The Demonstration Division evaluates promising new atmospheric observing technologies developed by the Office of Atmospheric Research (OAR) laboratories and other organizations and determines their value in the operational domain. Activities range from the demonstration of scientific and engineering innovations to the management of new systems and technologies. Currently the division is engaged in five major projects:
NOAA Profiler Network
The division manages, operates, and maintains the NPN comprising the Profiler Control Center (PCC) in Boulder, Colorado, and 35 tropospheric wind profilers located mostly in the central United States (Figure 26). The NPN wind profilers are upward looking, sensitive 404- and 449-MHz Doppler radars that measure the winds above the profiler site. They are specifically designed to measure vertical profiles of the winds. Typically, once an hour, the wind profilers produce a vertical "stack" of winds from near the surface to about 53,000 feet. They can operate in all weather conditions, including cloudy and precipitating weather. The radars are sensitive enough to detect fluctuations in refractive index caused by the turbulent mixing of air with slightly different temperature and moisture content. The resulting fluctuations are used as a tracer of the mean wind in the clear air. Data are sent in real time from each profiler to the PCC for quality control, processing, and distribution. All profiler sites are augmented with surface observing systems that provide basic surface meteorological observations, and are equipped to measure precipitable water vapor using commercial GPS receivers. In addition, RASS temperature profiling in the lower troposphere is provided at one-third of the sites.
NPN Data Distribution, Usage, and Collaboration
In conjunction with the National Weather Service (NWS) modernization, NPN data are distributed to all NWS forecast offices, government and university atmospheric researchers, private meteorologists, the National Centers for Environmental Prediction (NCEP) including the Storm Prediction Center, and foreign agencies responsible for weather prediction via the World Meteorological Organization (WMO). In addition, NPN data are available to everyone on the World Wide Web in real time, and can be accessed by such users as local and long-duration balloonists, cross-country aviation racers, and soaring enthusiasts.
Forecasters and the research community routinely use profiler data to monitor the spatial and temporal evolution of atmospheric weather features to provide improved weather forecasts. Specific examples include:
The division also supports other federal agencies and foreign governments in their efforts to develop and implement wind profilers, as follows.
NOAA Profiler Network
Staff continued to operate and maintain the NPN, and to supply upper-air and surface observations to a wide range of users, including NWS forecasters and numerical weather prediction models. The NPN profiler sites are routinely sending data to the Boulder Profiler Control Center. The data from each site are transmitted to the Profiler Hub computer system within the PCC for processing and quality control. The hourly averaged profiles are then sent to the NWS Telecommunications Gateway for distribution to the regional NWS forecast offices and to NCEP. The datasets continue to be made available to about 150 universities, private sector subscribers, the government research community, WMO, and they are available on the Web. All six-minute and hourly averaged profiler data are archived by the National Climatic Data Center.
Increase in Delivery of Hourly Data to Customers – The availability of hourly winds to the NWS remained high for 1999 (Figure 27). This high availability follows the completion of previous significant enhancements to the power amplifiers, antennas, and lightning suppression. The decrease in the availability of hourly winds can be seen during the spring and summer, compared to slightly higher availability during the fall and winter seasons. This pattern, detected every year since the NPN became operational, can be attributed to increased lightning activity causing commercial power problems and profiler hardware damage, and site air conditioner failures.
New Profiler Locations – The profiler located at Wolcott, Indiana, was brought into full operation last year with the addition of a lightning protection and grounding system. After some initial startup problems, the system is now producing reliable data for the NPN. This profiler was relocated from near Homer, Alaska, after the installation of three new 449-MHz profilers in Alaska. The profiler located at the old Lockheed Martin test facility in Bloomfield, Connecticut, has been relocated, installed, and upgraded to a 449-MHz profiler. It is now at Hancock International Airport in Syracuse, New York, also the new location of the Lockheed Martin facilities supporting the NPN program. The system supports the NWS Eastern Region operations, particularly in forecasting lake-effect events.
