Issues & Analyses: Dark Skies And Safe Roads. Can We Have Both?  (Part 1)

Mode Shift: Supporting Sustainable, Equitable Transportation

Today’s Mode Shift column , and next week’s, take on the issue of streetlighting. This seemingly boring topic is both critical for a transportation mode shift in which the share of travel increases significantly toward bicycling, transit and walking. It is also challenging.  

Why critical? Research and the experience of many people in Amherst is very clear: to get more people to walk, bike, and bus, including at night, and to ensure they are safe when they do, we need considerably more streetlighting. I’ll provide more information next week on why, how, and where streetlighting is crucial for a mode shift. 

But why is this issue so challenging?. First, the usual: budget limitations, and a public infrastructure system that is oriented largely towards cars. Second, there is now considerable research on the importance of less lighting, to protect the night skies and the health of people and nonhumans. Along with this research there is also an increasingly powerful advocacy movement to stop light pollution, including here in Amherst.

There are ways to do both, that is, to provide lighting for pedestrians, bicyclists, and transit users, and at the same time reduce light pollution. However, there are some inevitable tradeoffs among more light, less light, and cost. For a community to provide adequate lighting for pedestrians, bicyclists, and transit users, and to protect the night sky and reduce light pollution, the community needs three things:

(1) a commitment to both sets of goals;
(2) a clear analysis of where and how they conflict;
(3) a coherent public process to find win-win solutions whenever and wherever possible, and to work through both general and site-specific tradeoffs when necessary. 

A current Amherst Town Council proposal provides a microcosm of the potential problems when these pieces are missing. 

The proposal for a new streetlights policy before the Amherst Town Council is rich with well-thought-out detail. The problem is, it is narrowly framed, and for this reason dangerous. Its performance standards aim to achieve a one-sided set of goals: “to minimize light pollution, glare, uplight, and light trespass, as well as to regulate illumination levels.” Its proponents make only limited acknowledgement of tradeoffs with safety and cost, and the proposal provides only limited, cumbersome, and non-public processes to increase lighting, or to “waive” the policy on a case-by-case basis. As written, it promises continued night-time and bad-weather injuries and deaths of pedestrians, bicyclists, transit riders, and car drivers in Amherst. It will also obstruct a mode shift toward just and sustainable multi-modal transportation in Amherst. 

However, there is potential for a win-win solution. To support a transportation mode shift, Amherst needs a new streetlight policy. With some strong advocacy we may be able to revise the proposed policy before it gets passed. Then we could meet both the goals of the current proposal and provide the lighting necessary to advance a shift toward more just, sustainable transportation.

This week’s column covers the rationale for the proposed policy,  a description of the current street lighting context, and an analysis of what it will change and not change on our streetscapes. Next week, I will focus on street lighting that would support safer walking, biking, and transit use; and how to build a policy that could support both goals. 

My strategy in this pair of columns is to focus on “arterial,” “collector,” or connector streets. These are core walking, biking, and transit corridors. I also use diagrams to illustrate and summarize technical information. For those interested in my take-home advocacy message right away, you can see the comments I wrote to the Council last week, and also my initial model for a draft revision of the policy by looking here.

Background and Rationale: The Amherst Streetlights Policy Proposal
The proposed Amherst streetlights policy aims to protect the night sky and reduce light pollution and light “trespass.” It was put together by Councilors Mandi Jo Hanneke (at large) and Ana Devlin Gauthier (District 5), in consultation with James Lowenthal of Northampton. Lowenthal is an astronomy professor at Smith College, and an educator with the International Dark-Sky Association. Considerable research shows that excessive night light —especially in the higher wavelengths of many newer LED streetlights and electronic screens—negatively impacts human circadian rhythms and health disturbs wildlife, and obscures the night sky. As Lowenthal has argued, even lights that are installed in the name of safety are often brighter than they need to be, and illuminate much wider areas and directions than needed. They often create light pollution and nuisance, and may create glare that reduces rather than aids safety. 

