Attachment D - NOAA - National Weather Service Radar Functional Requirements.pdf

NOAA/NWS Radar Functional Requirements

NOAA/National Weather Service

Radar Functional Requirements

Approved by the NOAA Observing Systems Council

September 2015

Attachment D

Signatures

Radar Functional Requirements Validation

NOSC Endorsement The NOAA Observing Systems Council (NOSC) has received the National Weather Service Radar Functional Requirements with Line Office, Subject Matter Expert and Research Program concurrence, and is satisfied with the Level-of-Validation provided for the threshold and “optimal for 2030” requirements.

NOAA/NWS Radar Functional Requirements Validation

Line Office Endorsement The Directors of the National Weather Service (NWS) Office of Science and Technology (OST), Office of Climate, Water and Weather Services (OCWWS), and Office of Operational Systems (OOS) Radar Operations Center (ROC), and the Director of the Office of Oceanic and Atmospheric Research (OAR) National Severe Storms Laboratory (NSSL) have received NOAA/NWS Radar Functional Requirements (RFR) document with Project Lead and Developmental Team concurrence and are satisfied with the level of validation for requirements that extend from present day through 2030, given information known at the date of signing.

Radar Functional Requirements Lead Validation The NWS OST Radar Improvement Development Manager, and OCWWS Meteorological Services Division Chief have validated the process and contents of this NOAA/NWS Radar Functional Requirements document.

Radar Functional Requirements Integrated Working Team Concurrence The Radar Functional Requirements Integrated Working Team (IWT) membership concur that the requirements contained herein, comprise NOAA/NWS Radar Functional Requirements from the present day through 2030, given information known at the date of signing.

Mr. Donald W. Burgess, OAR/National Severe Storms Laboratory (Associate) Mr. Joe N. Chrisman, NWS/OOS, Radar Operations Center Dr. Jeffrey G. Cunningham, NWS/OOS, Radar Operations Center (Associate) Mr. John T. Ferree, NWS/OCWWS, Public and Fire Weather Services Division Mr. Richard L. Ice, NWS/OOS, Radar Operations Center (Associate) Dr. Daniel Melendez-Alvira, NWS/OST, Program Plans Division Mr. Dennis A. Miller, NWS/OHD, Hydrologic Science and Modeling Branch Mr. Robert E. Saffle, NWS/OCWWS (Contractor) Dr. David J. Stensrud, OAR/National Severe Storms Laboratory Mr. Christopher C. Wamsley, NWS/OCWWS, Integrated Services Division

Subject Matter Expert/Consultant Contributors The following subject matter experts provided information to both support the requirements formulation process, and over the horizon (~2030) requirements linked to strategic research and future modeling programs.

Mr. Louis Cantrell, NESDIS, Technology, Planning and Integration for Observation (TPIO) (Contractor) Dr. Jacob Carley, NWS/NCEP, Environmental Modeling Center (Contractor) Mr. Carlos Diaz, NWS/OCWWS, Requirements Division Dr. David Dowell, OAR/ESRL/GSD, Assimilation and Modeling Mr. Kurt Hondl, OAR/NSSL, MPAR Program Manager Mr. Mark B. Miller, NWS/OST, NextGen Weather Program Manager Mr. Fred Toepfer, NWS/OST, Hurricane Forecast Improvement Program Dr. Sebastian Torres, OAR/NSSL, Senior Research Scientist (Associate)

Contents

1. Introduction

2. Scope

3. Intended Use of Document

4. Terms and Definitions

5. Radar Data Acquisition Functional Requirements

5.1. NOAA/NWS Strategic Performance Goals

5.2. Threshold Requirements

5.3. Optimal Requirements

5.4. RDA Transmit and Receive

5.4.1. Weather radar variables

5.4.2. Wavelength

5.4.3. Beamwidth

5.4.4. Pulse length

5.4.5. Pulse repetition frequencies (PRFs)

5.4.6. Minimum detectable signal (MDS) and Sensitivity

5.4.7. Dynamic range

5.4.8. System bias

5.4.9. Minimum elevation angle

5.4.10. Maximum elevation angle

5.4.11. Beam elevation and azimuth positioning accuracy

5.4.12. Side lobes

5.5. RDA Signal Processing

5.5.1. Standard deviation of estimates of radar variables

5.5.2. Quantization of radar variables

5.5.3. Linear (radial) range resolution

5.5.4. Effective angular resolution

5.5.5. Complex waveforms

5.5.6. Clutter detection and filtering

5.5.7. Coverage

5.6. Scan Strategy Adaptability

6. Radar Functionality Constraints and Tradeoffs

7. Accuracy

7.1. Radar hardware design and calibration

7.2. Operational settings for scan strategy adaptable parameters

7.3. Complexity and strength of weather and other targets

8. Radar Observations for Meteorological Phenomena and Scenarios

8.1. Clear air

8.2. Stratiform rain

8.3. Winter weather

8.4. Heavy rainfall

8.5. Convective storms

8.6. Tropical cyclones

9. System-independent Observation Requirements for Weather Phenomena

10. Individual Radars and Network Considerations

10.1. Geographic coverage

10.2. Individual radar products

10.3. Multi-radar and multi-parameter products

10.4. Local and downstream users

Appendix A: References

Appendix B: Terms and Definitions

Appendix C: Requirements Validation

Appendix D: Subject Matter Expert Validation Statements

Appendix E: Integrated Working Team and Subject Matter Expert Concurrence

1. Introduction The mission of the National Oceanic and Atmospheric Administration’s (NOAA) National Weather Service (NWS) is to provide weather, water, and climate data, and forecasts and warnings for the protection of life and property and enhancement of the national economy. Weather radars are a critical component of the weather data observing systems that support this mission. This document formalizes NOAA/NWS Radar Functional Requirements (RFR) that are radar-design independent (as much as possible), address current and future (~2030) multiple mission needs, and support agency planning associated with next generation weather radar surveillance within NOAA/NWS.