New NPN Boulder Facility – During April 1999, the division, along with the rest of FSL, relocated to the new NOAA Boulder facility, the David Skaggs Research Center (DSRC), collocated on the Department of Commerce campus. With in-depth planning, coordination, and the implementation of partial system redundancy, the loss of profiler data transmitted to NWS was minimal. A very dedicated, hardworking staff moved the Hub computer and communications systems and quickly got them back online. Uptime for the Hub and communications the week before, during ,and after were 99.4%, 93.0% and 93.5%, respectively. In late 1999, the computer facility's power was upgraded by connecting it to the building's diesel-powered emergency power grid. Since the NPN is considered "mission critical," it was necessary to ensure that date-related problems would not affect the ingest and distribution of profiler data as the country moved into the new millennium. No complications occurred during the crossover. The connection to emergency power, along with a new uninterruptible power supply unit, has also allowed the flow of data during times of local or building power problems, and thus continues to increase the reliability of data.
Operational Use of the NPN – On 3 May 1999, one of the largest tornado outbreaks occurred across west-central Oklahoma and southern Kansas, killing 48 people and causing over $1 billion in property damage. The NPN system at Tucumcari, New Mexico, played a very crucial role by providing information to the NWS Storm Prediction Center that contributed to the their ability to issue timely warnings of the impending outbreak. Without the early warnings, surely more deaths would have occurred. Following the outbreak, the NWS performed a "Service Assessment." This report of facts, findings, and recommendations listed as Recommendation 1: "The NWS should make a decision on how to support the existing profiler network so that the current data suite becomes a reliable, operational data source."
In concert with NWS headquarters, the Demonstration Division performed an in-depth analysis of NPN performance to identify specific areas to increase data availability in an affordable manner. Two areas were targeted: repair response time and downtime due to site power interruptions. The NWS Central and Southern Regions will increase efforts to improve Electronic Technician response time to perform remedial repair of NPN sites. The division will accelerate its internal program to provide a "remote breaker reset" capability at each site. Work is proceeding on a design that would permit reset of the main power circuit breaker via voice telephone. This capability would eliminate the need to visit the site to simply reset the circuit breakers. NPN power has always been problematic due to the remote locations of the systems. To alleviate problems resulting from loss of power caused by lightning and power surges, an auxiliary power service box (Figure 28) was designed and a prototype was built and successfully tested.
NOAA Profiler Network
The division will continue to operate and maintain the NPN, improve the quality and quantity of data for operations and research, and constantly monitor the performance of the profiler network. The main focus will be on increasing data reliability. During the summer of 2000, the remote breaker reset function will be installed. All NPN sites, including the three Alaska profilers, will receive this critical upgrade. Additionally, a prototype of a surge protection and lightning protection system will be built and tested.
The Neodesha, Kansas, profiler site will be relocated because of a change of ownership and its unsuitability for RASS operations. A new site will be found nearby, and site preparation and system relocation will be undertaken. It is hoped that loss of data from this relocation will not exceed two weeks. This field activity should be completed by early fall.
Modernized data processors will be installed at selected sites to replace the aging micro-VAXes, which are becoming more difficult to maintain due to obsolescence. The replaced micro-Vaxes will then be utilized as spares for the NPN.
In cooperation with NWS staff, several features will be implemented to allow earlier notification of loss of data delivery. A Web-based feature will be included that will allow individual NWS forecast offices to initiate repair actions during standard business hours.
All of the above activities should result in even better delivery of data to NPN customers.
Radio Acoustic Sounding System (RASS) for Temperature Profiling
The original concept for an operational profiler network envisioned the Doppler radar profiler as part of an integrated upper-air remote sensing system capable of measuring winds, temperature, and humidity. The Demonstration Division is involved in an effort to achieve this goal. Progress includes the addition of Radio Acoustic Sounding Systems (RASS) for temperature profiling in the lower troposphere and GPS water vapor systems for moisture measurements to the NPN sites.
Although the temperature measurements produced by the RASS-equipped profilers are accurate to better than 1oC to the uppermost limit of their coverage, the altitude to which they can measure is somewhat limited. RASS measurements with the 404-MHz profilers typically extend up to 2.5 to 4 km above the ground. In general, the velocity of the lower tropospheric wind limits the maximum height coverage of RASS by advecting the acoustic signal outside the radar beam. To minimize this, RASS acoustic sources are placed at each of the four corners of the wind profiler antenna.
