Welcome, everyone. This is Bruce Kinzey with the Pacific Northwest National Laboratory. Thank you for joining today's webinar brought to you by the US Department of Energy's Solid State Lighting Program. This is a repeat of a presentation I made at the Illuminating Engineering Society's Street and Area Lighting Conference a couple of weeks ago. Following that event, we had a number of folks suggest that this might hold a wider appeal to a larger audience. So we decided to do it as a webinar.

And wow. Here we are. Our 1,000 available connection points filled up within about five days. And even right up until now, we're turning people away. We've turned away literally more than 100 people. So I think this really just goes to show how much attention these issues have stirred up among the lighting community-- things related to blue wavelength, and the potential health consequences, and so on.

I want to congratulate all of you on getting through the window before it slammed shut behind you. As we mentioned, this is being recorded and should be available on the DoE solid state lighting program website. And you'll be able to find links to that from that address shown in the lower left hand corner of this slide. That should be up within the next couple of weeks in case you have other colleagues or people you think might benefit from watching this.

I think this is also a good time to note that the Department of Energy is holding its next technology development workshop next month, November 16th and 17th in Denver. And we will be continuing to talk about these topics and related topics. And there will be lots of input from your colleagues around the country. So I really recommend that you consider joining us there if these topics are of continuing interest to you. And again, you'll be able to find links to the registration and so on from our home page there.

OK. One more thing before I get started. I just want to preface my remarks today with a couple points here. We're not picking on anyone here, either any individual or any organization. We're basically pursuing this in the interest of providing better information. The mantra I'm going to be using today is better information leads to better decisions. And along those lines, we can expect that as the level of scientific knowledge progresses on this-- research is conducted and so on-- that our understanding of these issues is probably also going to change in parallel with those. And so we will try our best to keep up with this and try to report information that is as current as we can get it at the time that it is presented.

OK. So I first want to provide just a brief basic background to bring everybody up to the same page here. There's a tremendous amount of energy that arrives to the Earth's surface via sunlight every day. It comes in a wide range of wavelengths. And among those, a much smaller-- relatively speaking-- a much smaller band of wavelength that we call blue in the visible spectrum. The physical relationship between those wavelengths and the majority of the components of our atmosphere, which is primarily nitrogen and oxygen, separates them from the pack and gives our sky a characteristic blue color.

Over the last billion or more years, all manner of natural systems have evolved under this blue canopy, and use its various characteristics like its intensity, and length of day, etc., to trigger all sorts of natural responses. And we're continuing to learn just how much influence these blue wavelengths have over our lives. We're also continuing to learn just how sensitive our receptors are to these blue wavelengths, and to the extent that light, even at very low illuminance levels such as that provided by electric lighting, can provoke some of these same responses.

And the reason we're all here today is, of course, that street lighting has received a fair amount of negative press lately for its ability to stimulate some of these same responses, which of course, coming at unnatural timing relative to our circadian rhythms may bring accompanying undesirable health consequences.

The concerns being raised are being leveled at LED products in particular due to their increased blue content relative to the sodium paper products that they are largely replacing. The two major issues being raised-- sky glow and ensuing health concerns-- are both related. There's a high correlation between those perceived issues and blue content.

Now I'm going to try to remember to use the word potential a lot today, because there are effects that have been scientifically established in the laboratory and so on. But there are still plenty of areas of scientific uncertainty, and data gaps, and so on. And a perfect example is how well do the results of research that's been conducted in a laboratory setting under very carefully controlled conditions and maybe involving non-humans-- other kinds of animals and things-- how do these translate to a practical situation of someone in their bedroom? Light filtering in from a street light outside, maybe through open curtains or maybe not, maybe closed curtains, the person on the bed with their eyes closed. How much exposure are they likely to receive? And then what, in terms of ensuing impacts to their health?

There's a huge link in between those that we're not even close to resolving yet. And that's the reason why there's so much ongoing research. They're basically trying to establish trying to resolve some of these uncertainties, trying to fill in these data gaps, and so forth.

In the meantime, there is no lack of projections being made that typically start with one of these established facts. And then maybe leap over sort of all of these other unknowns, using some heroic assumptions and often erroneous assumptions. And arrive at what is essentially sort of a worst case scenario, that then is some conclusion that's used to justify some sort of recommended action.

An example of this that's very common right now-- and this may be explicitly stated or it may just be implicit in the values being used for comparison-- is all other things remaining equal, substituting a very high temperature LED spectral power distribution or a low high-pressure sodium spectral power distribution will result in some impact.

So I'll call your attention to this photograph at the bottom of the page. This is a photo of the US border down around Yuma, Arizona where we were involved with a field demonstration-- just a small pilot involving the retrofit of three poles that had quartz metal halide. These aren't high-pressure sodium, but the same point's going to apply. These had quartz metal halide on the incumbent technology used. These were replaced with LEDs.

Well first let me just ask this question-- which of these three poles is not like the others? If you selected these three, you can give yourself a point. The LEDs are putting out something less than half the lumen output of the HID products. They had to lower those poles in order to maintain the uniformity, to meet the uniformity requirements on the illuminated space. But they were able to do so while still maintaining a horizontal orientation. That wasn't the case with the quartz metal halide, which had to be tilted at an average angle of about 45 degrees. And just look at the appearance here. This is from about a mile away, this photo. And look at the difference in glare coming from the HID products and, presumed to uplight everything else.

I didn't cherry pick this photo. I mean, this is a typical kind of result. I could have put any one of a dozen photos in here. And the point is that things are never the same, following the conversion to LED products. And the differences are not subtle, just like you see in this photograph. So relying on these kind of erroneous assumptions like I've got listed above there leads to erroneous results. And ultimately, conclusions that may not be at all optimized in terms of their practical application towards addressing the issues being raised.

So the media reports, we see discussing these issues. And there's been a lot of them lately as everybody knows. They often acknowledge the kind of capabilities that LEDs bring to the table that enable the kind of differences that I was just pointing out there. These are usually contained in the form of recommendations, such as what you see listed here. Avoid overlighting and incorporate dimming technologies when it's possible. Just use exactly the amount of light and direct it exactly where you want it to go. So eliminate light trespass and uplight. Select an appropriate correlated color temperature.

This is like a list of distinguishing characteristics for LED products. I mean, we agree wholeheartedly with all of these. But then when it comes time to sort of settle down to a final recommendation, there seems to be this overarching emphasis on really just one of these characteristics. Don't install anything over 3000k color temperature or even less. Or maybe even get rid of the blue wavelength content altogether. I imagine everybody on this webinar right now has heard these kind of recommendations.

