Warning Decision Support System–Integrated Information (WDSS-II)
)
Interrogating the 3D structure of a hurricane using the 4D, interactive display. The 3D structure (at 1 km resolution every 5 minutes) was created in real-time by combining high-resolution radar data. Larger image
The WDSS-II is a multi-radar/sensor real-time data ingest and processing system that can be used to evaluate experimental applications in an operational setting. It is also a powerful application development tool. It is easy to add new products and concepts, and it provides a seamless path from data ingest, data processing, and output using standard formats. This improves the pace of science and technology infusion into operational warning decision systems.
The WDSS-II system is primarily used for research, prototype application development, and application evaluation, but the system is run 24/7 to evaluate the performance and scalability of the system for operational uses. SWAT partners with the Real-time Applications and Display Development team in NSSL's radar division to develop this technology. The WDSS-II is the result of over 10 years of research, application development, and operational testing at NSSL and NWS forecast offices.
History
The Warning Decision Support System – Integrated Information (WDSS-II) is the second generation of a system of tools for the analysis, diagnosis and visualization of remotely sensed weather data. The first version of the Warning Decision Support System (WDSS), now known as the legacy WDSS was developed in the early 1990s and was based on data from individual radars. The NSSL tornado vortex signature and mesocyclone detection algorithms currently used in operations by the NWS were first implemented, tested, and validated within the WDSS framework. To support university and other researchers, workstation versions (called WATADS) of these operational algorithms were distributed freely.
Since the WDSS was developed, there have been two major advancements. Computer networking and compression methods have improved significantly through a project called CRAFT. CRAFT allows data from individual radars to be transmitted, in real-time, over the Internet to interested users, and has made the development of new weather applications possible. The second advancement is a data access application programming interface (API) that enables access to data from various sensors – enabling rapid development of new meteorological applications.
The current NSSL Warning Decision Support System – Integrated Information (WDSS-II) uses these advancements and now contains over 100 multi-radar/sensor applications of all kinds.
WDSS-II 2D Multi-Radar CONUS Composite Reflectivity
(multi-sensor QC, smoothed/thresholded)
(click image for loop)
Current WDSS-II application development at NSSL
- Multiple-radar SCIT and HDA
Reflectivity information from multiple radars is used to detect and diagnose storm cells.
- Four-Dimensional Multiple-radar applications
NSSL has developed the capability to merge multiple-radar data into four-dimensional grids.
- Gridded Hail and Hail Swath diagnostic products
Using the Severe Hail Index (developed by NSSL in 1990) along with contributions of higher reflectivities above freezing levels, the probability of severe hail and maximum expected hail size products are derived. The gridded hail size data can also be accumulated over time to provide precise hail swath maps, showing both maximum hail size by location, and hail damage potential (combination of hail size and duration of hail).
- Vortex Detection and Diagnosis Algorithm (VDDA)
More sophisticated techniques are being developed to accurately detect and diagnose rotation in radar velocity data. Shear fields from single and multiple radars can be accumulated over time within specific height layers – providing a proxy for "rotation tracks" of mesocyclone features. Within one rotation track image is information about the past track of the events (which can be used to nowcast the future position) as well as the trend of the strength of the rotation in those events. Also, the rotation track product can serve as a very valuable verification tool to help determine where unreported or unobserved tornadoes may have occurred. The benefit of this image is that it can replace the time-consuming process of replaying radar data and manually tracking individual mesocyclones, and can be used to help deploy damage survey teams.
- Motion Estimation
NSSL is currently developing a sophisticated technique to forecast the motion, growth, and decay of two-dimensional storm fields. The technique also produces a high-resolution motion field that can be used to advect any two-dimensional product, such as precipitation accumulation, VIL, hail, rotation, or lightning fields to provide up to 60-minute forecasts of these phenomena. The high-resolution motion estimates are also used within the 4D multiple radar grids to advect the slightly older data (up to 10 minutes old) forward in time.