Investigations continue on the impact of additional acoustic sources located upwind of the profiler site. Seven NPN sites in the central United States have RASS temperature profiling capabilities, including Platteville, Colorado; White Sands Missile Range, New Mexico; Hillsboro and Haviland, Kansas; and Vici, Purcell, and Haskell, Oklahoma. As shown in Figure 29, each RASS-equipped site has four acoustic sources (crown-like structures) that are located inside the antenna field fence near the corners of the wind profiler antenna. This arrangement provides security for the acoustic sources and allows for some lateral advection of the acoustic signal above the profiler site due to the local winds.
Maximum RASS height coverage of 5–6 kilometers above the ground typically occurs during periods of very low wind speeds (minimum acoustic advection) in the low to midtroposphere. Substantially greater than normal height coverage has been observed with stronger winds, when the winds reverse direction with height. In this case, the acoustic signal is first advected away from the site by lower altitude winds, and then with a 180o reversal of the wind direction at higher altitudes, the acoustic signal is advected back over the profiler site resulting in increased RASS height coverage.
During periods of strong low-level winds, a decrease in the RASS height coverage has been noted. This phenomenon is generally confined to the RASS profilers located in Oklahoma and Kansas during the spring and early summer months. A nocturnal low-level jet often develops during these times with southerly winds in excess of 25 m s-1 that rapidly advect the acoustic signal away from the profiler site as the sound waves propagate vertically. The propagation speed of the RASS acoustic signal is approximately 330 m s-1. Therefore, about 6 seconds of propagation time is needed in order for the acoustic signal to reach an altitude of 2 km above the ground. During this time the RASS acoustic signal would be laterally advected 150 meters in a 25 m s-1 wind, and possibly outside the profilers region of sensitivity. Ongoing experiments are being conducted at Platteville, Colorado, and Purcell, Oklahoma, to investigate this effect with acoustic sources placed 35–140 meters upwind of the profiler sites. Typical improvements of 500–1,000 meters in the RASS height coverage are observed when the 70–140 meters upwind acoustic sources are activated.
The Demonstration Division will continue to operate and maintain the RASS-equipped profilers which have operated for nearly 10 years. Experiments will continue at Platteville and Purcell to investigate the optimum acoustic source locations (distance upwind) and acoustic output power. With the upcoming move of the Neodesha, Kansas, profiler to a new nearby site, plans include eventually adding RASS at that site. Improvements are expected in the quality control of RASS data, primarily during heavy rain events. The presentation of RASS data on the division Web page (http://www-dd.fsl.noaa.gov/profiler.html) will be enhanced with contours of specific temperatures.
Boundary Layer Profilers
Low-power profilers that measure winds and temperature in the boundary layer have begun operating in greater numbers around North America in recent years. They primarily support air quality measurements and meteorological research programs. There are about 25 Boundary Layer Profilers (BLPs) currently operating and providing data to the Profiler Control Center. The division is working in cooperation with other agencies to acquire BLP wind and temperature data which are processed into hourly quality-controlled products, and ultimately distributed along with products from the NPN. The data from these profilers have applications to numerical weather prediction, subjective weather forecasting, and air quality research and monitoring.
A Boundary Layer Profiler Hub processing system has been under development within the Demonstration Division for several years. The system includes the basic capabilities required to ingest BLP wind and temperature data, apply automated quality control, monitor the quality of the data, and distribute products in real time. A key feature of the processing system is the ability to combine data from disparate radars into homogeneous products. It also allows for the rapid addition of new sites as quickly as arrangements can be made with individual agencies that provide the data.
The automated quality control algorithms used for the NOAA Profiler Network wind and temperature data were adapted and extended for use with the BLP data. Procedures used to subjectively monitor the quality and reliability of NPN communications and data were also adapted for the BLPs. Real-time data acquired from the profilers arrive via the Internet or dial-up communications. Input formats include several versions of primarily two different formats used by the BLPs. The BLP Hub ingests the data, which are then converted to a single format. The data are being distributed to users in FSL, and have been incorporated into the Mesoscale Analysis and Prediction System (MAPS) and Local Analysis and Prediction System (LAPS) data assimilation systems. Data are also available through the division Web page in graphical displays and numerical text files.
In 2000, staff will continue to operate the BLP Hub and use it to acquire data of opportunity, and quality control these data and disseminate them to users through the Web.