2. Scope This document considers multiple areas, including the types of weather phenomena for which radar observation is a significant factor; NOAA/NWS current mission performance goals and strategic goals for the future; evolving weather observation needs from activities such as storm scale numerical modeling and anticipated growth in aviation traffic; and support to other Government agency and private sector partners. The functional requirements reflect current needs as well as capabilities anticipated to be needed for agency performance goals extending to approximately 2030.

The focus of this document is on single site radar data acquisition and signal processing functionality, the Radar Data Acquisition (RDA) component of a weather radar system, but discussions are included on the broad range of uses for these observations. The extent and quality of geographic and vertical coverage of individual radars are encompassed in RFRs such as wavelength, beamwidth, minimum/maximum elevation angles and scan strategy flexibility. National geographic coverage is a function of the number of NOAA/NWS radars deployed and the availability of data from non-NOAA radars. The number of NOAA/NWS radars deployed is a programmatic determination involving cost and climatology across the nation, and is beyond the scope of this document.

3. Intended Use of Document

This document informs NOAA/NWS management on the substance and rationale for future radar functional requirements. The RFRs provide a basis for ongoing evaluations of the role of NOAA/NWS radar capability improvements in meeting strategic performance goals.

The document also helps to inform potential radar acquisition programs, including NOAA/NWS engagement with the Federal Aviation Administration (FAA) in the FAA Next Generation Surveillance and Weather Radar Capability (NSWRC) program to replace current FAA terminal tracking, aircraft surveillance, and weather radars. These NOAA/NWS RFRs would inform the acquisition program in the development of engineering system specifications for any particular radar acquisition.

It is important to recognize that the RFRs are not, in themselves, sufficient to act as system acquisition requirements. In any system acquisition, the procuring agency must balance cost, technical performance and schedule when developing a system specification and evaluating potential system improvements. Although this RFR document largely represents the functional capabilities achieved with the present WSR-88D system, it does not preclude agency consideration of potential future radar systems that would address the RFRs with different radar designs (e.g., Phased Array Antenna radars; low power, short wavelength radars). Further, agency management may well judge that the overall functional capability of a potential radar design will improve operations even if certain RFRs are not fully met.

This is a “living document” which is meant to be revisited and revised at approximately five year intervals to take advantage of changing technologies and potentially refined requirements.

4. Terms and Definitions

It is not practical to discuss weather radar functionality requirements without using technical parameters and terminology. Appendix B, Terms and Definitions, contains explanations of the terminology used in this document. These explanations are not meant to be comprehensive engineering definitions, but rather, guidance to the intended audience for this document, e.g., managers, budget analysts, program planners. More rigorous definitions can be found in sources such as the American Meteorological Society Glossary of Meteorology, Federal Meteorological Handbook No. 11, or the Warning Decision Training Branch’s Distance Learning Operations Course (WDTB DLOC) radar training web site.

5. Radar Data Acquisition Functional Requirements This section provides a concise listing of NOAA/NWS RFRs grouped into three categories: RDA transmit and receive, RDA data signal processing and RDA scan strategy adaptability. A brief example of operational impact of each RFR is included. More comprehensive discussions of the use of various radar capabilities for particular weather scenarios are contained in Section 8.

5.1. NOAA/NWS Strategic Performance Goals

These RFRs are directly related to meteorological phenomena observations needed to maintain the current high level of forecast and warning performance, and to meet future performance goals as stated in NOAA and NWS Strategic Plans. These strategic performance goals include the following:

Increase warning lead times for tornado and other high-impact events (e.g., flash floods, severe thunderstorms)

Reduce tornado warning false alarm rate without degrading probability of detection

Promote comprehensive weather situational awareness

Improve weather decision services and convey uncertainties associated with data and products

5.2. Threshold Requirements

NOAA/NWS must maintain the high quality of radar information provided by the operational Weather Surveillance Radar – 1988 Doppler (WSR-88D) with Dual Polarization. Threshold RFRs generally represent the achieved operational capabilities of the WSR-88D. In cases where the achieved WSR-88D capability is greater than the minimum requirements of the WSR-88D System Specification, the achieved capability is cited in the relevant RFR. The WSR-88D capabilities continue to be improved, and these RFRs represent minimum required functionality

– not limits to system improvements. Future updates to the RFRs will reflect any approved modifications based on these improvements.

5.3. Optimal Requirements

For some areas, more advanced functionality would be needed to better support science and analysis tools necessary to accomplish NOAA/NWS strategic goals for mission performance improvements. These strategic goals include developing a capability to provide large increases in tornado warning lead times through the application of storm scale numerical models to forecast the formation of tornadoes, a capability termed Warn-on-Forecast (WoF). Such RFRs will be identified as ‘Optimal for 2030’ operations. Certain Optimal RFRs are also related to the general observing requirements for particular weather phenomena contained in the NOAA Consolidated Observation Requirements List (CORL) as described in Chapter 9.