- Quality control neural network (QCNN)
NSSL has developed a Quality Control Neural Network (QCNN) designed to look at properties of the three moments of radar data (reflectivity, radial velocity, spectrum width) as well as multi-sensor “cloud cover” data (combined IR satellite and surface temperatures) to segregate precipitation from non-precipitation (and data artifact) echo. By segregating the precipitation areas, the false mesocyclone detections are removed without impacting the true detections within precipitation areas.
- Near-storm environment (NSE)
NSSL has developed an algorithm that analyzes mesoscale numerical model output and derives a large number of sounding parameters. These derived gridded data are used as source input to a number of our current and proposed algorithms. The model initial analysis fields are used to provide greater temporal and spatial resolution of important environmental data for the multiple-sensor applications. The rapidly updating information can be used to capture rapidly evolving thermodynamic fields or fields with large spatial gradients much better than rawinsonde information.
- Enhanced Hail Diagnosis Algorithm (EHDA)
NSSL has enhanced the original single-radar cell-based HDA, known as the Enhanced Hail Diagnosis Algorithm . This improved hail diagnosis uses a sophisticated and more-accurate Neural Network that integrates the traditional reflectivity radar information with velocity radar information (for rotation and storm-top divergence) as well as NSE data from a mesoscale model. Additional outputs include hail size conditional probabilities.
- Using Google Earth to share experimental products
NSSL has begun utilizing Google Earth as a way to share experimental severe weather products with other researchers and operational meteorologists for evaluation and feedback. A variety of multi-sensor severe weather products are generated by NSSL and shared to Google Earth users via the internet at http://wdssii.nssl.noaa.gov. These products include spatially gridded fields of Vertically Integrated Liquid, Maximum Expected Hail Size, tracks of circulations derived from Doppler velocity data, composite reflectivity, and 30-to-60 minute forecast reflectivity fields, among others. These products, which have a spatial resolution of approximately 1 km by 1 km, are generated every two or five minutes within the WDSS-II. The WDSS-II system provides the images in GeoTIFF format which may be imported into most Geographic Information Systems software including Google Earth.
View Google Earth real-time data. (
.kmz
file, requires Google
Earth 
for
viewing)
Radar Severe Weather Case Studies
Experimental algorithms and displays developed at NSSL were used in the analysis of a number of classic, as well as unique, severe storm case studies from a radar and warning perspective.
WDSS-II Proof-of-concept tests at NWSFOs
The NSSL Warning Applications Group (SWAT) has a long history of collaboration with NWS forecast offices (NWSFOs) in the development and testing of new warning decision-making tools. During the 1990s, NSSL conducted many proof-of-concept tests at NWSFOs in more than a dozen states to examine the applicability of experimental severe weather detection algorithms during warning operations. Since 2002, SWAT has continued to conduct these experiments, using WDSS-II tools in select offices. Additionally, SWAT experimental products are available online for the conterminous United States (CONUS) at http://wdssii.nssl.noaa.gov. As NSSL's primary customer, understanding the NWSFOs' operational requirements and developing techniques to improve warning accuracy and services are paramount. Our interactions with NWSFOs serve several purposes:
- To determine if the algorithms are region-independent by conducting them in different parts of the country
- To test the long-term robustness of the algorithms by running them continuously for long-periods under varied environmental conditions
- To have algorithms evaluated qualitatively by NWS personnel through feedback questionnaires
- To be able to assist NWS staff in collecting real-time verification data and conducting post-storm damage surveys to enhance the quality and quantity of ground truth
- To gain feedback from operational warning meteorologists on the utility and effectiveness of the new WDSS-II concept and its enhanced algorithm and product displays before consideration of their inclusion in official NWS operational systems (e.g., AWIPS, ORPG).
- To gain feedback on the utility of the WDSSII display. Needed additions and enhancements will be acquired from the NWS meteorologists via post-shift questionnaires and by observations of real-time operational use by NSSL staff. This feedback is very important to help NSSL determine how the algorithm output will be displayed on NWS operational systems.
- To provide operational experience to NSSL meteorologists and developers during real-time warning situations in order to better understand user requirements.
- To offer operational meteorologists a first-hand experience with next-generation algorithms and product display concepts.
- To foster collaboration between NSSL scientists and operational meteorologists.
