12 January 2019, Blog No. 1
Per NOAA's Radar Operations Center1, three WSR-88D (NEXRAD)2 Doppler weather radars operated by the National Weather Service lie within the confines of North Carolina: KRAX in Clayton, KMHX in Newport, and KLTX in Shallotte (pronounced sha-LOAT). Eight Doppler radars have coverage areas that overlap into North Carolina: the three aforementioned sites plus KFCX in Roanoke, VA, KAKQ in Wakefield, VA, KCAE in West Columbia, SC, KGSP in Greer, SC, and KMRX in Morristown, TN (Fig. 1). TRDU and TCLT are Terminal Doppler Weather Radars (TDWRs) near the RDU International Airport and Charlotte-Douglas International Airports, respectively, but are not included in this analysis because they do not possess the same capabilities of the NEXRAD system radars.
Thus, we are limited to how much a radar can "see" based on its proximity to a given location. When the initial round of Doppler radars were installed in the late 1980s/early 1990s, no radar was installed in the western half of North Carolina. Why? I cannot find an answer to that - I can only conclude that the National Weather Service designed the NEXRAD network in the manner that they sought best for warning significant weather. However, as a result, the NC Foothills, Metrolina (greater Charlotte metro area), and the Interstate 85 corridor as a whole up through the Piedmont Triad region rely on out-of-state radars for all of their area's coverage.
We can play around with this plot above and visualize how the problem of the radar beam missing near-surface weather arises. Below in Figure 3, I superimposed the Charlotte, NC (pop. 859,000 est. 2017) skyline and exact point that Salisbury, NC (pop. 33,800 est. 2017) lies in distance from the KGSP radar in Greer, SC. In Figure 4, I repeated the process except for instead superimposing the Winston-Salem, NC (pop. 244,600 est. 2017) skyline as a depiction of distance from the KFCX radar in Roanoke, VA:
Eight Doppler radars may seem like a lot for the state, but keep in mind that all of these radars work in tandem to produce as much coverage as possible over the entire 48,617 square miles3 of land area in the Old North State. The reason that so many Doppler radars are needed across the nation is due to an unchangeable property of the earth surface itself - its curvature.
As the radar beam shoots in a perfect line, the surface of the earth slowly curves away from the beam and samples higher up in the sky as a consequence. Think of taking an exit off of the interstate where the interstate is the radar beam and you are the earth surface slowly diverging away from the busy highway as you drive forward on the exit ramp. The same principle is happening in a vertical manner in the atmosphere. For simplicity, we can model this in a basic plot where the x-axis (earth surface) is flat and the radar beam instead curves upward with increasing distance (Fig. 2) (this is done because we cannot easily plot on a curved graph axis).
Figure 1. North Carolina map with eight NEXRAD radar locations in violet, mock range rings with radius of approximately 70 miles in blue, and TDWR locations in light green. County borders are outlined in brown. Source: NOAA's Weather and Climate Toolkit 4.2.0
Figure 2. Standard NWS Volume Coverage Pattern4 precipitation scan strategy (VCP 215) plotted using MATLAB 2017b. Radar would be located in the bottom left corner. Height on y-axis and distance from radar on x-axis, both in miles.
Figure 3. Not to scale. Same plot as above with general range of Charlotte area depicted by skyline overlay and location of Salisbury, NC from KGSP radar. Notice how the radar beam is sampling at its lowest at 0.75 miles above the ground over the Charlotte metro region, and over 1 mile above ground for Salisbury.
Figure 4. Not to scale. Same plot as above two with general range of Winston-Salem area depicted by skyline overlay and location of Salisbury, NC from KFCX radar. Radar coverage is better here over Winston-Salem proper, but still sampling just shy of a mile above the surface over Salisbury.
So what does it matter if the radar beam is sampling a little bit higher in the sky? Well, tornadoes are very shallow phenomena (< 1/2 mile high) that the radar beam would be overshooting. Over Salisbury, severe winds would be hard to monitor because the beam would be sampling winds that are too far above the ground to actually make it to the surface. This creates a disconnect between what the radar sees and the sensible weather at the surface. I took a screen grab a while back from NCEI's NEXRAD Interactive Map Tool5 (currently down because of the government shutdown) that depicts these radar coverage holes in a different manner (Fig. 5):
Figure 5. Screen grab from NCEI webtool showing NEXRAD coverage. Excellent coverage is green, good coverage is yellow-green, fair coverage is blue, and unshaded areas indicate poor to no radar coverage. Salisbury, NC sits right in the radar coverage hole. Nags Head and Kitty Hawk, NC also fall in this range.
When we zoom into those zones with poor radar coverage (Fig. 6), we can identify the communities that are then inherently at higher risk of a quick spin-up tornado or severe winds going undetected or at least detected too late for weather warning issuance:
Figure 6. Zoomed in map of Salisbury poor coverage zone (left) and Outer Banks poor coverage zone (right)
Using the nifty site populationexplorer.com6, I mimicked the shape of the poor coverage zones and estimated a population of approximately 300,000 in the Salisbury poor coverage zone, and approximately 60,000 permanent residents in the Outer Banks, not including the significant tourism population. That is 360,000 people with that inherent disadvantage of being under less than optimal radar coverage. Yes, the TDWRs do exist and fill in some of the area in the Salisbury zone, but these smaller scale radars do not have the same capabilities, such as dual-polarization7, of the WSR-88Ds installed for the NEXRAD program.
So why not just go and install a new WSR-88D radar near Salisbury? What is the ubiquitous limitation that exists with nearly every facet of scientific research and advancement? Say it with me: "money!"
Per an August 1999 Forecast International technical document8, WSR-88D radars cost between US $2.25 - 2.5 million a pop excluding siting and maintenance costs, and that was the pricing in 1999 dollars. Just shy of twenty years later, the National Weather Service is in the midst of securing funds for a round of maintenance and tweaks tabbed as "SLEP" or a service live extension program (Government Technology, ¶8). The product of the SLEP will suffice until a new phase of weather surveillance technology is implemented at an undefined time in the future. With the current administration, the completion of securing these funds and the progress of any maintenance is unclear.
What does remain clear is that substantial gaps in adequate radar coverage are present over populated areas in North Carolina. Upon a glance, it appears as if the Salisbury poor coverage zone is perhaps the most populated area without at least fair radar coverage in the contiguous United States of America. While further research is warranted on this topic, each passing storm system begs the question of whether installing a Metrolina-area Doppler radar is a worthwhile investment.
NOAA's Weather and Climate Toolkit 4.2.0
Government Technology Article published 17 July 2015
Terminal Doppler Weather Radars
Supplemental WSR-88D Information from NWS Northern Indiana