5.4. RDA Transmit and Receive

These RFRs comprise the basic functionality of weather radar hardware design.

5.4.1. Weather radar variables

The radar must provide observations for the following basic variables of polarimetric Doppler weather radar. Forecasters typically utilize multiple radar variables, along with non-radar information, to analyze weather events and issue forecasts and warnings.

Threshold: WSR-88D capability

Reflectivity (Z)

Radial Velocity (V)

Spectrum Width (SW)

Differential Reflectivity (ZDR)

Correlation Coefficient (CC)

Differential Phase (PHI)

Numerical model initial conditions and forecasts are sensitive to the radar-observation errors specified during data assimilation, but the characteristics of these observation errors are currently poorly known. Improved specification of observation errors, both in a mean sense and on an observation-by-observation basis, would be possible if sample variances for the observation estimates were available. Extensive assimilation of radar data for input to storm scale numerical models is critical to achieving the capability for forecasting the development and evolution of tornadic storms, and thus extending tornado warning lead times significantly.

Optimal for 2030: Within each sample bin

Sample variance for reflectivity

Sample variance for radial velocity

Sample variance for differential reflectivity

Sample variance for correlation coefficient

Sample variance for differential phase

5.4.2. Wavelength

The radar must provide high quality information throughout, and beyond, regions of heavy rainfall to long ranges. The radar must also provide operationally acceptable (e.g., provision of data that significantly enhance forecast and warning operations) values of maximum unambiguous range and maximum unambiguous velocity for any given pulse repetition frequency.

Threshold: Performance of WSR-88D (S-Band, ~10 cm)

The WSR-88D wavelength of ~10 cm is the optimal choice as it supports a maximum range of 460 km due to minimized attenuation by heavy rainfall and also provides the best combination of maximum unambiguous range and maximum unambiguous velocity compared to shorter wavelengths. This wavelength ensures that each radar will provide critical coverage within its umbrella.

5.4.3. Beamwidth

The radar must provide sufficient spatial resolution (both horizontally and vertically) to detect fine scale features, such as tornadic circulations, to an operationally acceptable (e.g., provision of data that significantly enhance forecast and warning operations) range.

Threshold: Performance of WSR-88D (1.0 deg Beamwidth in azimuth and elevation)

Due to antenna rotation and processing of multiple pulses per data radial, the WSR-88D hardware beamwidth (~ 1 deg) provides an effective beamwidth of 1.5 deg in azimuth.

A signal processing windowing technique, termed Super Resolution, provides high quality data radials at 0.5 deg increments with an effective beamwidth of 1.1 deg.

5.4.4. Pulse length

A short pulse length is necessary to provide the linear resolution required for detection of complex, small scale storm features such as tornadic circulations. An additional, longer pulse length is also necessary to enable greater ability to detect returns from weaker targets such as light snow and clear air.

Threshold: Performance of WSR-88D (1.57 µs (short pulse) and 4.71 µs (long pulse))

Future use of complex pulse generation methods such as Pulse Compression could provide increased sensitivity without degrading other requirements such as linear range resolution. Increased sensitivity could extend the range of coverage of weaker targets, thereby enhancing forecast and warning operations.

5.4.5. Pulse repetition frequencies (PRFs)

The radar must provide useful information for a wide variety of weather scenarios, including the concurrent presence of multiple storms along the same radial, multiple storms at different ranges in different sectors, and strong radial velocities in the storms.

The radar must support multiple PRF values to minimize the effects of range folding on particular storms while preserving the best possible estimations of radial velocity for those storms.

Threshold: WSR-88D capability

A PRF, or combinations of PRFs, to support a maximum unambiguous velocity of ± 32 m/s

A PRF to support a maximum unambiguous range of 460 km

A range of PRFs between 318 hertz and 1310 hertz

Radial velocities often exceed 32 m/s in association with tornadic supercells and strong mid-latitude cyclones. Since there is a strong relationship between observed radial velocities, storm intensification, and wind damage of all kinds, increasing maximum unambiguous velocity would lead to improved decision support services and more effective assimilation of these observations into numerical forecast models.

Furthermore, an increase in maximum unambiguous velocity in the lower elevation angles would help assure that strong circulations occurring within the planetary boundary layer are routinely well sampled within a radar’s coverage area.

Optimal for 2030:

A PRF, or combinations of PRFs, to support a maximum unambiguous velocity of ± 50 m/s

5.4.6. Minimum detectable signal (MDS) and Sensitivity

A low MDS is necessary to provide the sensitivity to detect weak, fine scale targets such as gust fronts and weak circulations, and to obtain valid velocity measurements from targets such as insects and moisture discontinuities. Radar sensitivity is closely related to the MDS. Higher sensitivity corresponds to an ability to detect weak signal strength features to longer ranges. The Threshold RFR is cited for a range of 50 km to provide reference point values of the required MDS.

