In 2021, hundreds of citizen scientists walked a distance of 600 km, counting up and classifying all of the outdoor lights they could see while they did it. The results of their observations have just been published in the journal Nature Cities: https://www.nature.com/articles/s44284-025-00239-5
Find out why they did it, and what they found out in this thread 👇
(1/)
#CitizenScience #LightPollution #OpenAccess #NighttimeLights #RemoteSensing #Nachtlichter
First off, some background. In 2017, I published a paper that showed that most countries in the world are getting brighter at night in nighttime satellite observations: https://doi.org/10.1126/sciadv.1701528
A number of journalists who reported on that paper asked me “what is actually changing on the ground?” Are streetlights getting more numerous or brighter? Or is it because bike paths are getting lit up now? (seriously got that question from multiple journalists…) Or is some other kind of light changing?
I realized that not only did I not know the answer, but that no one even knows how much the different light types contribute to the brightness of the ground (as seen from space), or the brightness of the night sky (as seen from the ground).
(2/)
That question from the journalists had a major impact on my research career, and I’ve spent several years trying to understand the amount that different types of light sources radiate into the night environment.
That led to an EXTREMELY COOL research paper: https://doi.org/10.1177/1477153520958463
The city of Tucson, Arizona alternatively dimmed and brightened nearly their entire street light network (tens of thousands of lights!), in order to see how much light comes from the streetlights, and how much comes from other lights.
(3/)
By changing the lights on different nights, it let us extrapolate to find out how bright the city would be if it turned off all the streetlights completely.
The result was surprising: if all the streetlights were turned off completely, Tucson would barely change in the satellite image. After midnight, only about 13% of the light from Tucson that's seen from space comes from streetlights.
(4/)
There was a difference between suburbs and the city center - streetlights made up a bigger portion of the light emissions from the outlying areas, but still only up to about a third of the total light.
(5/)
That's what motivated our #Nachtlichter project. In 2019, the @association issued a call to Helmholtz Centres for #CitizenScience projects. In addition to running a citizen science experiment, the projects had to involve at least two different centres.
Together with colleagues from @DLR and @ufz, we submitted a project idea called Nachtlicht-BüHNE: https://nachtlicht-buehne.de/
The name in English means: "Citizen Helmholtz network for research into nocturnal light phenomena".
The project had two parts: DLR (the German space agency) would create and app for reporting Fireballs: https://meteor.nachtlicht-buehne.de/
And our group would create an app for studying outdoor lights: https://lichter.nachtlicht-buehne.de/#l=14.7/49.57249/10.89506
(7/)
An extra special aspect of both apps is that they were co-designed with our teams of citizen scientists.
That means that from the very start, we had a team that helped us to design the entire app and project.
COVID hit in the midst of our starting up period, and since then we've been having meetings with the co-design team roughly every two weeks (94 in total).
(8/)
The #Nachtlichter app was designed to assist participants in counting and classifying light sources. We pre-defined street segments (usually from one street corner to the next), and participants went there at night and reported all the different lights they could see.
One question that we had to deal with early on was "how do we categorize all the different light sources that are out there?" In the end, we came up with 18 categories (including "other" as the final category for uncommon things like flagpole lights, glowing park benches, illuminated water fountains, etc.).
(9/)
We had originally proposed that we would cover at least 6 square kilometers in at least 3 German communities.
In the end, our fantastic team of citizen scientists made observations in 33 communities in 9 countries (but mainly in Germany), covering a total area of 22 square kilometers!
(10/)
So what did we find? Well, first, we compared the number of lights that our participants counted (per square kilometer) to the brightness of the #VIIRS_DNB satellite instrument.
We see a clear correlation between the two, and this gives us a "conversion" factor from "nW/cm^2sr" to "lights/km^2".
We used this factor to calculate how many light sources out team would count if we covered the entire area of Germany exactly at midnight. Our result is 78 ± 3 million lights, or roughly one per person.
(11/)
You can see that while the relationships are obvious, the number of each light of different light types is not directly proportional to the satellite brightness. That's similar to what we saw for Tucson: as you go from more rural to more urban places, the relative importance of street lights decreases.
We looked at this by comparing the fraction of lights of different types for observations in different land cover types.
(12/)
These results are particularly important for people who model artificial night sky brightness (skyglow), because up until now, most skyglow models assume that all of the light comes from streetlight-style lights.
This shows that there are a great number of light sources that shine sideways (signs, and windows), that are right now being largely ignored.
This is also important for efforts to reduce #LightPollution, because it means that while streetlights are important, fixing them is not going to come close to fixing the entire problem.
Illuminated advertisements and lights mounted on buildings are also important - and the lights from residential houses probably matter a lot more than people generally think!
(13/)
Huge thanks go to the hundreds of people who took part in the project!
Thanks also to @helmholtz and @bmbf_bund for funding our research!
If you'd like to learn more, we've also written a "questions and answers" document for non-academics. It's available in both German: https://doi.org/10.48440/gfz.1.4.2025.001
And English: https://doi.org/10.48440/gfz.1.4.2025.002
Check out also this "behind the paper" blog post written by two of our citizen scientists: https://communities.springernature.com/posts/behind-the-paper-citizen-science-illuminates-the-nature-of-city-lights-citizen-scientists-insights-into-the-research-project-nachtlichter
And thanks to everyone who read and shared parts of this thread!
