Q12. Has the FAA diverted aircraft as a result of wind turbine clutter?
Q1. How far away from the NEXRAD should we site a wind farm? Do you
have a benchmark distance? If so, what is it?
REPLY: Our benchmark for requesting consultations with developers to mitigate potential interference is when a proposed
wind farm's turbine blades protrude into the NEXRAD's radar line of sight (RLOS). The benchmark RLOS is based on the propagation of the radar beam in the
Standard Atmosphere transmitted at the lowest scanning angle (0.5 degrees above horizontal). The beam does not propagate in a straight line
but follows the curvature of the earth at approximately 4/3 the earth's radius. This curving of the beam (called refraction) is due to changes
in atmospheric density. The wind turbine is considered to be in the RLOS when the bottom of the main radar beam intersects the highest
point reached by the turbine blades. The distance to the edge of the RLOS from the radar depends on the surrounding terrain and the
maximum height of the wind turbine blades. For level terrain, the RLOS would extend out about 23 nm from the radar for a typical
400ft turbin. However, the RLOS is different for every NEXRAD location primarily due to varying terrain, and there is not one benchmark distance.
Q2. What is the "Radar-Line-of-Sight" and why is that important?
REPLY: The radar line of sight/radar beam width can be considered analogous to the beam of light coming from a flash
light. Most of the energy of the flashlight, just as with the radar, is in the beam of light/radar beam. In radars this is the distance
between the "half power" points or where the energy in the beam is down 3 dB from that at the center of the beam. For the NEXRAD the
beam width is approximately 1 degree. As the beam propagates away from the radar, its width increases. For NEXRADs, at 60 nm from the radar
beam is approximately 1 nm wide. Obstacles in the radar line of sight can block the radar signal and reduce the ability of the radar to
see targets further downrange. The figure below is a depiction of the radar line of sight.
Main beam/radar line of sight is defined by half-power points
Q3. Does the RLOS ever change?
REPLY: Yes. The actual RLOS (not the RLOS based on the Standard Atmosphere, but the RLOS based on actual day to day weather) changes
during the day as a result of temperature and humidity changes. It also changes as fronts pass or with nearby thunderstorm outflows.
Typically after sunset, the surface temperature cools causing the radar beam to bend more towards the earth's surface. This is called super-refraction
or "ducting". The net result is that wind farms that are normally out of the RLOS may be in the RLOS at certain times of the day and during certain
weather conditions. So, even if wind farm developers site their projects outside the benchmark RLOS, the weather forecasters will
occasionally "see" the wind farms on the radar imagery.
Q4. How can NEXRAD systems "see" wind towers/turbines when I can't
visually see the radar from the wind farm?
REPLY: The path that emitted radar energy (i.e., the radar line of sight) takes, depends upon
atmospheric density. Density differences are caused by variations in pressure, temperature and moisture.
In a "standard atmosphere" representative of the atmosphere on a day with enough wind to mix the lower
atmosphere well, the radar beam takes a path that is approximately 4/3 of the Earth's radius. This bending is
called "refraction." So, the NEXRAD, like other radars, can "see" targets well beyond the optical line of sight.
The figure below is a depiction of the beam's path in a standard atmosphere.
Q5. How powerful is the NEXRAD's transmitted microwave energy?
REPLY: The NEXRAD radar transmits a pulsed signal at 750 kilowatts. The maximum time-averaged
power (transmitting and listening periods) is about 1500 watts.
Q6. Why can't the NEXRAD be reprogrammed to filter out returns from
wind turbines?
REPLY: The NEXRAD/s clutter filter scheme only removes clutter that is stationary,
such as buildings, trees, and terrain. Unfortunately, both precipitation and wind turbine blades are
moving and the filter is not applied to them. Trying to filter out moving blades will inevitably alter
how the radar sees real precipitation. Here's why. A single radar volume sample (gate) at 30 miles from the radar is approximately
a square kilometer. Thus, for a typical wind farm, the radar may receive reflected energy from many turbines
within that gate, each with multiple rotating blades. These numerous rotating blades appear similar to
precipitation, which is also made up of numerous distributed moving targets. Yes, there are fewer blades
than raindrops within a sample volume, but the blades make up for their smaller numbers by reflecting significantly
more energy back to the radar. However, the radar has no way to determine the number of targets it is sampling
within a particular gate. Also, the reflected energy is constantly changing as the blades change their pitch and
orientation relative to radar, with some blades moving towards the radar, some moving away, and some not appearing
to move at all (perpendicular). This is analogous to the movement of precipitation within a volume sample. Studies are underway
at the University of Oklahoma and other institutions to find a solution. However, at this time there is no known
way to filter out turbine blade clutter.