We will contend-- and continue to contend-- that one size does not necessarily fit all. And what works, what was the best choice in one location, may not be the best choice somewhere else. And really a user needs to understand the collective contribution of all the key factors that contribute to sky glow and to these other ensuing impacts. In order to make the best decision, the city, say, is trying to achieve a balance among many objectives, many of which are competing. They're trying to be fiscally responsible, trying to minimize energy use, carbon emissions, trying to provide safety and security for their citizens, and so on.

They can't necessarily afford to chase one characteristic of LEDs or chase one of these objectives out to its extreme end at the possible expense of the others, because what they need to achieve is this balance. So better information, again, leads to better decisions. And with that in mind, the Department of Energy is supporting a variety of investigations to try to fill in some of these data gaps. And this is in the areas that we know where we have expertise. So that pertains to the lighting technology itself, and ultimately trying to better quantify the influences of these different factors that affect sky glow.

And then we will leave the ensuing what happens beyond that-- if there are health consequences and other things like that-- we're leaving those to the other experts in other fields that know these areas much better than we do.

So what I'm going to talk about in the remainder of this presentation now is I'm going to describe some of these investigations that we're pursuing.

The first of these is an investigation of sky glow. And this has turned out to be much more complicated than we anticipated when we first set out. Our expectations were set based on the number of published papers and so on appearing out there. We assumed that these were using very established techniques and so on. And it's turned out to be, a lot of these tools out there are actually in more of the developmental type of stage than we originally anticipated.

Our original plan was that we would select an established sky glow model that's being used out there in the astronomical modeling community. And there are a number of these. But perhaps they're relying on some of those erroneous kinds of assumptions now, like I mentioned earlier. We felt we might be able to work with the developers of these, improve the model by infusing some more realistic characteristics of the light source and use. And ideally, we would maybe identify a number of sites that were getting ready to undergo a conversion, get the specific details of the light technologies before and after, put those into the model, create some projections. Then at the same time, go out to those sites and take some measurements of the night sky, and then compare those two data sets. So we're comparing our projected values with actual measured values, use those to calibrate the model.

And then once we've done this at maybe five or six sites, we might wind up with a pretty good predictive tool that we could turn over to the lighting community and say, here's an empirically calibrated model that you can use to try to hone your designs to get the best result for your given set of circumstances.

Well that's when we were pretty young and naive in this whole business. The reality is, here we are more than a year later and we have just recently launched our first set of models runs. We had some false starts. We spent many months working with more than one modeler and so on. We also eventually abandoned the idea. We learned that this process of taking night sky measurements, sky brightness measurements, is much more complicated than we anticipated. Particularly to get the levels of precision that we would require for what we're trying to do here, it generally requires very expensive equipment. And it's not even just one measurement. Usually you're going to ideally take multiple measurements and calibrate those against one another and do some data post-processing.

Really what you want to hire an expert to come do this for you that's been doing it for years, has already come up the learning curve, knows where all the potential pitfalls are, and et cetera. And we just realized that we were not going to be able to achieve this in the time frame that we had available to us. We hold ourselves to a very high standard and we do not want to report information that we do not have a high degree of confidence in, or frankly that we can't even really explain. We don't want to have results that we don't understand.

So we ultimately walked away from this part of the project as well.

The good news is that we were able to identify a model that's been very well published in the literature, seems to be well respected among the astronomical modeling community. We were able to connect with the modeler, Miroslav Kocifaj in Slovania. We've been working with him for a few months now. His model was already calibrated using a city in Eastern Europe. And they were able to do some useful things such as turn the street lighting system on and off and take measurements under both conditions so that they could get a very precise contribution that was specific to the street lighting system of sky glow.

And then this graphic in the lower right is actually a photograph of the night sky. So this was taken from the ground looking straight up. It's a 360 degree photograph. And each of those points depicted on there is a point of comparison between projected values of the model and measured values. And looking across all of these, they found an average error rate of less than 2%. So that's pretty phenomenal actually.

That's all I'm going to say about that. If you want to read more about the validation process and so on, you can Google the information there at the end of that bullet under the source. And it will be the first report that comes up.

So I'm not an expert on the model. Fortunately, we have my colleague, Tess Perrin, that will be available to us during the question and answer period to answer questions about the model if anyone dares to go there. Some of the features that are valuable to us, of most attractiveness to us, is the ability to adjust these different variables to test the sensitivity of sky glow to them.

Now any model, particularly any model that's modeling the atmosphere, has to incorporate huge simplifications just to bring the calculational effort down to something manageable, because you're otherwise working with a virtually infinite medium. You could just look at the accuracy of weather forecasting, for example, that they've been doing for decades, and how often they don't get it exactly right just to get an appreciation for how complicated this is.

So down in the lower right hand portion of this slide, you'll see a depiction of the geometric setup of this model. If you want to read more about that, there is a complete citation of the first paper that describes the model at the bottom. In addition, the tabulated output that also produces these polar plots like you see in the lower left hand side of this slide, you can also read more about those if you like.

The only other thing I want to say about the model is that one of the features of particular interest to us is the ability to adjust characteristics of the luminaires. So this includes things like luminaire output, percent uplight, spectral power distribution, among others. And of course, these are precisely the features that we're interested in investigating.

So we came up with a series of scenarios for different input parameters. And we're looking at this at about the 10,000-foot level here right now. One scenario is a model run that has one setting for all these parameters. Then we increment one of those parameters, one value, that's another scenario. We run the model again and we increment again, and so on through this entire set. Until ultimately, what we wind up with is a scenario for every single combination of every one of those different parameters.

And then in addition to the 11 spectral power distributions, you see there in the middle that were selected to represent 11 different types of products, we also ran the model with every other combination in five nanometer increments of wavelength in order to test the sensitivity of sky glow to wavelength at that granular level of five nanometers. So when you add all this up, it comes to something more than 430,000 runs of this model.

It turns out we didn't actually have to run that many, because there was some duplication in there that was unbeknownst to us at the time that we set out to do this. But at the time we went to run the model, this is what we were facing, as well as several weeks running on a standard desktop computer.

One of the benefits to working for a big bureaucracy like we do is that we can call our friends down the hall who spend most of their days on the playground playing with our two-year-old supercomputer, Constance. She was in the top 300 fastest computers in the world when she came online in 2014. She's quickly dropping, of course. Nothing stands still in that area. I think she's still within the top 500, but just barely.

But nevertheless, she was able to crank through our scenarios in about 17 hours. And so now we've got this mountain of data that we are still sifting through. It turns out this is another one of these complicated processes. We're looking for universal truths, and major trends, and things like that. A universal truth is an impact that you can achieve across all scenarios. The only example of that that we've uncovered, and think we're likely to uncover, is luminaire output. And that has a scalar impact. So if you say, reduce luminaire output by 50% through, say, downsizing or dimming a luminaire, you will achieve a corresponding 50% reduction to the contribution to sky glow from that luminaire.