Threshold: Performance of WSR-88D

0.0 dB Signal to Noise Ratio (SNR) for a -9.5 dBZe target at 50 km in short pulse and a -18.5 dBZe target at 50 km in long pulse

5.4.7. Dynamic range

The radar must process returns from both weak and strong targets concurrently (e.g., mix of fine lines and strong thunderstorms). The receiver dynamic range must be large enough to accommodate typical weather scenarios.

Threshold: WSR-88D capability of 93 dB

A larger dynamic range will support improved processing of strong clutter targets and increased sensitivity for weak returns. The improved data quality will enhance the utility of radar data in the numerical models that are crucial to providing tornado and other severe storm warnings based on forecasts of storms.

Optimal for 2030: 97 dB

5.4.8. System bias

The radar hardware calibration must determine the measurement uncertainties related to the system hardware components as accurately as possible for all radar variables.

Accurate system bias calculation ensures that observations accurately represent the returned signals for those variables.

Threshold: WSR-88D capability, in the absence of clutter filtering.

Reflectivity: 1 dBZ for target with true spectrum width of 4 m/s and SNR > 10 dB

Velocity: 0.0 m/s for target with true spectrum width of 4 m/s and SNR > 8 dB

Spectrum Width: 0.2 m/s for target with true spectrum width of 4 m/s and SNR > 10 dB

Differential Reflectivity: 0.1 dB for target with true differential reflectivity (ZDR) of less than ±1 dB, true spectrum width of 2 m/s, Correlation Coefficient ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB (for ZDR with a magnitude greater than 1 dB, bias should be less than 10% of the ZDR magnitude)

Correlation Coefficient: 0.006 for target with true spectrum width of 2 m/s, Correlation Coefficient ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB

Differential Phase: 1 deg for target with true spectrum width of 2 m/s, Correlation Coefficient of ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB

5.4.9. Minimum elevation angle

The radar must be able to scan at negative elevation angles to improve low altitude weather detection from high altitude sites (e.g., mountain sites).

Threshold: WSR-88D capability of -1.0 deg

5.4.10. Maximum elevation angle

The radar must be able to observe the mid-level and upper portions of convective storms close to the radar site. These are the regions of storms where the parent rotation of tornadoes, large hail, and the beginnings of microbursts are often first observed. The radar must also be able to observe high elevations to support calibration procedures based on measurements of the sun and the clear blue sky.

Threshold: WSR-88D capability of 60.0 deg

5.4.11. Beam elevation and azimuth positioning accuracy

Beam location accuracy is critical to properly locate tornadoes and other severe weather phenomena for accurate specification of warning areas and for assimilation of radar data into numerical models.

Threshold: WSR-88D capability of 0.15 deg

Further improvements to Differential Reflectivity calibration of system bias and general data quality will require better beam positioning accuracy and precision. Research radars have positioning accuracy of 0.0055 deg.

Optimal for 2030: Elevation and azimuth positioning accuracy within 0.0055 deg

5.4.12. Side lobes

The radar observation data must be calculated predominantly from the signal power returned within the main lobe of the beam pattern to ensure the data represent the nominal spatial resolution of each sample volume. Signal power returns from energy transmitted in nearby lobes (side lobes) must be minimized.

Threshold: WSR-88D capability, two-way side lobes:

First side lobe: ≤ -70 dB relative to the peak of the main lobe

Side lobes beyond ± 2 deg from beam center decrease to ≤ –96 dB at ± 6 deg relative to the peak of the main lobe, decreasing to -110 dB at ± 20 deg relative to the peak of the main lobe.

5.5. RDA Signal Processing

Sophisticated RDA signal processing of target returns is necessary to generate accurate, reliable radar variable estimates, and to utilize flexibility in setting radar transmit and receive parameters.

5.5.1. Standard deviation of estimates of radar variables

The radar variable values generated by signal processing of target return signals must be consistent from observation to observation for any given type of spatially homogeneous weather target to ensure reliability and dependability for subjective and objective applications of the data. Accurate estimations of system bias and low standard deviations of estimates of radar variables are critical components of the accuracy and representativeness of the radar variable estimates.

Reflectivity: ≤ 1 dB for target with true spectrum width of 4 m/s and SNR ≥ 10 dB

Velocity: ≤ 1 m/s for target with true spectrum width of 4 m/s and SNR > 8 dB

Spectrum Width: ≤ 0.5 m/s for target with true spectrum width of 2 m/s and <

0.95 m/s for true spectrum width of 4 m/s and SNR > 10 dB

Differential Reflectivity: < 0.4 dB for target with true spectrum width of 2 m/s, Correlation Coefficient ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB

Correlation Coefficient: < 0.006 for target with true spectrum width of 2 m/s, correlation coefficient ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB

Differential Phase: < 2.5 deg for target with true spectrum width of 2 m/s, correlation coefficient of ≥ 0.99, dwell time of 50 ms and SNR ≥ 20 dB

Optimal standard deviation goals for 2030 are not well established. However, techniques that reduce the standard deviation of estimates should be developed so that data quality can be traded for other things such as faster updates.

5.5.2. Quantization of radar variables

Radar variables must be provided to data processing systems with sufficient quantization to support data accuracy determinations and to provide the capability to reveal the small scale, often critical, variations of values within many weather targets (e.g., convective storms). Radar variable quantization is sometimes used interchangeably with precision of the variable estimates (e.g., the WSR-88D System Specification), even though such usage is not technically correct.