Unlike most policies, the proposed Amherst streetlight policy reads like an engineering society list of design specifications, evidently strongly informed by guidance from the Illuminating Engineering Society (IEA) and the International Dark-sky Association (IDA). Most of the substance is in the proposal’s Appendix, titled “Streetlight Performance Standards.”

Normally, legislative policy allows for more precision being added over time as implementation evolves.  Legislation sets goals, authorities, and some mechanisms; regulations provide details for how the legislation will be implemented by the designated authority; and protocols and practice govern the activities and locations of employees and equipment. It seems that Councilors Hanneke and Devlin Gauthier, with support from Lowenthal, decided that allowing a policy to evolve over time as standards change, or providing flexibility in implementation, are not as important as locking in place an exacting set of specifications that  cannot be altered or weakened later by regulations or in practice by town committees or staff. The inflexibility is enforced with the liberal use of the words “shall,” “shall not,” and “maximum,” mandating an absolute rigidity of implementation.

Because of the technical language, the likely impact of the policy is difficult to understand. Because of the legal language, the proposed policy would make it virtually impossible to install streetlights that don’t fit its strictures. 

I am not a lighting expert, but I did what any intelligent person should be able to do with any public policy: I read the policy and the definitions, cross-referenced them with other information sources on the web, particularly the expert sources referenced in the policy, the IEA and the IDA, and applied some trigonometry functions archived deep in my brain. I had to read a lot deeper and wider than the proposed policy’s technical appendix to understand its standards. I have posted a full layperson-friendly translation of the policy’s performance standards on the web. If I got something wrong, that’s another reason to revise the policy: to communicate its plain meaning. 

Current Street Lighting Conditions In Amherst 

Figure 1. Current conditions on a typical Amherst arterial street, as well as on many neighborhood through-streets. Both light pollution and safety lighting need improvement. Figure:Eve Vogel

Figure 1 illustrates a typical streetscape along an arterial or connector street in Amherst at night. Many of our arterials are also key pedestrian and bicycle routes. The diagram represents a roadway with shoulders and/or sidewalks on both sides. This is also a reasonable representation of many through streets in larger neighborhoods. It represents, for example, roads with significant populations of carless residents: North Pleasant north of campus, East Pleasant from Olympia Drive south, or East Hadley Road. It also represents roads that could be much more well-used by cyclists, pedestrians, and transit users at night if the roads were safer and more inviting. Think Pelham Road, Belchertown Road, West Street, Heatherstone Road, East Pleasant north of Eastman Lane, for example. Insert your own favorite.

Under current conditions, streetlights in Amherst are located mainly at intersections and special locations such as cul-de-sacs, downtown streets, near schools, and hazardous curves. The town keeps a database on its approximately 500 streetlights. Most lights are 6,300 lumens (a measure of brightness), while those at locations considered more critical—fire stations, downtown, etc.—are 27,500 or even 50,000 lumens. Most streetlights in Amherst outside of downtown are mounted on utility poles. This saves the town from having to purchase, install, and maintain its own light poles, but it requires the lights to be mounted high up. In the figures, I approximated the height of these lights as 30 feet. 

What problem do streetlights cause for night sky visibility and light pollution? As shown in Figure 1, most streetlights in Amherst have opaque tops, helping to prevent light pollution into the night sky. Still, a lot of indirect light goes up into the sky and much direct light shines into adjacent properties. Also, many of the current lights are LED lights on the blueish-white end of the spectrum, which irritates human and other animal eyes and brains. The greatest contributor to excess upward and outward lighting in Amherst is UMass, and the greatest contributor to glare is probably cars’ headlights, and many of those are on that blueish-white end of the color spectrum. The town of Amherst cannot easily control these. (UMass lighting might be a topic for a future column.) The town’s ability to reduce light pollution, glare, and blue light is more limited, hence the focus on Amherst’s streetlights.

in Amherst today, the problems for pedestrian, bicyclist, and transit rider safety come not primarily from excess street lighting but from excess darkness. 

Some excess streetlighting can be dangerous for bicyclists, pedestrians, and transit users—for example, if it shines in their eyes, or if it directs drivers’ attention to other places, or if it accustoms their eyes to bright conditions so they cannot see well in the dark. Higher-mounted lights like those on utility poles are less likely to create glare for passing pedestrians, bicyclists, bus riders, and drivers, although they may cause glare in adjacent property owners’ second-floor windows.

