Assume temperatures decrease steadily with heightSurface. Assume temperatures decrease steadily with height
Radar Cross Section Radar Cross Section CharacteristicsCharacteristics
Z/ZDR/CC Z/ZDR/CC CharacteristicsCharacteristics
High DensityHigh Density
High ConcentrationHigh ConcentrationOblate, Horizontal OrientationOblate, Horizontal Orientation
Small sizeSmall size
Z < 35 dBZZ < 35 dBZ
ZDR 0-6 dBZDR 0-6 dB
CC > 0.95CC > 0.95
Decreasing densityDecreasing density
Decreasing ConcentrationDecreasing Concentration
Less oblateLess oblate
Larger sizeLarger size
Z increasingZ increasing
ZDR decreasingZDR decreasing
0 > ZDR > 0.5 dB0 > ZDR > 0.5 dB
CC > 0.95CC > 0.95
Rapid increase in densityRapid increase in density
Rapid increase in oblatenessRapid increase in oblateness
Z increasing but < 45 Z increasing but < 45 dBZdBZ
ZDR rapidly increasingZDR rapidly increasing
0.50 > CC > 0.90.50 > CC > 0.9
Rain Snow Discrimination
Z ZDR
KDP CC
Snow
Rain
One Hour Later…
Z ZDR
KDP CC
-SN
Data Quality Improvement
• Ground clutter/Anomalous propagation– High reflectivity (Z) -- (> 35 dBZ)
– Near zero or slightly negative ZDR
– Noisy, lower correlation coefficient (CC) -- (< 0.90)
• Insects/Biological scatterers– Low reflectivity (Z) -- (< 35 dBZ)– Horizontally-oriented with elongated shape: very high
ZDR (> 2 dB up to 6 dB)
– Heterogeneity causes very low correlation coefficients (< 0.70)
Tornado Detection
• Tornado debris is large (from radar perspective), irregularly shaped and randomly oriented– Z > 45 dBZ– ZDR near 0 dB– CC very low (< 0.8)
• A good sign that a tornado is already in progress!– Diagnostic ONLY– Has only been verified for EF-1 or greater
tornadoes at relatively close ranges
Tornadic Debris Signature (TDS)
Z ZDR
CC
TDS!
Debris cloud near GM Plant