Reflectivity: 0.5 dB

Velocity: 0.5 m/s

Spectrum Width: 0.5 m/s

Differential Reflectivity: 0.0625 dB

Correlation Coefficient: 0.00333

Differential Phase: 0.35 deg

WSR-88D’s quantization of spectrum width is too coarse, given the standard deviation of estimate requirement and typical range of spectrum widths. Reducing the quantization would provide improved discrimination of turbulence intensity.

Optimal for 2030: Spectrum Width Quantization of 0.1 m/s

5.5.3. Linear (radial) range resolution

The radar must provide sufficient spatial resolution to detect fine scale features, such as tornadic circulations. Range Resolution is determined by pulse length and the Range Weighting Function (RWF). The RWF is a combination of transmitted pulse characteristics and processing within the receiver and signal processor. The RWF delivers optimum values of SNR and range resolution for the radar variables.

250 m for the Short Pulse Length

5.5.4. Effective angular resolution

Signal processing techniques have been developed to use windows of overlapping azimuthal samples to significantly increase the antenna’s nominal effective angular resolution at the cost of a slight increase in the standard deviation of radar variable estimates. Combined with 250 m linear range resolution, the resultant data have proven to be highly valuable in identifying tornadic circulation patterns and extending the effective range of such identifications.

Threshold: Performance of WSR-88D: Effective angular resolution of 1.1 deg

5.5.5. Complex waveforms

The radar must provide useful information for a wide variety of weather scenarios, including the concurrent presence of multiple storms along the same radial, multiple storms at different ranges in different sectors, and strong radial velocities in the storms.

To optimize the radar operation, the radar must provide for the use of complex radar waveform transmission and signal processing in order to minimize range-velocity ambiguities.

Multi-PRF

Systematic Phase coding

Staggered Pulse-Repetition-Time (PRT)

5.5.6. Clutter detection and filtering

Undesirable signal artifacts such as ground clutter and electromagnetic interference can degrade the meteorological utility of the radar data (e.g., rainfall estimation, tornadic circulation analysis). The radar system must identify and provide a means of removing such artifacts. Artifact removal must minimize any degradation of the desirable meteorological data.

Threshold: Performance of the WSR-88D

Automated detection of clutter at Clutter to precipitation Signal Ratios (CSRs) down to -13 dB.

Clutter filtering resulting in reduction of CSRs to -13 dB or better.

Minimize the bias and standard deviation induced by the clutter filter process:

1 dB bias and standard deviation for reflectivity and 1 m/s bias and standard deviation in radial velocity and spectrum width estimates.

5.5.7. Coverage

Radar data should be provided to the maximum range and altitude for which the data provide operationally useful meteorological information.

Maximum range: 460 km for reflectivity; 300 km for velocity and polarimetric variables

Maximum altitude: 70,000 ft

5.6. Scan Strategy Adaptability

The forecaster must have the ability to optimize radar operations for weather scenarios ranging from clear air to fast moving tornadic storms. This functionality must include defining preset scanning strategies (also termed Volume Coverage Patterns, or VCPs) to sample specific elevation and azimuth positions (i.e., the spatial sampling grid), with certain RDA parameters such as waveform, PRF, and antenna beam scan speed for each slice. The functionality must also support forecaster and automatic selection of PRF values to dynamically tune the preset scan strategy to minimize range folding and velocity aliasing in the area of significant storms.

Through selection of preset scan strategies appropriate for different general weather scenarios, and dynamic tuning of the PRF within the scan strategy in use, forecasters exercise the best possible operational trade-offs concerning sensitivity, standard error of estimates, maximum unambiguous range, maximum unambiguous velocity and volume coverage time.

Multiple Clear Air and Precipitation Mode VCPs with adaptable parameters including:

o Number of elevation slices and the elevation angle for each slice o For each elevation slice:

Transmit waveform

PRF

Antenna scan speed Pulse length Number of pulses to process for each sample bin Edge angles of PRF sectors SNR thresholds for each of the 6 radar variables estimated Azimuth angular resolution Velocity quantization (0.5 m/s and 1 m/s)

Support for PRFs between 318 hertz and 1310 hertz

Manual and automatic selection of different PRFs for three azimuthal sectors for each of two sets of elevation slices

Ability to repeat the lowest elevation angle mid-way through a volume scan and/or to eliminate high elevation angle scans when insufficient weather echo exists.

The design parameters of the WSR-88D limit the temporal resolution of volume updates to three to five minutes (two to three minutes update of the lowest elevation slice), depending on the locations and heights of storms in a particular weather scenario, while retaining the required sensitivity, spatial resolution and standard deviation of estimates. While this temporal resolution has been sufficient to support dramatic increases in tornado warning lead times since the introduction of the radars, it does not adequately represent the true speed of formation and evolution of tornadoes. Faster scanning capability, with retention of the data quality attributes stated above, would improve the timeliness of tornado detection, thus increasing the average tornado warning lead time. Further, data update times on the same scale as tornado evolution would better support the development of storm scale numerical models intended to forecast tornado development, leading to greater increases in tornado warning lead times with the Warn-on-Forecast approach.