Another major trend-- this one is not scalar, because it's impacted greatly sometimes-- by some of those other parameters, especially things like atmospheric aerosol content-- that's referring to how much moisture is in the air, or dust, or other particles, and so on. But what's a major trend is that direct uplight has a significant to sometimes tremendous contribution to sky glow. Or conversely, reducing or eliminating direct uplight has a significant to sometimes tremendous impact on reducing sky glow.

So I've depicted this on this slide as replacing a drop lens fixture with a full cut off fixture. But it's really true for getting rid of any direct uplight. But in this case, replacing a 5% uplight fixture with a 0% uplight fixture achieves reductions in sky glow ranging from about 13% to about 52% with a mean value of 28%. And we say this is a subset of runs, because we held several other things constant, such as luminaire output and spectral power distribution. We didn't want to mix and match these things. We wanted to test, in this case, we're just looking for what is the impact specific to eliminating direct uplight.

So this may not come as really that big of a surprise if we take a look at a real world example. This is down in Los Angeles, courtesy of the Bureau of Street Lighting. We're looking at this street from above. So everything here is uplights on these fixtures. You'll also notice that there's a fair amount of back light that's illuminating the branches and things behind these. If we move to the post conversion photo, of course, that largely disappears to the point that you might actually almost have difficulty locating these luminaires in this photograph if they weren't illuminating the poles beneath them.

This still has uplight. But it's primarily that that is being reflected from the pavement now. And let's not forget that the previous lighting system also had that different spectral power distribution, but much higher intensity. These lights-- I don't know exactly what the reduction was in Los Angeles-- but I'm sure it's at least 50% fewer lumens output. And you can see the impact of these if you look at, especially like those hotspots up on the left hand side there, which all but disappeared here under the LEDs.

So the next question that we're very interested in-- and unfortunately, I don't have the complete response for you yet-- but under what circumstances and how often, if ever, does the increased blue content of this reflected light more than offset the impact achieved by the reductions achieved by eliminating those drop lenses in the previous photograph? And like I said, I don't have the answer to that right now. But you can bet that we will have that in a final report, in addition to things like additional contribution from reduced luminaire output and a number of other factors. So just to head off this question that I expect is coming in Q&A, I don't have an expected date for that final report yet. But we're working on it and we're going to try to get that out as soon as possible.

OK. I'm going to shift gears here now a little bit. This is a table that we generated, the Municipal Solid State Street Lighting Consortium generated in direct response to the AMA release that came out last June. This has been modified just slightly. We've added a couple other rows, but otherwise it's still the same.

The purpose of this was really, in addition to just providing numbers that people would be able to use in their own work and be able to cite, it has really just established the fact that these issues being raised are nothing new to our lighting world, and they are by no means unique to LEDs. The right hand column there-- relative melanopic content-- this is an indicator of the ability of a lighting source to stimulate some of those non-visual receptors that I mentioned earlier. These are all listed relative. They're normalized to high-pressure sodium.

So if you look at this, maybe later if you look at this a little bit more, one thing you'll discover is that light sources at generally similar color temperatures have generally similar percent blue contents and also fairly similar melanopic content. There's a lot more variation in the melanopic content, because it's not exclusive to blue wavelength. But I think people looking at this will be very, very surprised, for example, to see down at row T, incandescent on this table, and how much melanopic content it has. All of these white light sources contain varying degrees of melanopic content and of blue content.

Look a little bit higher on the table there-- mercury vapor. This is a product that's been used in street lighting since the 1940s. In fact, we still have a lot of mercury vapor out there today. A lot of the conversions taking place around the country are actually replacing mercury vapor with LED. So the situation is different, I think, than is really being depicted. And one thing that is very different, an advantage that LEDs bring to this situation, is all of these capabilities that I mentioned earlier that will enable them to address the issues being raised. They have more capabilities along these lines than any of the other products on this table.

And here's my go to example for that. This was also in our MSSLC response. Cambridge, Massachusetts underwent a conversion in the 2013-2014 time frame. They also did a pole-by-pole inventory at the time. So our baseline numbers here come from that inventory. They replaced about 54 million lumens of high-pressure sodium lighting across their entire system with about 32 million lumens of 4000k LED lighting. That also includes some excess to compensate for future lumen depreciation of those fixtures. So they also put a control system on there that they are currently operating at 70%.

So at dusk when the lights come on, they're only actually operating at 70% power in order to deliver just exactly the amount of light they need to meet their illumination requirements on the street. And the idea is that they will, over the years as these fade, they will ramp that power up to continue to maintain that necessary level.

But what that means is that they're actually only using about just under 22 and a half million lumens now in their system. And then furthermore, at midnight-- from midnight to morning-- they dim that system a further 50%. So now they take that output down to around 11 million lumen.

So there are a couple of different ways to calculate blue content, melanopic, and whatever. I just took some values off of the table that you just saw just to illustrate the point here. At dusk, when their system comes on, it's introducing about 1/3 more blue content to their local atmosphere than the high-pressure sodium system did. But after midnight-- from midnight till morning-- when they've done that additional 50% dimming, they are reducing their blue content in their atmosphere also by about 1/3.

So let me just put that succinctly. Cambridge's 4000k LED product system, during the hours from midnight to morning, produces less blue content in their atmosphere than the high-pressure sodium system that it replaced. I think that's a message that isn't getting conveyed nearly enough out there. And this is what I mean when I say LEDs that bring all these capabilities to the table. This is a nonstarter with any of the other street lighting technologies on that table. HID technologies, there's no way. This is a message I think we need to spread around a little more.

Also as a result of that AMA release, we started getting a number of inquiries in from municipalities, and utilities, and other lighting owners. Again, they were trying to achieve a balance among all these different competing objectives. And suddenly, here's a new concern that they have to worry about, the potential for health consequences. So they were coming to us, a lot of cases just conveying questions from their citizens and other people wondering, where does this fit in this set of priorities that we've got? Is this stuff established? Is this serious? We know there are lots of other sources of light we're hearing about from our interior lighting sources, for example.

About this time, Apple Computer came out with its night shift mode that eliminates the blue content after hours. And we're getting warnings about not using our electronic devices within two hours of going to bed. So how does this light source, this small amount of light we expect, infiltrating from street light outside, where does this fit? Can you put this into context for us?

So a colleague of mine, Naomi Miller, and I looked around to see what kind of data on the relative contributions of these different sources existed out there. The Lighting Research Center at Rensselaer has published a few papers on this. And really that was about all that we could find.

So we thought, well why don't we just take the lighting meters home to our own homes and take some measurements so at least we've got a point of departure here for discussion. And we did so, and we were impressed enough with the results that then we thought, well maybe we ought to recruit a few others to do this. So this table reflects a set of five readings. The people that had participated are listed up there at the top.