GPS Water Vapor Demonstration Network
The purpose of ground-based GPS water vapor observing systems that comprise the GPS Water Vapor Demonstration Network is three-fold. The systems are used to: 1) demonstrate an operational surface-based integrated precipitable water vapor monitoring system using the GPS, 2) facilitate an assessment of the impact of this observing system on weather forecast accuracy, and 3) investigate alternative GPS data acquisition and processing techniques to derive more information about the vertical moisture structure of the atmosphere.
As identified by the U.S. Weather Research Program, enhancements in the ability to monitor atmospheric water vapor are essential for improved weather forecasting, climate monitoring, and research. Nowhere is the need greater than in operational cloud and precipitation forecasts, but the lack of timely water vapor data has severely limited progress in this area. The satellite GPS, designed primarily for military navigation, positioning, and time transfer, provides a largely unanticipated opportunity for meteorologists to measure atmospheric moisture inexpensively – under all weather conditions and with great accuracy. International interest in GPS meteorology grew dramatically in 1999, and advanced programs to exploit this technique are underway on five continents. Japan and Western Europe, especially interested, are implementing many aggressive programs to exploit this promising new observing technique. In extending its role to help NOAA meet its mission, the GPS-Met Observing Systems Branch has been tasked with demonstrating the utility of GPS for improved short-term weather forecasting. It is also investigating new GPS data acquisition and processing techniques to provide significantly improved upper-air moisture observations.
The division continued to deploy GPS water vapor systems with the goal of having a demonstration network of approximately 200 sites in place by 2005. Implementation of the network is made possible by a collaborative effort between FSL and NOAA's National Geodetic Survey and National Weather Service, Scripps Institution of Oceanography, the University of Hawaii, the U.S. Coast Guard, and the Federal Highways Administration. In recognition of this effort, members of the GPS team were awarded the Vice President's Hammer Award for their contributions to the development and implementation of the Maritime Differential GPS Service and Nationwide Differential GPS Service. The number of sites in the network grew to 58 in 1999 (Figure 30), primarily through the installation of GPS and surface meteorological sensors at the 19 remaining NPN sites. The division also installed a GPS water vapor system at the NWS Forecast Office at Blacksburg, Virginia, and incorporated into the demonstration network four Nationwide Differential GPS sites owned by the U.S. Department of Transportation. Typical sites in the developing GPS Integrated Precipitable Water Vapor (GPS-IPW) Demonstration Network are illustrated in Figure 31.
GPS and surface meteorological sensor data availability from the demonstration network averaged 97% and 94% respectively for the year. The mean time between failures for the network was about 2,800 hours (117 days) and the mean time to repair was less than 1 hour. Most failures appear to result from electrical surges associated with thunderstorms. A significant number of GPS receivers failed soon after initial installation, due most likely to a previously undetected firmware problem, which is being corrected by installing a new version of firmware at all sites.
Major changes to the Web GPS data interface were implemented that facilitate the viewing and retrieval of GPS and surface meteorological sensor data via an improved Web browser, available at http://www-dd.fsl.noaa.gov/gps.html. By late 1999, the division was providing GPS and surface meteorological data from 58 sites to forecasters, researchers, and modelers about 16 hours after the end of the day. FSL's Forecast Research Division (FRD) assimilated these data into MAPS as part of an ongoing assessment of the sensitivity of numerical weather prediction (NWP) models to GPS data and the impact that these data have on weather forecast accuracy. Results from the first 18 months of continuous parallel NWP model runs with and without GPS indicated a systematic improvement in forecast accuracy through the addition of these data. The improvement was seen despite the fact that data from only an average of 15 stations routinely went into the models during this period because of a timing problem. In general, the largest improvements were seen during active weather events usually associated with heavy precipitation.
A detailed analysis of data acquired during the 1997 water vapor intensive observing period at the Department of Energy (DOE) Atmospheric Radiation Monitoring (ARM) Cloud and Radiation Testbed facility near Lamont, Oklahoma, was completed this year. The analysis, carried out in conjunction with DOE and the University of Wisconsin at Madison, revealed that the accuracy with which column-integrated precipitable water vapor could be measured by different observing systems, including GPS-IPW, was consistent to within 5% (Figure 32). This result has importance for very high accuracy measurement applications such as climate monitoring, quality control for radiosonde integrated moisture soundings, and the correction of infrared astronomical observations.