However, in Amherst today, the problems for pedestrian, bicyclist, and transit rider safety come not primarily from excess street lighting but from excess darkness. An additional problem is the inconsistent lighting on most routes. As shown in the figure, many sections of arterial streets and neighborhood through-streets, including sections of many important bicycle and pedestrian connectors, have only intermittent streetlights. 

Although the distances in the diagram depend on the brightness of the lamp as well as the specific height at which it is mounted, the diagram gives a good idea of the visibility of pedestrians and bicycle riders at varying distances from the light. In the figure, the pedestrian who is 50 feet from the streetlight, and the bicyclist who is slightly further, are visible and can see easily. The light is well above them, so there is no glare in their eyes, even for the pedestrian headed toward the light. 

The pedestrian 130 feet from the streetlight and on the opposite side of the road from the light probably has just enough light reflecting off her that she is visible to cars. She will be far less visible, however, on a rainy, foggy, snowy, or smoky night; if there is a bend, hill, or dip in the road that blocks direct lighting; or if tree branches on her side of the road block the light from reaching the sidewalk or shoulder where she is walking.

The bicyclist who is 200 feet from the streetlight is visible to cars only if they have lights (legally required but not universally owned), but even with a bicycle light or headlight, this rider is not able to see potholes or debris on the roadway unless their bicycle has a top-of-the-line,  extra-bright headlight, which is more expensive and less common than a regular light. If there is oncoming traffic blinding the rider with headlights far brighter than the remnant illumination from the distant streetlight, the shoulder in front of them will look pitch black (I write this from experience). Hitting an unseen pothole, storm drain, stick, or rock, or a sandy or slick stretch can cause a bad bicycle crash and injury. Similarly, of course, if a car driver headed toward the bicyclist is blinded by headlights from an oncoming car (or from poorly angled streetlights or building lights), they may not see the cyclist, even if the rider has a light–and certainly not if they don’t. Then, there may be an even worse crash.

The pedestrian who is headed to the bus stop 275 feet from the light pole is all but invisible. Even at his slow speed, he cannot see cracks or debris on the sidewalk or shoulder. These are especially dangerous if he happens to be someone who is more likely to get injured in a fall, such as an older adult with a walker or cane. If he’s walking on a shoulder and not a sidewalk, a car driver could hit him without ever even seeing him.

Bicyclists and pedestrians in dark conditions are far more vulnerable to injury or death. A pedestrian is three to seven times more likely to be killed by a car at night than during a clear day. Children are particularly vulnerable, as their small size makes them especially hard to see.

The darkness at so many points along walking, biking, and transit routes is a problem because bicyclists and pedestrians in dark conditions are far more vulnerable to injury or death. A pedestrian is three to seven times more likely to be killed by a car at night than during a clear day. Children are particularly vulnerable, as their small size makes them especially hard to see. BIPOC and low-income communities experience disproportionate burdens, often because of more dangerous roads and more limited infrastructure for pedestrians and bicyclists near where they live. If a person is hit by a car or bus going over 25 mph—a traffic speed that is exceeded on virtually all arterials in Amherst and most neighborhood through-streets—there is a greater than 50% chance they will die. And car drivers themselves are also far more likely to die in crashes at night.

In the last 10 years, five pedestrians and five drivers have died on Amherst roads.