And we have now, likewise looking at this, we think we actually want to expand this to probably take this table to the International Association of Lighting Designers and see if we might bump up the number from five readings up to maybe into the low hundreds-- 100 to 200, or more than that. One of the downsides of working for a big bureaucracy is we have these scary offices called Human Subjects Research. And I called again. Miller had to run the gauntlet with this form, showing that we weren't going to have any negative impact. And thanks to her, she's almost through. So we're just about set to go out with this to the IALD.

But back to this particular table, which is obviously a set of preliminary results with just a few sample units. But I still think there are some initial results here that are worth noting. The first is the relatively small amount of light that is coming in from the street lights outside. So standing in their bedroom in front of the window with the curtains open, holding the light meter at their eye level and looking at the nearest street light and taking a reading. In all cases, this was 0.1 Lux or less.

Two of our participants here actually didn't have any lights visible from their bedroom. So their readings were actually 0. But three of us did. And in all cases, that was 0.1 Lux or less.

The second thing I think is very interesting here is to look at what generated the greatest range. Now the kitchen was pretty large too, but I think in the current context, this is actually much more interesting. This is from the bedside lamp. So this is laying in bed with a book at reading distance, maybe on your lap or wherever-- a book or a magazine-- holding that meter at eye level and looking at the reflected light from that page. We generated a range of results between 35 and 350 Lux. And of course, it's not like people are lighting these books with high-pressure sodium, or low-pressure sodium, or something. They're using incandescent, or fluorescent, or maybe LEDs.

If you go back to the table, these are all white light sources with plenty of melanopic content and so on. And look at the rows just immediately above that where we measured phone and tablet output. We're getting these warnings about not using our electronic devices. I think it's almost becoming an urban myth, something about the light coming from these electronic devices that is somehow different from light coming from other sources. And it's just not the case. And if you look here, we're talking about, in terms of intensity, maybe an order of magnitude greater or more.

So I think what this is all going to boil down to-- again, these results are still preliminary-- but I think what this will boil down to is that if some of these current scientific uncertainties are resolved and it turns out that these issues are something that we very much need to take seriously, we really need to do something about, it's going to require changes in our lifestyles across. And it's really not going to be sufficient to just focus all of our attention on this one relatively -- at least so far-- relatively small source.

OK. So I just wanted to show what the situation looks like outside of my own house to give people an idea. We live in a normal neighborhood and all that. When I showed these photos at SALC, at Street and Area Lighting Conference, I was challenged by a gentleman afterwards who suggested that the photo on the left was somewhat misleading, because of the sensors used in cameras that do not necessarily adequately record all of the blue wavelength and can, therefore, change the appearance of two different sources in a photograph like this. And he suggested that this was misleading.

So all I can say is that this does present an accurate representation of what I saw with my eye that morning, which is part of the reason why I took the photo. And furthermore, the values here actually correspond with the values that we've got in this table. So I'm going to go out on a limb here and say that the 4000k LED street lights in our town are both dimmer and warmer than the moon. OK, moving on.

So this is probably the last set of general issues that I want to talk about. And this is, how do the relative contributions from street lighting compare? Or how do they fit in with all the other sources of exterior lighting that we have in our environment? And of course, we have a lot of them. In this case, this is street lighting versus lighting coming from buildings. I'm not talking about being inside the building like the table we just saw, but from a point outside like where this photograph was taken.

This is Chicago. You can see in the upper left hand portion there the effect of the street lighting, the uplight and all coming from the street lighting. Whereas in the foreground, everything there mostly is building interior lighting, which is escaping from the window. So you also have some architectural lighting and some other sources there. We're above the city, again, so all of this is uncontrolled uplight coming in. That building interior lighting is probably 4100k fluorescent. Or I see some there that look even higher probably.

Chicago is getting ready to undertake-- they're in the planning stage right now-- of converting their street lighting system to LEDs. And we can expect that they'll achieve some of the same kind of results that we saw in that earlier Los Angeles photograph. So probably that upper left hand portion of this photo is going to change greatly over the next two or three years, whenever they get that system installed.

But what's going to happen in the foreground here? Nothing short of taking any other actions. And right now, as far as I know, nobody can really accurately predict what the change in sky glow is going to be over the city of Chicago after they undergo this conversion to LED products. But this seems central to this whole point. If light at night, and melanopic content, and all that contains a lot of hazards for us and we need to get serious about addressing them, we really need to understand what the various contributions are if we really want to come up with effective solutions.

OK. Similarly, I would like to call your attention to the street lighting in this photo by first noting that there actually is street lighting in this photo. It's currently being entirely obliterated by that monster screen up in the right hand corner. This is obviously Las Vegas, The Strip. This and Times Square in New York, these are extreme types of examples. But I think that probably most of us can think of examples in our own cities where you have lots of other sources of light-- maybe building architectural lighting, or signage, or landscape lighting that is being pointed up into the air and some of it's hitting the target and some of it isn't. And we saw earlier what the impact of direct uplight was.

So again, there are many sources that are really going to need to be understood and addressed if we're really serious, if it turns out we really need to do something about these issues being raised.

And finally, last but not least, everyone's personal favorite, this source of light at night that can go for hours into the evening, especially starting around this time of year when nightfall starts falling earlier. Here we are above the roadway again. So everything we see here is uncontrolled uplight. At least on the oncoming traffic, there's a fair amount of melanopic content in that. And if you think about going to the other side of this photograph and looking back, of course, there's a lot more going in the other direction.

Again, the contribution of this is significant around the nation, particularly around our populated areas. And in this particular photograph, I'll just point out that there isn't any street lighting even in this photo. So again, we need to understand the contribution from all these different sources if you want to be able to take the most effective actions.

Right now, we don't have any official activities for that last topic, trying to put these different things into context. So if anybody out there actually knows of studies, or documentation, or anything of other relative contributions of these different sources, we'd be very interested to see them. Again, we're just trying to assemble the best set of information that we can.

What will we do with this information once we get it? The first thing is, we'll be able to respond to municipalities and other lighting owners that, again, are trying to balance all these different objectives. And they want to know, what do they need to do? They're trying to be fiscally responsible, trying to minimize carbon emissions, trying to provide safety and security. But they don't want to do this at the expense of the health of their citizens, certainly. So we just want to try to provide some insight on this, try to give them the best information that's available so they're able to make the best decisions for their particular given set of circumstances.

We think we might also be able to provide some related tools to the lighting community such as a graphical user interface that we might put on our sky glow results that would enable a lighting designer to, say, test the impact of scenario A versus scenario B, and might be able to refine their design and put some economic numbers associated with that. And maybe come up with, ultimately, sort of the most effective and most cost effective solution for a given set of circumstances.