Currently, the lowest elevation angle scanned by the WSR-88D is 0.5 deg (0.17 deg at one site, Langley Hill, WA) even though the physical lower limit of the radar is -1.0 deg. Phased array technology could provide adaptability of acquisition parameters (e.g., PRT or PRT sequence, number of pulse samples) for every point in the spatial sampling grid. On top of this, the spatial sampling grid could be adaptable as well (e.g., super-resolution vs. legacy resolution, different elevation spacing).

Optimal for 2030:

One minute or less volume coverage time with no degradation of the sensitivity, spatial resolution or standard deviation of measurement for radar variable estimates.

Capability to define scan strategies with elevation angles as low as the operational siting and radar design allow

6. Radar Functionality Constraints and Tradeoffs

Coverage, spatial resolution, sensitivity, maximum unambiguous velocity and range parameters vary in an inter-related manner with the radar’s transmitted power, wavelength of the transmitted signal, beamwidth, and radar waveform including pulse length and PRF. For rotating dish antenna radars, the transmitted power, wavelength and beamwidth are generally constant for a given design. For phased array antenna radars, the wavelength may change with sub-array partitioning, and the transmitted power and beamwidth are adaptable. The pulse length and PRF are adaptable for both types of radars. In operations, a given scanning strategy takes advantage of the available adaptable parameters to best sample the existing weather scenario, with phased array antenna radars having more such adaptability. Often, a particular scanning strategy may degrade one parameter in order to enhance another with more value to operations for the given weather.

The following bullets summarize the effects of differing radar operating parameters on radar functionality.

Effective radar coverage increases with:

o Greater sensitivity, broader scan angles, lower elevation angles (extending the spatial sampling grid), increased pulse samples and/or signal processing techniques to lower variance of estimates and permit use of data at lower signal to noise ratios.

Spatial resolution increases with:

o Narrower effective beamwidth, shorter wavelength, shorter pulse length, and sharper range weighting function

Temporal resolution increases with:

o Faster effective beam scanning speed, smaller spatial sampling grid, and shorter dwell times.

Sensitivity increases with:

o Shorter wavelength, higher transmit power, longer pulse length, narrower beamwidth

(for a given transmit power), lower receiver noise power

Attenuation decreases with:

o Longer wavelength

Maximum unambiguous velocity increases with:

o Longer wavelength, shorter PRT, multiple PRTs and signal processing methods

Maximum unambiguous range increases with:

o Longer PRT

7. Accuracy

The accuracy of radar estimates of reflectivity and other radar variables is a complex issue involving radar hardware design and calibration, appropriateness of selected values for scanning strategy adaptable parameters, and the complexity and strength of weather and other targets. The following sub-sections briefly discuss these major factors affecting accuracy. These discussions do not include specific requirements, but are intended to inform the readers of this document and to highlight the inter-related nature of the RFRs listed in section 5.

7.1. Radar hardware design and calibration

The system acquisition choices of radar design aspects such as wavelength, beamwidth and transmit power determine fundamental limits of the applicability of the radar for different weather scenarios. For example, long range surveillance of weather scenarios that include both light returns (e.g., gust fronts) and heavy rainfall requires a radar with little attenuation in heavy rain (e.g., longer wavelength) and high sensitivity.

Regardless of the inherent capabilities of a radar’s design, there will be some signal degradation due to the combination of hardware components and their inter-connections. The sum of the signal degradations due to these sources is termed system bias. Radar engineering calibration estimates the signal bias, and must be very accurate to support valid estimates of radar variables. Typically, however, engineering calibration techniques cannot be absolutely accurate, and more stringent techniques can be very expensive to implement operationally.

The calibration techniques chosen for a given radar represent a compromise between system bias estimation accuracy and cost.

7.2. Operational settings for scan strategy adaptable parameters

Operational weather radar functionality typically offers control over certain scan strategy parameters such as PRF, antenna beam scanning speed, and elevations and azimuths processed. These choices are used to create preset scan strategies (also termed Volume Coverage Patterns) to achieve reasonable tradeoffs among volume update time, maximum unambiguous velocity, maximum unambiguous range, coverage, spatial resolution and clutter mitigation for different weather scenarios. Dynamic tuning of parameters for specific scanning sectors, such as modifying the PRF to mitigate range folding, is used to further enhance performance in the vicinity of significant storms. Selection of the proper scanning strategy and dynamic tuning of adaptable parameters are critical to achieving accurate estimates of radar variables.

7.3. Complexity and strength of weather and other targets

Even with high quality radar hardware design and calibration, and with appropriate selection of scan strategy parameters, real world weather returns present significant challenges to accurate estimations of radar variables. Weather, insects, birds, anomalous propagation and ground targets contribute to a complex radar return signal in many different weather scenarios. The distribution, intensities and velocities of precipitation areas may exceed the sampling and signal processing capabilities for mitigating range and velocity ambiguities. The strength and velocity of a storm may vary significantly within a single sample volume, especially for the larger sample volumes at longer ranges. Precipitation areas may fall partially, or wholly, below the lowest beam of a scanning strategy, or may be obstructed by terrain or man-made objects. For certain radar variable estimates, such as Differential Reflectivity, the radar estimate may be accurate but the scientific association with a type of precipitation (e.g., snow) may be problematical due to the variability of the weather target compared to other meteorological data (e.g., temperature).

Given a well-designed and calibrated radar, operations with appropriate scanning parameters and signal processing techniques, and an understanding of the complexities of weather radar estimations of radar variables, the final test of a radar’s accuracy is the usefulness of its data for support to operational forecasts and warnings.