Figure 31. Typical sites comprising the NOAA GPS Water Vapor Demonstration Network. Sites belong to FSL and other NOAA organizations and other federal agencies.
One of the major sources of error in high accuracy geodetic surveying comes from temporal and spatial variability of the constituents of the troposphere, primarily water vapor. NOAA's National Geodetic Survey (NGS) performed an analysis last year of techniques to mitigate the impact of this variability on GPS measurements of position, especially height. In collaboration with FSL, NGS determined that information provided by numerical weather prediction models could be used to estimate and correct for these errors with very good results. NGS also determined that the best results were achieved when the signal delays measured at FSL GPS-IPW sites were used, thus demonstrating the potential of using meteorological models that assimilate GPS-IPW data for improved survey accuracy.
Significant progress was made in the generation of the improved GPS satellite orbits needed to calculate water vapor. In addition to improved orbit accuracy, the time required by the International GPS Service for Geodynamics (IGS) to produce these orbits was reduced from 24 hours to about 8 hours. This will make it possible to reduce the latency of daily GPS-IPW solutions from 36 hours to about 13 hours after the end of the day.
Progress was also made on the generation of "ultra-rapid orbits" that are calculated hourly with about one-hour latency. This was successfully demonstrated by our university partner at Scripps Institution of Oceanography using hourly data acquired from a subset of the IGS global tracking stations.
Agreements with the U.S. Coast Guard (USCG) and U.S. Department of Transportation Federal Highways Administration (FHWA), expected to be completed in Fiscal Year 2000, will permit FSL to add USCG and FHWA sites to the GPS Water Vapor Demonstration Network. They involve the installation of surface meteorological sensors at 43 additional USCG sites and U.S. Army Corps of Engineers Maritime Differential GPS (DGPS) sites, and at an additional 76 FHWA Nationwide Differential (NDGPS) sites. The sites scheduled for integration in 2000 are identified in Figure 33.
Last year marked significant changes to the FSL GPS-Met data processing paradigm. These changes, listed below, are expected to be validated and implemented in 2000:
Testing and validation of ultra-rapid orbits for GPS-IPW will be conducted, and if all goes well, near real-time data processing will be implemented in the spring of 2000. GPS-IPW data should be available (initially) every hour with about a 30-minute latency.
Techniques to measure line-of-sight or slant-path GPS signal delays will be explored. A major source of error in slant-path measurements will be mitigated once real-time quality control of ultra-rapid orbits is implemented. The next step is to mitigate the adverse impact of signal multipath on data quality by upgrading the GPS antennas at FSL GPS sites and verifying the results. The last step is to perform an experiment to objectively assess the various sources of error in making slant-path signal delay measurements. Then it may be possible to develop strategies for dealing with these error sources operationally.
Alaska Profiler Network
The concept for a profiler network in Alaska dates back to December 1989 when the Mt. Redoubt volcano erupted sending volcanic ash 38,000 feet into the atmosphere. The ash caused extensive damage to a KLM 747 aircraft, but good piloting prevented a major disaster. The NWS Alaska Region management quickly investigated ways to forecast aviation hazards resulting from airborne volcanic ash. The profiler was chosen as a tool for predicting ash trajectories because it can measure wind profiles up to 53,000 feet and provide updates every 6 minutes. NOAA redirected one of the profilers coming off the production line for installation near Homer, Alaska. Operational within months, the Homer profiler aided NWS in monitoring winds during subsequent volcanic eruptions, although sea and ground clutter caused by nonoptimum siting limited its full potential.
Since then, the Demonstration Division, in collaboration with DOD and NOAA's Environmental Technology Laboratory, implemented a plan for the deployment of three new profilers in Alaska with better siting. Data from these profilers are used to warn of hazards to commercial aviation in the event of a volcano eruption, and to support other NWS activities in Alaska. When planning for the Alaska project began, the U.S. Air Force was using a 404-MHz NPN profiler in support of launch operations at Vandenberg AFB, California. The profiler's frequency needed to be changed to 449 MHz to accommodate DOD operational requirements. NOAA and DOD agreed to jointly develop 449-MHz profilers by combining their similar schedules, requirements, and funding. The Vandenberg AFB profiler upgrades included a profiler data processor supplied by Radian International and a new 449-MHz antenna and power amplifier supplied by Lockheed Martin. The Alaska profilers use existing NPN data processing systems as well as newly designed antennas and power amplifiers. This collaborative effort further reduced logistics costs through the joint use of certain spare parts.