This one's sort of related to that. This is talking about maximizing the effectiveness of our efforts that we're undertaking. Again, you start to run into diminishing returns at some point if you try to pursue any one single tool to try to solve all of your issues. At some point, you'd be better off spending effort and maybe money trying to address somewhere else. And so hopefully by understanding the balance between all of these different contributions and so on, we will be able to increase our effectiveness in addressing these issues that have been raised.

Despite the longstanding nature of these issues, the fact that they've been around since electric lighting was invented, I think it's being lost because of the increasing knowledge from our medical community and from others about the potential impact that these have. And the fact that that coincides with the rise of LEDs as a major lighting product is forming an association in the public's mind right now that these are LED problems, or that they're even exclusive to LEDs.

And I will say that I don't think that this perception is largely being dissuaded right now in many of the media reports that I see out there. But like I said, LEDs bring to the table all kinds of characteristics, more than anything else that ever been out there before, to actually try to address the issues that are being raised. So we're going to continue to try to hammer home this message. LEDs are not the source of these problems, but they are part of the solution.

And then finally, there seems to be a fair amount of unnecessary contention out there, finger pointing at DOE and manufactures for developing what's essentially the most efficient and most flexible lighting source on the planet. As well as their local cities who, again, have been just trying to do the right thing. Suddenly, they have this sort of uprising, accusing them of not doing the right thing.

Hopefully by getting improved information out there and improving our effectiveness and so on, we can ultimately arrive at a more collaborative type of working environment here. And hopefully spend less time in this sort of unproductive opposition, which I think-- I'm sure-- many of the people out there listening right now are encountering in their own situations.

So here is a list of just some of the resources we have on the DOE website that you can look to if you want to investigate some of these issues more, or learn more about them, and so on. Then there's tons of other useful information on the DOE website. If you haven't been there, I really recommend you check it out.

And with that, that concludes my presentation. Now as I said, we're going to hold this Q&A session open for 30 minutes starting right now. I'm going to be joined by my colleagues, Tess Perrin, who as I said, will be handling any questions that are coming in related to the model or the sky glow investigation. And my colleague, Michael Royer, who will be responding to questions pertaining to color temperature, or blue content, or other matters related there.

So let's see. We have got a number of questions coming in. Let's see here.

I'll step in and answer one question while Bruce gets a chance to look at some of these things. There was a question related to, what are you considering blue light? I think that's an important consideration. It gets thrown around a lot. LEDs have more blue light. What exactly is blue? We have a range in the electromagnetic spectrum. And it's sort of a continuous spectrum. So there's not sharp cutoffs where we have one thing that's defined as blue and the next wavelength is not defined as blue.

So something that's important to keep in mind is that the photosensors in our eyes are all broadband photo sensors. And they have sort of peaks. They look like, for lack of a better description, a bell curve or something like that. And you can't just have-- how do I want to put this-- one wavelength and take it out, or make little changes in wavelengths and have the sensors see things completely differently.

So we need to look at things sort of more holistically. And we have functions to find to address these things. So just saying, look at LEDs. And you say, you see a blue spike there. Well if you compare it to another source that's roughly the same color of light, well it will frequently have a dip or less energy in surrounding wavelength regions. And so the way our eyes are integrating that information, that spike may or may not cause much of a problem.

In general, you're integrating over large enough areas that it's not as problematic as you might assume if you were just looking at that spectral power distribution.

So I'll go ahead and throw it over to Tess with a question. This is getting into more of the modeling. And we'll continue to monitor the questions as they're coming in. There is a question, did you factor in the surface of the streets in the model?

Thanks, Michael. Although different street surfaces could include asphalt to concrete or grass, we actually test the reflections at 15%. So the model actually follows [INAUDIBLE] equation. So there is a contribution of the streetlight that is from ground reflection. So as I said, that would be at 15%. And then there's another portion of direct uplight that you can assess the percent of uplight. So in our runs, we either set that as full cutoff, 0%, 5%, or 10%. But otherwise, we did not consider changes in surface topology.

So there are, I see, a number of questions related to the surface of the street-- asphalt, concrete, brick, different reflectivities, and so on. This is something that we have looked at. There was an earlier study by the Portland Cement Association that was showing that the actual spectral content of reflected light varies quite a bit. But I don't know that that is universally accepted. So we haven't adjusted. We haven't tried. We're still looking at that. We haven't tried to adjust for different--

And part of the issue is that this quickly becomes extremely complicated, because if you start, what sort of assumptions are you going to make? OK what color is the pavement? How much is covering pavement versus grass, or other kinds of surfaces? It's sort of a slippery slope. Once you start down there, then it turns out it's a rabbit hole. At the bottom, you go down and try to come up. So this is like one of the simplifications that most models use. They just assume an surface albedo for all conditions.

Let me take one here. Somebody asked, in simple terms, the 3000k LED, is it superior over a 4000k LED for sky glow? So if you look at the table, in general, the answer is yes. There's generally less blue content in a 3000k light source than a 4000k light source. But you're not guaranteed that. There's variance between different products and even different LED products, and certainly different sources have different wavelengths in there. So the answer isn't universally yes.

But in general, you're going to find a little bit less blue content in a 3000k than you would over a 4000k. And I should say here that we don't have any preference for one versus the other. Our preference is that the sites that are making the decision understand everything about their decisions so that they are making these in a fully educated way. And it may turn out that by the time you have taken into account things like elimination of direct uplight, reduction in luminaire output, and some of these other factors that that percentage change is going to turn out to be quite small.

And it might be, I mentioned earlier, that there are diminishing returns. Someone may look at that and say, you know what? We get a lot bigger bang for the buck if we invest our efforts in doing some other measure or something. We actually get a much higher impact for something. We don't know that. What we want to do is put all the information out there so it will enable people to make the best decision on their own terms. But to answer that succinctly, yes. In general, there is a little bit less blue content in 3000k versus 4000k light sources.

Let's see. That's essentially the same question Tess already answered.

Somebody asked a question. I'm not sure which slide this pertains to. This is not the same lumen output, HPS against LED white light. Were they concerned? Maybe that's talking about the Los Angeles.

The thing is it's the collective contribution of all of these factors is what, in the end, analysis is what's important.

We're trying to understand the influence of the individual characteristics of those lighting sources so that somebody might be able to design an optimal type of system for their given set of circumstances. But ultimately, what the sky glow results in is that overall combination. Obviously, whatever lights they end up putting in in the end, it's all of these things together, what creates the results.

Let's get a different set of questions here.

Yeah. I can go ahead, give Bruce a chance to read some of these. There's a question, does blue light exposure during non-sleep hours affect circadian rhythm or only exposure during sleeping? So there's a whole lot of literature out there on the effect of light on our non-visual systems, circadian rhythms. And essentially, all of the light we're exposed to throughout the day and the pattern of that light is affecting this.