Q7. What is the NEXRAD Program doing to solve the WTC problem?
REPLY: The NEXRAD Program has provided research funds to the University of Oklahoma
to devise potential short-term and long-term solutions. Advanced signal processing techniques, such as non-stationary clutter filtering,
are currently being explored as sophisticated and robust solutions, but these are long-term efforts. In
addition, innovative radar designs, such as adaptive phased array antennas, are being explored as potential
solutions. Finally, knowledge-based techniques, which would exploit information (blade phase, rotation speed,
etc.) from wind turbines are being conceived. The Atmospheric Radar Research Center (ARRC) at the University
of Oklahoma plans to make use of its Electromagnetic Microphysics Laboratory (EML) for
this final concept. With the appropriate funding stream, experiments could be conducted to simulate a working turbine within the
lab, providing the ability to test knowledge-based algorithms in a controlled setting.
Q8. How close is a NEXRAD-based solution to the WTC problem?
REPLY: There may not be a NEXRAD-based solution (i.e. no signal processing solution). The simplest solutions,
such as identifying and flagging wind turbine-corrupted data are at least 5 years away and they are only partial solutions. Signal processing
solutions, in general, are very complicated and are at least 5 years away, assuming an acceptable solution can even be found.
Q9. Can't you just move the NEXRAD to a new location, or build a new one?
REPLY: Moving a NEXRAD radar is
very expensive--$1.5Million(M) to $4M—and a new
weather radar with similar NEXRAD capabilities could
be $10M depending on site acquisition costs and
other site-specific costs like radar tower height.
In general, moving a radar is not a good solution
since these radars were strategically sited to work
as a national network with proper coverage while
minimizing operating costs. Moving one radar can
affect coverage relative to surrounding radars in
the network. Given the ever increasing number of
wind farms being installed, this can quickly become
a costly and futile exercise as new wind farms
encroach on the moved radar.
Q10. Can the NEXRAD impact a wind turbine or its maintenance personnel?
REPLY: Yes, if a wind turbine is
sited very close to the radar. When wind turbines
are sited very close to NEXRAD radars, the turbines
can be adversely affected by the high power, 750 KW,
radar transmission. Within 600 ft of a NEXRAD and in
the transmitted beam, this energy can exceed the
OSHA (29 CFR Part 1910--Subpart G-Occupational
Health and Environmental Control Ch.1910.97)
threshold for occupational exposure to microwave
energy for construction, operation, and maintenance
personnel. Within 10 miles of a NEXRAD, the
microwave radio frequency field strength can cause
bulk cable interference (inductive coupling) with
the turbines electronic controls if they are not
properly shielded (MIL-STD-461D).
Q11. Has the National Weather Service ever missed a weather warning
to public, or given a false weather warning to the public as a result of the wind turbine clutter problem?
REPLY: No, not yet. The number
of wind farms close to NEXRADs is not very great at
this time and we have a relatively short experience
working with WTC. Operational forecasters can often
distinguish (WTC) from weather signals using their
experience. However, WTC can be a distraction and
can take forecasters’ time away from evaluating
developing weather. Another major concern is the
effect of these echoes on automated detection
algorithms and users (e.g. media and public) not as
experienced or used to the appearance of WTC. And,
while the WTC problem is causing relatively minor
operational impacts at this time, the expected
exponential increase in the number of wind farms
near NEXRAD radars is cause for concern. It is easy
to envision some NEXRADs becoming surrounded by many
wind farms and forecasters and other users having to
work around significantly large areas of
contaminated radar data.
Q12. Has the FAA diverted aircraft as a result of wind turbine clutter?
REPLY: Yes. The FAA has
re-routed air traffic due to false returns from wind
turbine clutter. NEXRAD data streams are fed
directly into the FAA's Weather and Radar Processor
System at Air Route Traffic Control Centers (ARTCC)
and FAA controllers use the data to route aircraft
safely around weather. ARTCCs have contacted the
NEXRAD Radar Operations Center asking about NEXRAD
radar data showing what appeared to be significant
weather that required rerouting, but pilots reported
not seeing weather in the area. This confusion
causes unnecessary and expensive aircraft re-routing
and excess fuel consumption.