8. Radar Observations for Meteorological Phenomena and Scenarios

Weather radar provides valuable information on a broad range of meteorological phenomena and weather scenarios. The following sections describe several scenarios and note the radar variables most important to their analysis.

8.1. Clear air

This category includes all non-precipitation targets, including insects, birds, bats, smoke and volcanic ash plumes, chaff, Bragg scatter, and fine lines (e.g., frontal boundary, gust front, dryline).

Velocity estimates from insect targets are particularly useful for wind information, including wind shift boundaries. The resultant wind information is useful for real-time activities (e.g., airport runway switching) and as input to numerical model initialization.

Radar signals from clear air targets are often weak, and degrade rapidly with range. The relative absence of insects in cold weather (e.g., winter) reduces the quality and coverage of velocity estimates. Special clear air scanning strategies have been developed to maximize the radar coverage by employing a longer pulse length, and by collecting more samples per estimate.

8.2. Stratiform rain

Stratiform rain typically forms in the lowest few kilometers of altitude, and extends over relatively large areas with weak horizontal gradients of intensity and limited vertical extent. In some scenarios, stratiform rain may be associated with other weather phenomena such as mesoscale convective systems and tropical or extra-tropical cyclones.

Reflectivity and the polarimetric variables are used to analyze the intensity and areal extent of stratiform rain, and to estimate rainfall accumulations. Velocity data are used to calculate vertical wind profiles and as a general indication of wind direction and speed. Polarimetric variables are also used, in combination with other types of observations, to analyze rain/snow transition zones and to assess the likelihood of freezing rain and aircraft icing areas.

Radar-derived quantitative stratiform precipitation estimates can be degraded by several factors, including: overshooting precipitation areas with the lowest beam, anomalously enhanced reflectivity values for precipitation in the melting layer, and beam blockage. These factors limit the range to which such quantitative estimates are valid.

8.3. Winter weather

Winter weather includes a variety of precipitation types: freezing rain, ice pellets (sleet), snow, snow pellets, and graupel. Some or all of these precipitation types may occur concurrently within the radar umbrella. Winter weather precipitation has stratiform characteristics similar to those of stratiform rain. One of the most important uses of radar for winter weather is the identification of transition zones between the different types of precipitation (e.g., the rain-snow line). However, much winter precipitation occurs at low altitudes – presenting a challenge for single radar coverage at longer ranges. In areas where the winter precipitation is detected on the lowest radar beam at longer ranges, other effects (e.g., melting, refreezing, sublimation, drifting of snow) add to the uncertainty of radar estimates of precipitation types, amounts and surface locations.

Reflectivity and the polarimetric variables are used to analyze the intensity, type and areal extent of precipitation, and to estimate snowfall accumulations. Velocity data are used to calculate vertical wind profiles and as a general indication of wind direction and speed.

Polarimetric variables are also used, in combination with other types of observations, to analyze precipitation type at the ground and aloft important to surface and air transportation.

Clear air scanning strategies with longer sampling times are often used to detect the weaker reflectivity. Higher scan angles are typically not required since the precipitation areas do not have a great vertical extent. Coverage limitations are similar to those of stratiform rain.

8.4. Heavy rainfall

Heavy rainfall can be showery (intense rainfall rates for tens of minutes) or stratiform (moderate rainfall rates for hours) in nature. Heavy rainfall from convective storms may have larger raindrops resulting from melting hail or graupel, and its intensity typically varies greatly both spatially and temporally. Heavy rainfall can also occur with densely packed small raindrops formed from ‘warm process’ coalescence growth of raindrops in temperatures warmer than 0 deg C. Warm process rainfall tends to be stratiform (i.e., lower spatial and temporal variation in intensity), and can occur with tropical or non-tropical systems.

Radar estimates of heavy precipitation amounts have much greater spatial and temporal resolution than those from rain gage networks. Reflectivity, Differential Reflectivity and Differential Phase (as converted to Specific Differential Phase in downstream processing) are the main radar variables used in precipitation estimation algorithms. Combined with rain gage data, the radar rainfall estimates have led to dramatic advances in flash flood warning performance. The introduction of dual polarization capability to the WSR-88D has improved precipitation estimates versus using reflectivity only, with the promise of further improvements as the polarimetric precipitation estimation algorithms are refined.

There are significant inherent limitations in capabilities of a single radar for precipitation estimation. Precipitation estimates typically degrade with range due to several factors, e.g., overshooting of low altitude precipitation, overestimation in bright band areas, uncertainties of estimating surface rainfall rates from observations in snow above the bright band, and evaporation below lowest sampling beam. However, single radar data are also used in downstream data fusion and data assimilation systems that integrate networks of radar and other operational datasets into advanced precipitation estimation models.

8.5. Convective storms

This category includes convective rain showers, general and severe thunderstorms (i.e. those that produce NWS-defined severe wind gusts or severe-sized hail), and tornadic storms.

Convective precipitation forms in cumulus or cumulonimbus clouds and occurs where updraft velocity is large relative to the fall speed of precipitation particles. Convective precipitation often has strong horizontal reflectivity gradients and can extend several kilometers vertically.