So it's the light-dark cycle that essentially trains our circadian system. It's your history of exposure. It's the intensity of the exposure. It's the spectrum of exposure. It's all these things combined that are affecting your circadian rhythms. And I'm not a health researcher. But it's an incredibly complex system, our eye-brain system and our non-visual response. And so it's not just that, when it's time to sleep if you're exposed to light, that causes a problem. It's a much bigger picture than that.

Bruce? Got one?

Yeah. OK so this is the pitch. Somebody says, given that there are warmer LED fixtures available, and given that blue is known to be a problem for sky glow, melatonin, and in many settings aesthetically unpleasant, why is your agency so hot to defend LEDs? One can't help but wonder-- well hot LEDs. So they're talking about high-color temperature LEDs. One can't help but wonder if this is not industry driven.

It's not industry driven. And there's a couple responses I want to give to this. The first is an assumption. You mentioned some things that might be perceived. These are not necessarily all established yet. But they are perceived to be issues for higher color temperatures. But there's an underlying assumption here that blue is all bad, that there's no benefit from having blue wavelength in your source. That's not the case.

If you go out and look at, for example, IES TM-12 it acknowledges that there are some visual acuity benefits in street lighting applications specifically, meaning some improved visibility from additional blue wavelengths. So right now, these kind of assumptions sort of assume that blue wavelengths are all bad. There's no impact from reducing or eliminating blue wavelength. Whereas getting rid of them only brings benefits.

But I would say-- and there is not a study yet-- but here's an example of a data gap. Here's a study that might be interesting if they could ever establish it, would be there is probably-- I'll just go out on a limb-- there's probably at least an increased risk of, say, pedestrians being struck by vehicles that didn't see them in sufficient time as a result of eliminating blue wavelength from a lighting source than there is an increased risk of cancer from the additional blue content in those sources. There hasn't been any study to that effect, but I'm sure it's non-zero, the additional risk to pedestrians. So that's my first response to that.

But there are reasons to have-- plus blue light does provide better color contrast. We've done numerous surveys. People really like the ability to see that the grass is green, et cetera. So there are other benefits to having blue wavelengths.

And then the other thing that I wanted to say is that we don't actually defend any-- DOE does not push one color temperature over another color temperature. But we want to ensure that people understand the consequences of their actions, understand the full complement. And I guess one thing that, I guess, is the appropriate time to say is that what we especially do not want to do is have some understanding or some implied understanding that blue light actually brings the health consequences to somebody in their house.

That has been claimed since that is not yet scientifically established. We have a lot of people out there that have done a lot of hard work, and they've spent a lot of money and a lot of effort trying to design systems, trying to meet all these objectives. And in some cases, they have installed entire systems. Or in some cases, they're midway through. And these objections are coming up and stopping this work midstream. This costs a lot of money and also opens up cities, for example, for future potential litigation maybe and stuff. People say, well they left those lights in.

These things have not been established yet. And so we want to continue to defend those until these uncertainties and so on are completed, and are established. If they ever are established, we will change at that point. But right now, it's not justified. These things are not justified by the evidence that has been developed to date.

Michael, have you got another one there ready?

I don't have a specific question. I'm kind of watching these and seeing some trends emerge. There's a lot of people asking for specific recommendations. You know, I have this situation or that situation. Would you, in general, recommend this over that?

And I think one of the points that Bruce has tried to make is that it depends on the specific situation. And I think setting up these generalized recommendations can sometimes lead us down a path that doesn't give us an optimized solution for any given situation. So we can see trends emerge in terms of how much blue content versus CCT. Again, I'm sort of on the spectral side here. CCT is not a perfect correlate for melatonin or IPRGC sensitivity. It's not a perfect correlate for blue light hazard or retinal damage. It's not a perfect correlate for sky glow and the amount of blue.

CCT is based on three color matching functions that weights, essentially, the different portions of the spectrum. So it has a function that's based in the blue. And so it generally gives us an idea of how much blue content is in a source. Bruce showed that table. A lot of data went into that table. You can see that in any given CCT, there's a range for any of these more specific and related blue content sensitivity functions.

So there's a specific question here. That 4000 LED source is no worse than conventional sources. That's the news in terms of health impacts. I think the question there is, what exactly are you comparing it to? Is that compared to HPS? And I think the other point that Bruce has tried to hammer home is that just switching from one to one is not all other things being equal. And in the case of all other things being equal, if it was that perfect world, there'd be slightly more blue content in that LED light source.

Laboratory studies have shown that that blue light at the wrong time can have negative health consequences. But we don't have a lot of data in terms of in application, what kind of effect that might have on people.

There's a lot of other questions here related to my screens, when I'm driving, or if I go to a store at night, or even in my home. So we have all these exposures to light at night. And street lighting has kind of become this magical thing that's causing a bunch of problems. I wouldn't say that we should ignore it because of all these other factors, because it is something that individuals don't necessarily have control over. But I think, again, to reiterate some things Bruce has said, putting it into context of the effect that might have versus all these other things is important to consider.

Yeah. There's two other factors here that I think you alluded to but didn't mention explicitly, Michael. And that is, it's not just a function of the blue wavelength content, but also of the intensity of that light and that time of exposure. How long are you exposed to those light sources?

So I'm getting other questions. Like somebody said, wouldn't you think there's more exposure to blue light from the office and store lighting than home lighting? Well of course, it always depends on your individual situation. There is blue light coming maybe from an LED street light. But how much of that is actually infiltrating into a bedroom window? You know, it's probably very small. Now you've got a much higher intensity from your night stand light. But then you turn that off and you go to bed.

There's all of these different variables in there. So these questions are very difficult. It's very difficult to come up with sort of like a generalized overall statement. And this is part of the challenge right now is that we're focusing a lot of attention on street lighting. But there's also many other factors that go into how much exposure somebody gets. And even on an individual basis-- an individual biological basis-- things like, how much exposure a person has had to daylight during the day, and how, I guess, sensitive their responders are to this at night, and so on. So it's very difficult to come up with these general conclusions.

I think, Tess wants to jump in with an answer about something on the model.

Sure. One of the questions was whether or not our decision to use only one sky glow model and to not validate our runs with our own ground measurements undercut the reliability of our results. Well given all of the sky glow models take quite a bit of time to understand how they work and what variables we can input to see if they're going to give us the kind of incremented data that we're looking for basically to be able to isolate how one variable is performing under different circumstances.

We decided to go with Miroslav Kocifaj’s sky glow model simply because it has gone through three sets of validations that are deemed to be worthy in the atmospheric community. So his model was compared with a more complex model that had already been validated. Various sensitivity studies were also conducted in order to test that the model predicted well-known behavior. And Bruce presented other results where measurements had actually been taken and were validated with his model.