(14/)
Nature Cities has also published a "research briefing", with comments about the paper from me, one of the reviewers (Noam Levin), and the editor. Check it out: https://doi.org/10.1038/s44284-025-00242-w
(An addition to the thread later in the day, 15/15)
Another article about #LightPollution also came out in Nature Cities today: https://www.nature.com/articles/s44284-025-00258-2
@travislongcore is a co-author, and the paper's title is "Artificial light at night outweighs temperature in lengthening urban growing seasons"
The authors looked at how temperature, artificial light, and how urban a city is (impervious surface fraction) affect the start and end of the growing season for plants.
They found that artificial light "increased exponentially toward urban centers, and exerted stronger influence than air temperature in lengthening the urban growing season, especially by delaying its end, although the effects varied across climate zones."
In case you are interested in light pollution and didn't see it yet, check out the thread above 👆
@skyglowberlin Ahhh I was literally just wondering on the bus this morning about whether increased lighting vs 'just' climate change might be behind some of the unseasonal flowers @joncounts and others have been noting....
@cosmos4u yeah, at first I thought it was good that they came out together, but that may have sucked the air out of the room for ours...
It was reported in El Pais, Deutschlandfunk Nova, and (maybe?) BR.
Coming out a week before the summer solstice probably also doesn't fit well to the stories the editors want to tell - the "hot time of year" for interviews with me has nearly always been January and February, when the nights are long.
@skyglowberlin @travislongcore
This is cool.
I also noticed that certain wildflowers bloomed near the glass door of my house at least 2 weeks earlier in the Spring than it did in the surrounding yard.
There might have been some affect due to temperature leakage, but it really seemed to be due to the light, and this supports that.
@skyglowberlin great thread, thank you.
I read a couple of times about vehicle lights being indirectly counted but couldn't see the contribution quantified?
🙏🏽
@jbenjamint Yeah, we asked people to rate how many cars were driving. The scale was something like:
o continuous traffic
o several per minute
o one car per minute
o little to no traffic
We didn't have a good way to bring that into the paper in the same way as the count results, but also, traffic pretty much ends on the vast majority of German streets around 9-10 pm. I mean, there are still cars, but it's nothing compared to the early evening.
Car headlights might help to explain why we see such a discrepancy between citizen scientist observations of star visibility in the early evening and satellite measures of light emissions well after midnight (see: https://www.science.org/doi/10.1126/science.abq7781)
@skyglowberlin @helmholtz @bmbf_bund
Thanks to cheap 360 cameras, it might be quite simple to get a bunch of volunteers to actually survey a target at midnight.
I'm thinking a group of motorcyclists could be prepositioned with chosen routes, and at the right time, ride out. Send the videos back, run them through some image processing, presto.
Similarly, riders willing to do this over moderate distances could get you rural data.
@Benhm3 We thought about ideas like that, but it's difficult to clearly identify light sources from imagery alone. In this example, a human can figure out (especially if they are there) that the light here is coming from floodlights and count them. But it would be much more challenging to try to get that just from imagery.
There are lots of other examples where lights can only really be seen from the sidewalk and not the street.
@skyglowberlin @helmholtz @bmbf_bund
Wow. Fascinating
@skyglowberlin Yep, our GMN network in NZ is augmented very nicely by receiving public reports, they're helpful.
Dunedin in NZ is a "dark skies" location, and the council have been replacing traditional sodium orange streetlighting with downward focused LED, amongst other things. It makes a lot of difference driving around the suburban areas - at night you can no longer easily see houses etc at the sides of the road, just the road itself. It would be interesting to know if this really helps!
A customized and shielded street lighting design significantly reduces the lethal attraction of flying insects in various environments. The approach harmonizes human needs with minimizing the ecological impact of artificial light at night.
@skyglowberlin please tell me this isn't the last post in this brilliant thread!
Edit: phew, glad to read that it's not!
@jbenjamint I don't know if
@skyglowberlin intends to write more in this thread or not but I spent the whole thread thinking "Is it advertising billboards?" and then went to the article at https://www.nature.com/articles/s44284-025-00239-5 and read in the abstract "in German city centers, more total light sources are used for advertising and aesthetic purposes than for street lighting".
Now to go back and read the whole thing to get a bit more nuance. 🙂
This study employs a citizen science approach to identify and classify over 230,000 light sources in German city centers, suburbs and villages. The results underscore the pivotal role of citizen science in expanding knowledge of artificial light emissions and bolstering policymaking efforts to mitigate urban light pollution.
@jbenjamint Ah, important point from deeper in the article - "private windows are the most common outdoor light source, [so] policy encouraging the use of curtains after sunset could have a larger impact than one might first assume".
@zeborah @jbenjamint That's just "counts", of total lights, though. Private windows are both smaller and less bright than advertisements, so in terms of absolute photon numbers, they may be less important.
Many more private windows also turn off over the course of the night than do advertisements. So in the early evening, private windows matter more than you might think. But in the late night, my gut feeling is that advertisements that shine while nearly everyone is sleeping are actually an important part of light emissions from urban areas.
Christopher Kyba 🇨🇦🇪🇺
Moritz Negwer
Zeborah
nosfe
Daniel Fischer
Rozzychan
Ben Thompson 🐕
Large Format Projectionist
glasspshr
CMDR Yojimbosan 🅅⁂
Dilman Dila