Convective storms typically have rapid evolution (on the order of 1 min or less) in strength and size and given the right ingredients in the near-storm environment, can develop severe weather features such as tornadoes. Tornadic circulations are very small scale and can vary significantly across distances of just a few hundred meters. Convective storm scenarios, including tornadic storms, range from isolated storms in limited areas of the radar umbrella, to widespread storms throughout the umbrella, to squall lines oriented along the full diameter of the umbrella.

Convective storms, especially severe/tornadic ones, present the largest challenge for weather radars, and their surveillance drives the most stringent values of the various radar functionality requirements. These storms have precipitation and up/downdraft cores where the key processes determining rainfall rate/type, updraft and downdraft strengths are contained.

Specification of the structure (i.e., vertical extent) and contents of these cores is critical to warning operations. All of the standard and polarimetric radar variables are used extensively in the analysis of convective storms and are critical observation data used to issue severe thunderstorm, tornado, or flash flood warnings.

8.6. Tropical cyclones

Tropical cyclones are large scale warm-core storms with organized convection that exhibit cyclonic rotation around a low pressure center. These storms contain spiral rain bands and can contain very high gradient winds. The rain bands often generate very high rainfall rates. In addition, tropical cyclones can contain tornadic storms, which are often weak with very rapid development and short life times.

Timely detection of tropical cyclone tornadic potential requires fast scanning strategies with good low altitude coverage. Ground-based radar provides the main source of position fixes and other diagnostic information for near land and land-falling tropical cyclones. Such observations are also foundational to short-term forecasts as tropical cyclone conditions persist near and over land. NWS has developed guidance for operating WSR-88D units during tropical cyclone conditions. The application of dual polarization parameters to tropical cyclones is rather new but these parameters are already being used to identify strong updraft cores and to improve rainfall estimates.

Heavy rainfall and tornadoes associated with tropical cyclones often occur at low altitudes, and their detection is degraded at longer ranges as the radar beam overshoots the precipitation and rotational signatures.

9. System-independent Observation Requirements for Weather Phenomena

NOAA has developed the Consolidated Observation Requirements List (CORL), part of the NOAA Program Observation Requirements Document (PORD). Together, the CORL and PORD constitute an extensive database that summarizes specific NOAA program requirements. The Local Forecast and Warnings (LFW) CORL identifies 31 environmental observation requirements covering physical properties of the atmosphere and oceans. These observation requirements are further specified by spatial, temporal, accuracy and other attributes. The CORL requirements are based on the physical attributes of weather phenomena, e.g., size, gradients of storm features, rapidity of evolution, vertical extent, etc. They are purposely not observation system dependent, but provide guidance to NOAA/NWS efforts to improve observing capabilities and for comparing observing system interdependencies.

Weather radar provides information on 11 of the 31 CORL requirements, and is critical for convective storm scale observations. It is important to note, however, that CORL values are not meant to be used directly as NOAA/NWS radar functional requirements, nor as system specification requirements for any hardware acquisition. Instead, they should be used as input to the development of such requirements and this RFR document used them as guidance where applicable.

CORL values denote the observation attributes that would describe a particular weather phenomenon in enough detail to completely inform current and anticipated analysis and scientific techniques for detection and forecasting of that phenomenon. For example, although the CORL states 100 m spatial resolution for storm area precipitation rates, a single radar can only achieve this capability within a short range of the radar. Any specific radar acquisition program would need to balance affordability, performance and schedule to address CORL requirements.

The CORL requirements most relevant to single radar functionality are listed in the following table.

Phenomenon Vertical Resolution

Horizontal Resolution

Accuracy Sampling Interval

Precipitation Rate Profile:

Storm Area

100 m 100 m 1 mm/hr 30 sec

Precipitation Type: Storm Area

100 m 100 m Not Applicable

30 sec

Wind Direction Profile:

Storm Area

100 m 100 m 1 deg 30 sec

Wind Speed Profile:

Storm Area

100 m 100 m 0.5 m/s 30 sec

10. Individual Radars and Network Considerations

This document’s focus is on functional requirements for individual radar data acquisition. The following sections present additional discussions to help place these requirements within the context of overall radar information use by NOAA/NWS and other users.

10.1. Geographic coverage

NOAA/NWS forecast and warning operations use combined data from the network of radars present in a given region. The over-lapping radar umbrellas will often mitigate the coverage limitations of the individual radars concerning spatial resolution, overshooting of low altitude precipitation, and ability to detect weak targets. Individual radar coverage functionality remains critical, however, for areas where multiple radar coverage is not present, and for operations when one or more of the overlapping radars is down for maintenance.

Low altitude (below ~1500’ AGL) coverage will become increasingly important for initiation of storm scale numerical models as the complexity of these models’ underlying physics is improved. Quality information on the pre-storm environment is critical to the effectiveness of severe storm forecasts from these models, and to achieving the performance goals that depend on Warn-on-Forecast techniques. The requirements developed in this document should be used to inform the development of radar network coverage requirements for NOAA/NWS.

To supplement its own network data, NOAA/NWS explores opportunities to use data from non- NOAA radars in particular locations. Currently, NOAA/NWS acquires data from the 45 Terminal Doppler Weather Radar (TDWR) units operated by the Federal Aviation Administration (FAA).

In the future, such supplemental data acquisition efforts may include other FAA radars as well as local radars operated by television stations or local governments or any other public or private entities willing to share radar data.