So this investigation, on our end, is our first investigation. So our initial results, yes, they are only based on one model. But they're based on a model that has been validated much more thoroughly than many other models.

I'd like to jump in and answer one here.

Someone sent a slide set from another presentation talking about moonlight and 4000k LED. And saying, their spectral power distributions are very different. And look at this strong blue spike near the melanopic peak. So I'm going to be giving a whole presentation on this topic at the DOE Technology Development Workshop in November. So I'd encourage you to come to that.

I'll give a quick explanation of this one right now, because I've specifically looked at this slide before and evaluated what's going on here. So when I traced that LED that's claimed to be 4100k, I actually found that it was 3400k and also extremely green. So it had a DUV of something around 0.017, which is, it would look almost like green light to anyone who was viewing it. So I question that moonlight spectral power distribution a little bit.

Also these two spectral power distributions are presented as relative spectral power distributions, not equalized for lumen output. So the maximum peak in each is one. You get a very distorted view of spectral power distributions when you present them side by side this way. And so, again, part of my presentation is going to be talking about, you can't just look at relative spectral power distributions and understand these numerical quantities that can actually be calculated.

Because if I do go and plot those SPDs and put them into the table as Bruce showed in one of his slides, there is actually lower melanopic content in that LED that appears to have that blue spike. And one of the reasons is that blue spike is actually around 450 or 60 nanometers, where the peak of the IPRGCs non-opsin containing photoreceptors is around 490. Well that peak in the blue actually means there's a trough around 490.

So again, this sort of look at that giant blue spike really frustrates me, because when you look at the actual numbers and you do math rather than just a visual analysis, your answers might actually be quite different.

So I'll pass it on. Bruce, do you want to answer something?

Yeah. Actually, I've got to ask you a couple questions, because you've had a lot of people in doing subjective review of different color quality and these kind of issues. So here's a couple questions.

One question is, how does the appearance of 3000k compare with 4000k? I guess this would be more than just a street lighting context. But what kind of difference would people be expected to see between those?

So one of the things we have the benefit of in our visual system is chromatic adaptation. So you experience a broad range of CCTs throughout your day. You probably have a home where you're down 2700 or 3000k. Maybe not, but many people are. You probably go to an office and you see 4000k. You probably go outside and you often are seeing 5500, 6000k. But we have color constancy. And our visual system adapts so that your red apple generally looks red in all those conditions. Your clothing doesn't drastically change colors throughout the day.

So at least within a range for street light, for example, I wouldn't expect drastic visual differences between 3000k and 4000k when you're immersed in that environment. When you see them side by side, you have mixed chromatic adaptation. And you can detect obvious differences between the two.

So personally, I don't think there's necessarily going to be a lot of loss in terms of color contrast detection going from 4000 to 3000. When you start to get into ranges 2700k and below-- and the exact threshold is not really well-defined-- but our chromatic adaptation ability has limitations. So I think you throw in HPS, or even something that low that wasn't poor in color rendering, you are going to start to lose some of that color information that you gain from having white light.

So I'll stop there. Bruce?

OK. Yeah, I'll take this one. So the question was, what's the feedback from the public or residents on the CRI improvement if sodium vapor is replaced with 3000k LED street lights? So I'll answer this in a little bit larger-- this actually opens a nice door of another point that I wanted to make.

It's being conveyed right now that it sounds like there's a big uprising-- the public objecting to these very high color temperature streetlights, 4000k LED. I was managing our field demonstration program several years ago when I first came onto this project. And we worked with a number of cities installing lights at higher color temperature because that's what was available. So generally, we had a number of installations that-- well a few-- that started out around 6000. And then just small scale installations and down to 5000.

And then ultimately, 4000. And we would always make an opportunity for the public to comment on these lights. And sometimes these were brochures that were distributed throughout a neighborhood. Sometimes it was a website that was set up that people could go to and provide responses.

The few times that we did get, even with the higher color temperatures-- even at like 6000k the vast majority of responses were very positive. Most of the residents liked the light. They liked the improvement compared to the high pressure sodium very much. When we did get complaints, it typically did have to do with the high color temperature. These lights are too cold. We don't like them. This would be like less than 10% of the responses though.

But then as the technology improved, lower color temperatures became more widely available. And we started installing lower color temperatures. Like basically, they centered around 4000k. Those complaints largely went away in the surveys of the people, the users, of those spaces that were lighted.

Now recently here, we're seeing complaints. They're being conveyed in these articles like, well look. We've got three different people here that are saying something terrible about the color temperature in the lights. But I have worked with most of the major cities at this point-- or our team certainly has almost covered the waterfront working with cities that are converting lights. The vast, vast majority of responses we get from people-- this is being conveyed from maybe the facilities manager, the person in charge of the lighting conversion program-- has been very, very favorable.

You can never please everybody. You would never have any trouble actually going out somewhere and finding somebody to make a negative comment. And of course, that's what the papers are reporting. They're only reporting vocal comments from people that have whatever kind of beef to pick or something like that. But that does not reflect.

And I've heard this. I've had conversations with two major cities that we see in the news all the time. At SALC, I asked this question specifically. And that does not reflect the majority view. We're talking about cities in the multiple millions of residents. And the articles that have come out and talked about those do not reflect the majority view of the comments that they get from their populations.

And so in general, people like the conversion to LED quite a bit. They much, much appreciate the improved CRI, the ability to tell colors, color contrast, it looks much more natural. Of course, the higher color temperatures, sometimes people do complain about those. Some people do not like. There was an article from a guy in Brooklyn that called high pressure sodium light was light to fall in love under. Well OK. That's a very unique perspective on this issue, and I respect his opinion. But I don't think that would be very widely held among the population.

And so I think, in general, the reception to LEDs has been very, very positive. If people are objecting to 4000k as being too high, then a city might want to go ahead and put in 3000k. We always recommend that a city do a pilot project first, do a mock up, maybe do a couple of streets. Invite all their citizens, their police force, everybody else to come out-- the mayor-- everybody else to come out and take a look and to gather feedback. And not to move across the entire city until they've gotten that feedback and they're assured that what they're going to install is acceptable.

If everybody seems to like 4000k and one person doesn't, maybe that's not reason enough to go with 4000k. If they decide they want to go with-- or to change from 4000k. If they want to install 3000k because it's more politically expedient, that's enough reason in itself. If they decide they want to go that route, we don't advise them either way. Again, we just want them to understand that, here's what the state of knowledge is right now. And so you're making a fully informed decision, as fully informed as we can provide the information. We don't push one or the other.

OK. That was a long answer to that short question. Sorry. You have another one there, Michael?