Now that the three Alaska systems are completed and the Vandenberg system upgraded, these units are integrated with the existing NPN and use similar procedures for operations, maintenance, and logistics support. Data from the Alaska profilers flow to the NWS operations and to other recipients of NPN data.
The Alaska 449-MHz Profiler Network became operational last September. Division staff and Lockheed Martin engineers completed the acceptance of the three Alaska profiler sites at Talkeetna, Glennallen, and Central (Figures 34, 35, and 36). The sites were equipped with GPS precipitable water vapor systems, surface meteorological sensors, as well as FTS-2000 data communications. After experiencing problems with the transmitter and antenna, Lockheed Martin reworked the power amplifier and the antenna dividers. The reworked components were installed in August 1999. Division staff performed final acceptance tests and then brought the three sites online. Data from all three sites are transmitted to the Profiler Hub, distributed using the same process as for the NPN, and are also made available on the division home Web page.
The Glennallen profiler has operated over 90% of the time since September 1999. The antenna structure experienced some movement due to permafrost, and the problem was corrected by adding fill material and installing adjustable metal plates to the antenna's concrete piers that will allow up to 12 inches of height adjustment. When the antenna structure experienced a slight movement last year, the metal plates were adjusted to correct for the movement. Additional landscaping and groundwork around the perimeter of the site are expected to alleviate the permafrost problem.
The Talkeetna profiler has operated over 71% of the time since last September. Waiting for replacement parts attributed to a long downtime. During the winter the data quality at Talkeetna deteriorated significantly. Over three feet of snow and ice had accumulated on the antenna. NWS electronics technicians from Anchorage were dispatched to the site to melt the snow (Figure 37), and afterward the data quality returned to normal (Figure 38).
The Central profiler has operated over 53% of the time since last September. The remoteness and inaccessibility during the winter months caused a substantial downtime for this site. The same data quality degradation that happened at Talkeetna was also evident at Central, where about the same amount of snow fell. The Glennallen profiler received only small amounts of snow, with no data quality degradation.
Division staff will continue to operate and maintain the Alaska network. The NWS intends to use the three 449-MHz Alaska profilers as operational systems. A Memorandum of Agreement is being prepared with the NWS headquarters, the NWS Alaska Region, and the Office of Oceanic and Atmospheric Research/FSL for the implementation, support, maintenance, and operation of the profilers. The profilers will be transferred to NWS, with headquarters providing coordination and support to the Alaska Region for operational systems. FSL will operate the profilers as part of the NPN, and the Alaska Region will assume responsibility for onsite maintenance, logistics, and funding of these systems.
World Wide Web Products
The Demonstration Division acknowledges the growing importance of the Internet and Web and enthusiastically supports FSL-wide policy of finding creating ways to use this technology more effectively to accomplish its mission. The Software Development and Web Services Branch responds to the evolution of the Web as a dynamic tool for improving data access to the many customers of the NPN.
Since the division Website underwent an overall upgrade two years ago, many products have been added to enhance the Web pages, such as:
Other profiler products available on the division Website include a zoom map of the United States that shows the status of NPN profilers and serves as an interface to all division products. The addition of time animation to many of these Web products enhances their utility for meteorologists and other users. Most of the Web page products take advantage of Java client-server technology, and for users with computers that are not Java-enabled, static images are made available. Additional displays and availability of raw data and quality control information from BLP and NPN profilers and other collocated instruments are now provided. (Figure 39 is a screen from the division Website showing statistics on the NPN.) In addition to graphical products, the capability to obtain profiler data directly has attracted forecasters and researchers from universities and other collaborating organizations to these Web pages first for profiler data. Improvements to these Web pages garnered the developers an FSL award for the "Most Improved Website."
Improvements will continue to be made on the Demonstration Division Website. Also, a cooperative effort with NWS is planned which would allow NWS electronics technicians to diagnose profiler faults via a division Web page. Another plan involves the National Climatic Data Center (responsible for archiving wind profiler data) to collaborate in producing Web-based tools for viewing historical wind profiler data.