These are not “somewhere else” statistics. In the last 10 years, five pedestrians and five drivers have died on Amherst roads. I consulted with UMass Transportation Center’s Tracy Zafian to figure out which  of these deaths were caused by low visibility. She helped me look at and interpret data from the MassDOT crash database. Three of the pedestrians were hit because it was night and the driver did not see them. One was on North East Street, near the Strong Street intersection: Two pedestrians were walking side by side on an unlit part of the road where paint striping is often faded. The driver did not see them, and the person walking on the inside was killed. One was on North Pleasant, near the North Village apartments: After a graduate student got off the bus, he started to cross the road and got hit by a driver who did not see him. (Since then, the lighting in that spot has been improved, but other spots still have similar conditions.) The third was on Massachusetts Avenue, in front of some UMass construction, where the sidewalk was closed, and again, the driver did not see the person. Additionally, two of the driver deaths were in dark, low-visibility conditions, including a tragic one last year when an Amherst teen died. On average, then, over the last 10 years, there has been one death in our town every two years at least partly because of lack of visibility in the dark. According to MassDOT’s database, there have been 71 serious traffic accidents in Amherst since 2010. Of these, 25 were at night, and another eight occurred in poor visibility conditions (rain, snow, or sleet). Without delving into causality in each one, in 13 years, 33 people have had serious accidents in Amherst in poor visibility conditions. 

And for everyone who has been hurt or killed, there are 10 or 20 or 50 who never ventured out because simply by looking at the streetscape or starting to walk on a dark sidewalk or shoulder, they realized it was not safe, and so decided to take their car or call an Uber if they could, or to stay in for the night. Current dark and intermittent street lighting conditions, in other words, are a big obstacle to a transportation mode shift in Amherst.

Street Lighting Under The Proposed Policy

Figure 2: The same arterial street under the proposed streetlights policy. The policy solves night sky and light pollution problems at the expense of safety and a transportation mode shift. Figure: Eve Vogel

I turn now to the current policy proposal. Figure 2 shows an approximation of the same stretch of road if the proposed streetlight policy is implemented. To develop this diagram, I worked line by line through the performance standards in the policy proposal’s appendix (you can reference my full narrative translation of the policy into layperson-friendly language here). 

The proposed policy (not the appendix) leaves intact the existing bulleted list of locations where streetlights are to be located: “▪ at intersections; ▪ at ‘dead ends’ and ends of cul-de-sacs; ▪ for road conditions deemed potentially hazardous, such as severe curves or topographical changes or areas with a high accident history; ▪ on all downtown streets (generally, the Central Business District); ▪ on other streets with heavy pedestrian traffic, such as in the vicinity of schools and in other commercial areas.” The performance standards in the proposal’s appendix state that a lighting fixture height must be less than 25 feet unless it’s on a utility pole. In this diagram, the light stays in the same location (likely an intersection with a side street) and on the same utility pole at the same height, at 30 feet.

The light itself has changed though. It is now in a warmer shade of the color spectrum—the proposed policy requires 2700 Kelvin, which is slightly yellowish, or lower (warmer color). The light is fully shielded meaning it projects zero light upward into the sky. The lamp or “luminaire” itself has  a low rating for backlight, uplight, and glare (BUG). The glare portion of the BUG ratings mean that illumination from the light is limited in the thirty degrees below the horizontal line of the light (as shown in Figure 2, the horizontal light that extends from the light is 90°; the “very high” glare at 80°–90° has stricter standards than the “high” glare at 60°–80°). It is possible that glare may also have been prevented under this policy after inspection from the Department of Public Works, perhaps in response to complaints from neighboring property owners about light coming in their windows. (However, glare from vehicles’  headlights has not diminished.)

In addition to restrictions on the lamp or “luminaire” itself, the proposal includes strict limits on where it can shine. The proposed policy prohibits almost any light from spilling over into private property out of the public right-of-way (the policy calls this “trespass” and deems it a “nuisance”—with suggested legal implications for the right of property owners to abate the nuisance themselves or sue the town for damages). The figure shows the effect on the roadway and shoulder: to keep light from reaching the adjacent properties, the light is directed only to the middle of the road, leaving the shoulders and/or sidewalks dark.

Finally, there are no new lights in this streetscape. The performance standards of the proposed policy are explicit: “No streetlight will change the overall brightness or color of the night sky in the Town of Amherst.” Outside downtown and the village centers, the night sky will be required to be clear and visible enough that people can see the structure of the Milky Way. How exactly this “skyglow” standard is supposed to be met at all locations outside downtown and village centers without removing every streetlight is unclear, but it is reasonable to infer that the bulleted locations mentioned in the policy are grandfathered in to be allowed streetlights. However, any new light outside those locations would surely violate the skyglow standard. 