Yeah. I got an interesting one. There's a question of essentially, why not just go with higher CRI instead of higher CCT? That is certainly a consideration in terms of if you're trying to get accurate color reproduction with your lighting. Another interesting fact, though, is that with phosphor LEDs as a generalization-- and again, I hate to generalize LEDs in any way-- but in general, if you go to higher CRI with LEDs, you're going to increase the blue content whether you look at that in terms of melanopic content, or blue light hazard function, or probably sky glow. Because with LEDs, they're essentially optimized to give you good green rendering and lack red rendering. And to go to higher CRI means you're adding red.

Well if you're going to keep the same CCT-- say you're at 3000k and you want to go to higher CRI, if you add more red, you also have to add more blue. So you'll see some of the ranges that we listed in that table that Bruce presented for LEDs at any given nominal CCT. Some of that has to do because CCT is listed there as a nominal. CCT is also a very inadequate measure of spectrum, because we have other things like DUV. And then CRI is another factor in there that is causing that range, even for just phosphor-coated LEDs.

Now I think one of the other interesting things we can do with LEDs is that we can engineer them at any given CCT to have maximum or minimum effect on melatonin or any other quantity that you want to have. Because we can more carefully control the spectrum, we can sort of exploit the differences between the z-bar color matching function that's feeding information to calculate CCT and the sensitivity function of the IPRGCs. So we can move those peaks and troughs wherever we need to put them.

Now is it the case that we have those optimized sources today? Not necessarily. But the ability to engineer those sources is certainly available. And I would expect over the next few years we might start seeing some of those sources.

OK. I have another comment. This is more of a comment. The MSSLC standard-- that's referring to the Municipal Solid State Street Lighting Consortium-- standard specifications are 4100k plus or minus 200k. The luminaire specification that was issued by the MSSLC is intended to be a template. It's not a standard. Every value in there is intended to be-- and this is the way the whole thing was built-- was intended to be tailored to the needs of a particular site.

So we do have some default values in there. But we're not recommending that. The user can just as easily go in and type over the top 4100k, put in 3000k, or 2700k, or 6500k, whatever they want. We did not establish a standard. We do have just default values to illustrate the kind of thing that a user should input into those spaces. But it's at their discretion, whatever values they want to put in there.

Here's another one that somebody asked, so mercury vapor street lights would have lower pedestrian accidents than HPS? That's not what we're saying. And of course, any given situation, any given circumstance, has multiple factors contributing to what the likelihood, including just driver behavior and pedestrian behavior, and all kinds of other things. The statement that I made earlier is that blue content does provide some additional visual acuity benefits.

So all other things being equal-- and I can't even really go there-- but the idea is that there could be a quicker, say, somebody approaching the road from the side, they might be recognized just an instant earlier with better visual acuity and could be in enough time. You can't even predict this, but it could be enough of an additional warning that a driver was able to avoid a collision with a pedestrian. The point I was making is that there is undoubtedly some increased risk associated with removing blue wavelengths.

There's a lot of proponents out there right now of producing street lighting that has no blue wavelengths in it at all. And I'm contending that there's at least an equivalent increase in risk to pedestrians and other people probably as to whatever the potential risk is from having those blue wavelengths in there.

I guess I should pass it back to you, Michael.

OK. We get batches of about 15 questions and we kind of got to sort through them and see what we've answered or not answered. So I'm kind of struggling to see what is going on here.

OK. I got one here. I can take one here, Michael. Somebody asked, would you need to reduce the spacing between the light poles if you switched from 4000k to 3000k to maintain the same light uniformity on the roadways? So it's just the nature of the way LEDs produce light. It used to be that higher color temperature products tended to put out fewer lumens. There were more losses in the conversion of the light to the warmer color temperatures. And so you would have to put either a higher wattage-- you wouldn't typically change your pole spacing, because that's the most expensive component of the whole system-- but you might have to put higher wattage luminaires on in order to meet the same illuminance, the same lighting requirements.

That difference has been greatly reduced through various techniques the manufacturers are using now. So that essentially, the answer is no. I mean, from more than one manufacturer now, you can get the same product at either 4000k or 3000k in it. And it outputs lumens maybe within 5% or so. And they both are at the same wattage, like a 30-watt or a 50-watt product. And they output lumens within about 5%.

So the general answer there is, it depends on the product. Not every manufacturer has products like that, equivalent products at different color temperatures. But there's more than one available now. So that would not be necessary.

There's a question in here talking about DOE limiting its research on sky glow. How does that relate to the health issue? I'm going to take this in a broader context. Again, I'm not really the expert on street lighting. I'm more of a spectral engineering person. And the relation there is that all of these issues are related to the spectrum of light. And it goes beyond those. So our visibility is related to the spectrum. Spectrum has health impacts. Spectrum has sky glow impacts.

Spectrum has wildlife impacts. There was a question related to health impacts on animals and plants as well. And I think the question was phrased essentially, well why aren't we seeing problems in plants and animals? I don't think that's necessarily a fair assessment. I think it's hard for most of us to communicate with plants and animals and understand how they're being affected. But I do think there is some research out there that's showing that late at night certainly does affect plants and animals.

So it's all of these things combined when we need to make a decision about spectrum. And that's why all these questions asking about, can you recommend this over this, it really really, really depends on the situation. What is your primary concern? Is it wildlife impacts? Is it human health impacts? Is it pedestrians and roadway safety? Is it retinal damage concerns? Is it energy use? All of these things have to be weighed against one another. And multiple reasonable people can come up with different answers to that question.

OK. I think we're just about out of time. I will answer one more question here. And someone's asking, what kind of actions can we take to make this discussion more fact-based? They're talking about the discussion ongoing out there right now. Well this is exactly the kind of thing that we're trying to do here. That's what I mentioned earlier, is supporting a variety of investigations to try to fill in some of these data gaps and try to publish just better, more complete kind of information so people have a better understanding of the contributions of various factors. And we're not just empathizing or just focusing on one element of a perceived problem. When you put it into context, that undoubtedly will change the perspective on that problem.

And so that's basically why DOE is engaging in these different activities. And I guess that's it.

We hope that ultimately, we can get to a more, like I said, collaborative sort of working environment. We'd like to work with the different communities that are raising these issues and try to produce solutions that basically work for everyone. We're not likely to agree to a solution that just emphasizes one tool over all else. But rather, we want to try to put this into an appropriate context for the user community out there. And hopefully we can come up with solutions that, again, will never please everybody. But if we can address the concerns of the vast majority of people, we'll call that a success.

So with that, I think I'm going to go ahead-- we've used up our half hour. So I want to, again, thank everybody for participating in today's webinar brought to you by the US Department of Energy's Solid State Lighting Program. We do hope this was useful for your own street light planning and implementation efforts. Thanks again for attending and you all may disconnect now.