Wednesday, September 18, 2019

Bad lamp workarounds

Have you ever seen a badly designed or implemented outdoor lamp, and then actually done something about it? I've actually done this twice with floodlamps that were meant to illuminate buildings but were not well aimed. Just a little tilt, and suddenly all the light is on the building and no longer streaming into the sky... 

Here's a great example photographed by Tatjana Scheffler showing how someone decided to deal with a problematic buried spotlight:

This work by Tatjana Scheffler is licensed under a
Creative Commons Attribution 4.0 International License.

The problem was that when you go to to get your bike, the lamp shines up into your eyes and is very disturbing. Whoever found the pylon "solved" the problem for at least a portion of the bike parking area. Of course, this meant that the pillar was no longer floodlit, so this is a more drastic change than what I was talking about above.


Still, if light art is going to be used, it ought to be beautiful. And not just for the cars that are driving past, but for the customers that are shopping in the store employing the light art!

Thursday, September 5, 2019

First action of the "Nachtlicht-BüHNE" lamp project

I'm very excited to announce that the Helmholtz Association has decided to fund a new project related to light at night, called Nachtlicht-BüHNE. It is a co-designed citizen science project, funded under the CitizenScience@Helmholtz funding initiative.

BüHNE stands for "Bürger-Helmholtz-Netzwerk für die Erforschung von nächtlichen Lichtphänomenen". That translates roughly to the "citizen/Helmholtz network for the study of night light phenomena. The German word Bühne means "stage", and citizen scientists will really be at center stage in the project.

Within the project, citizen scientists will be developing two apps (in cooperation and consultation with the project coordinators). The first, coordinated by my colleague Friederike Klan at DLR, is about improving the system for reporting sightings of fireballs in Germany. The second, coordinated by me at GFZ, is about creating several large-scale inventories of ourdoor lighting. That's the part I am writing about in this post.

Our previous research has shown that light emissions are growing at a rate of over 2% per year worldwide, including in wealthy countries like Germany. The problem is, we don't know what types of lamps are responsible for the growth. For that matter, when we look at low-resolution satellite imagery, we don't even know what types of lamps are responsible for the emissions.


Composite image of Berlin from the Suomi NPP Day/Night Band.

Even with higher resolution imagery from space or the air, we still can't tell how much light comes from, for example, street lamps compared to advertising lighting:


Berlin photographed from the International Space Station.

Our goal within the artificial light part of Nachtlicht-BüHNE is to obtain more complete information from the ground that we can use to improve our understanding of the satellite imagery. For example, when the DNB instrument in space says that the radiance is "35 nW/cm2sr", how many light sources on the ground are responsible for it? 100? 1000? And how do the types of light sources change depending on whether you are in a small town, the suburbs, or a city center?

In the Nachtlicht-BüHNE lamp project, large teams of citizen scientists will work together to count all of the lamps that are visible from public spaces over very large areas. By the project end, we expect to have at least 3 regions in which we have counted all of the light sources visible from public spaces covering 2km2. Of course, we hope that the teams of citizen scientists will manage more than 3 regions, and maybe even larger areas, but that's getting ahead of ourselves...


Before we can develop an app, we need to figure out how to make it work well. Our first goal is therefore to start building up categories of what types of lamps are out there. It's easy to think of some obvious frequent types: street lights, illuminated windows, bus shelters, lit advertising, facade lighting. But there's a lot more! Gas station canopies, strings of lanterns, in ground luminaires, porch lighting...

The list goes on and on (and on and on), so in order to make sure we cover everything, and even more importantly, to make sure we organize things logically, we're going to need input from you. For the next 10 days, we're asking people around the world to go out at night and walk short transects with a friend, recording light source types, counting the number of each, and recording how long it takes

Here's what we suggest you do:
  • Find a friend
  • Choose a starting location together (eg. an intersection)
  • Get a large notepad and a pencil or pen, and then go to the starting location
  • Note the time
  • Start walking slowly down the street
  • Have your friend call out the different lamp types he or she sees, while you start recording them on a table
  • When you reach a logical stopping point (eg. the next intersection), write down the time you finished
  • Optional: do another street!
  • When you are finished, go home and report the data to us using this form in English (or this form for German)
  • Let us know your thoughts about the project from within the form, including ideas about how the app might work
    •  Note: DO NOT put your name, email, or any identifying information in the form! We want the data to be completely anonymous at this stage so that we can easily share it with our citizen scientist team without any privacy concerns. But we do welcome your correspondence. You can comment on this blog post, or contact Helga Kuechly at the email address at the bottom of the online form.
  • IMPORTANT SAFETY REMINDERS:
    • Do not make observations alone! With two people, you will do a better job of maintaining situational awareness. It will also be a lot more fun with a friend! 
    • If you live in an area with aggressive policing, we recommend you do not take part in this phase. Please wait until we have a printed form or the app which you could show to any police officers who wonder why you are taking notes about a neighborhood at night.

You can categorize the data in any way you want. Should you keep track of the color of the lamps? The height of the poles? How bright they seem? Whether they emit light upwards? It's entirely up to you! Do what you think is logical, needed, and/or easy. Our aim is to develop a system that provides useful data while also being really easy for anyone without background in outdoor lights to take part.

We are accepting data in this test phase until September 15. Our plan is then to meet with our team of citizen scientists developing the app to develop a standardized paper form, and then in late September or early October we'll run another campaign to see how well the standardized form works.

If you can speak German and would like to be part of the team of citizen scientists doing the app development and testing, please contact Helga Kuechly. Please also contact Helga if you would like to arrange an observation campaign in your community (no German required!), or help develop an English version of the app. If you would like to get our project newsletter by email, you can sign up here

Data submission form in English // data submission form in German

Monday, September 2, 2019

New Zealand imaged by starlight

It's never really dark at night. Besides starlight, the Earth's atmosphere itself glows. Both airglow and starlight can be reflected from the surface of the Earth. Since plants, rock, soil, and snow have higher reflectance than water, in a nighttime image they should appear brighter.

However, there is only one satellite that images the entire Earth each night, and it's just barely sensitive enough to see this. For example, here is what a portion of the South Island of New Zealand looks like when the data from all the clear moon-free nights of October 2017 is averaged together:

This work by Christopher Kyba is licensed under a
Creative Commons Attribution 4.0 International License.

You can see the mountain ranges, and you can just barely see the separation between land and water. But the more data you add, the more the signal comes out from the noise. My colleague Helga Kuechly produced this image by combining the seven October moon-free composite images from the years 2012-2018:


"Aotearoa by starlight" by Helga Kuechly is licensed under
a Creative Commons Attribution 4.0 International License.

In addition to making a stunning image, it demonstrates that if we continue to work on increasing the sensitivity of night observing satellites, it is possible to extend traditional visible band remote sensing even to nights without moonlight!