The policy does list a few circumstances in which lights outside the bulleted locations, even in violation of the performance standards, can be installed. But any such decision is completely controlled by the Town Council—which likely means a cumbersome, political, and very occasional process. The policy states the council’s deciding power three times under different circumstances. Requests for new lights for pedestrians in residential areas will only be approved if they fit within the bulleted location list above—or if the Town Council deems the location to need a new streetlight. No new locations will have lights—unless the Town Council explicitly approves them. And the restrictions are absolute unless the Town Manager, with support from the Department of Public Works or Public Safety, requests a policy waiver for a particular installation—and the Town Council approves the waiver.

In this streetscape figure, no additional location or waiver has been granted by the council and I have included a streetlight only in the location from Figure 1.

There are two other changes that come with the policy that I could not show in Figure 2, because they have to do with time. First, the street light in this scene is off before dusk and after dawn. Second, it is dimmed to 70% or less after 11 p.m. 

Impacts Of Proposed Policy On Road Safety
What are the effects of this new policy on the people using this changed streetscape? The benefits are significant: people can see the night sky better, experience less glare in their second-floor windows, don’t have streetlights shining on their property, and may have improved sleep and wake cycles and other health benefits.

But the effects on pedestrian, bicycle, and transit safety are starkly negative. A larger portion of the street has become so dark that travelers without bright headlamps cannot see what is in front of them, nor be seen by passing vehicles. The pedestrian at 130 feet from the light has become close to invisible—especially since she is walking in a now-completely unlit shoulder. She, and the bicyclist and pedestrian further from the light, are more likely to trip or hit a stone, and crash into something or fall, which is particularly dangerous for the bicyclist and the mobility-impaired walker. 

In other words, the kinds of areas with dangerous lack of visibility that led to three pedestrian and two driver deaths in Amherst since 2013, and up to 33 serious injuries since 2010, now encompass far more of the streetscape.

Additionally, the light that is permitted is often not bright enough to provide adequate visibility for safety. The yellow color of light makes it harder for drivers to see pedestrians; this will be worse if the color is orange or red. The strictures against street lighting until after dusk or before dawn make the darkening half-hour or so before dusk one of most dangerous times of the day, especially on east-facing and heavily shaded sides of the street. The same dusk and dawn restrictions also prevent the town from providing street lighting in low-visibility weather during the day even though fog, rain, snow, and smoke can all dramatically reduce visibility. 

The requirement for light dimming to 70% or less after 11 p.m. or an hour after nearby downtown or village center bars or venues close, whichever is later, means that when residents walk home late at night from work at the mall, or when students walk up North Pleasant Street to Pufton Village after a party at 2 AM, or residents of Village Park get off the bus, they are now less likely to be seen, and much more vulnerable to getting hit by a car or truck. 

In short, safety is worse.

Proposed Policy Inhibits Shift Toward Sustainable And Equitable Transportation
What about the effect on a transportation mode shift toward greater use of sustainable, equitable transportation? As I explained in my first Mode Shift column, a mode shift requires not only strong, confident, and well-equipped pedestrians, cyclists, and bus riders  to walk, ride, and take transit, and also not only those who have no choice because they are carless and cannot afford a rideshare or taxi, or do not have needed bus routes or not at the needed hours. A transportation mode shift requires people who could drive or go in a car but are inexperienced or unused to walking and riding to choose to walk, bike, take the bus, scoot, or use another “micro mobility” option. But this silent majority of travelers are mostly afraid of dark road shoulders and sidewalks and fast-moving traffic (and rightly so). Under the lighting conditions created by the proposed streetlights policy, these potential shifters toward sustainable and equitable transportation will not venture out unless they are in a car. 

In sum: The proposed streetlight policy will successfully improve the visibility of the night sky and reduce light pollution. But it will do so at the cost of more pedestrian, bicyclist, and transit rider accidents and deaths, and a reduced possibility of an Amherst mode shift toward those more active, sustainable, and equitable modes of transportation.

Tune in next week to learn what kind of lighting pedestrians, bicyclists, and transit riders need, and how we can build a policy that helps both the night sky and a transportation mode shift.