Skyglow: measuring light pollution in the UK

Skyglow is the halo-like artificial illumination of the night sky above towns and cities. It’s also the name of a major new study into light pollution in the UK by Hillarys. Using satellite images taken between 1992 and 2014, Skyglow shows how the UK’s night-time skies have changed. The research charts a 28% decrease in light pollution*, with falls in every region – and projects this data into the future. You can see the interactive satellite imagery and full study on the Skyglow project page. Or keep reading to find out the story behind the story.

Credit: Skyglow Project

Credit: Skyglow Project

We knew what to expect with Skyglow. From the start, we expected to find that light pollution in the UK is rising.

Our expectations stood to reason. The population has grown 10% since 1992 and infrastructure must have developed to meet its needs, so light pollution should have increased too. Except – it hasn’t.

Skyglow discovered something way more interesting: night skies are darker across the country.

Why we chose to investigate Skyglow

Hillarys is a gold sponsor the International Year of Light and Light-based Technologies (IYL 2015). With our sponsorship, we wanted to look at how light affects our lives in both small and big ways, from our smartphones to photosynthesis to looking at the night sky.

The idea of investigating the night sky had our attention. We started to think about how our use of artificial lighting is causing light pollution and affecting the brightness of the night.

Day by day, our research uncovered more and more about the consequences light pollution has for our lives. We spoke to Dr. Bob Mizon MBE FRAS, who told us many people – and even well-equipped astronomers – struggle to see star-filled skies because of light pollution.

Our investigations turned to the energy issue. Despite some improvements in lighting design and efficiency, we found in the EU more than half the energy used for outdoor lighting is wasted.

Dr. John Barentine explained such careless lighting has an economic cost – and a cost to our quality of life. He said our biology and ecology are ill-equipped to handle night-time brightness in extreme measures and can suffer even in the presence of relatively small amounts of light. We found this kind of exposure to night-time light can disrupt our brainwave patterns, hormone production and cell regulation.

We decided to conduct our own research into light pollution.

What is the Skyglow project?

Skyglow came out of this initial research. We had pondered light pollution and whether it would have changed over the years. To us, at this point in the process, we were confident of finding a change and probably an increase.

Our ideas developed. Soon we were investigating ways of researching light pollution – one potential method stood out immediately.

Skyglow would take statistical data from night-time satellite images and chart historical trends in light pollution. Our method would allow us to put a value on how much artificial light escapes into the atmosphere nationwide and to break our findings down to a regional level.

In addition, we would look at the effects light pollution has on us and the wider environment. And we would use digital design to raise awareness of our findings and the issue online.

Brightness of the night sky over the UK in 2014 calculated using images from two different satellites, the Operational Linescan System (OLS) and the Visible Infrared Imaging Radiometer Suite (VIIRS). Credit: Skyglow Project.

Brightness of the night sky over the UK in 2014 calculated using images from two different satellites, the Operational Linescan System (OLS) and the Visible Infrared Imaging Radiometer Suite (VIIRS). Credit: Skyglow Project.

How we created Skyglow

With the help of Jurij Stare from Dark Skies Slovenia, we settled on using images from the Defence Meteorological Satellite Programme (DMSP). For each year, we created a single composite image from hundreds of satellite photographs, which allowed us to take account of day-to-day variations.

The DMSP records analogue images, so we converted these into a greyscale digital format. On each image, different shades of grey showed the varying intensity of artificial light escaping into the atmosphere – or the amount of light pollution being created.

So to create the data needed to compare light pollution over time, we awarded each shade of grey a value from 0 to 63 (0 being the darkest and 63 being the brightest). From here, we could work pixel by pixel and show light pollution increases and decreases as percentages.

The next step was to change the map from a WGS84 coordinate system (the kind used on Google Earth 3D) to a spherical Mercator (the standard map projection, found on Google Maps).

Last but not least, we applied a colour ramp, turning the greyscale images into a coloured and easy-to-understand format.

What Skyglow found

Our work shows that light pollution in the UK is down 28% since 1992*.

The findings took us by surprise. Sure, we knew that the way energy is consumed and lighting technology has changed over the years, but the UK population has grown over the same period. And it would make sense if the rising population had caused a growth in infrastructure, which would have a knock-on effect for light pollution.

We needed a further opinion. So we turned again to Dr. John Barentine and Bob Mizon MBE FRAS, and also to Professor Martin Morgan Taylor. These established experts verified our research and helped us better understand the causes of our findings.

For example, the experts explained that while there are more streetlights, many modern designs feature guards that direct light towards the ground. This helps prevent light escaping into the atmosphere and causing Skyglow.

Region by region, the UK was experiencing a decrease in light pollution. Here are just a few of the statistics we discovered.

  • North Scotland down 40%
  • The West Country down 41%
  • Northern Ireland down 32%
  • Yorkshire down 29%
  • London down 14%

If the trend continues at the current rate, our data projects a further 21% decrease in light pollution between 2015 and 2025.

See how UK light pollution has changed – and use Skyglow

Skyglow didn’t end with the results. In many ways, this was the beginning.

The next step was to present our findings in a way that would help us raise awareness of the issue.

You can see our work at www.hillarys.co.uk/skyglow/ and use the interactive maps we created to show how UK light pollution has changed. You’ll also find out more about the project. We hope you enjoy Skyglow.

Note from the editor

1 – After the publication of the post, we have some comments from scientists pointing out that the study could be biased because the satellite data used might be no sensitive to LED lighting and to study skyglow measure consider surface light is not sufficient. For more information, see http://www.mdpi.com/2072-4292/7/1/1.


Sunny1Sandeep Matharu is the manager of Hillarys’ Living Light campaign and its International Year of Light sponsorship.

Advertisements

2 thoughts on “Skyglow: measuring light pollution in the UK

  1. The DMSP is not sensitive to light below about 500 nm – the region where LEDs have their strongest emission. So if a street switches to LED it would look like a decrease in light to this analysis. DMSP also doesn’t look at large angles, so even partially shielded lights are not visible to it. To study skyglow, you either need to use a radiative transfer model together with assumptions about the light emission, or else you need to actually measure skyglow. Measuring the surface light isn’t sufficient. We have a paper about this here: http://www.mdpi.com/2072-4292/7/1/1

    Furthermore, the DMSP is an uncalibrated sensor, and it’s unlikely that the gain was stable between 1992-2014. Despite being stable, the DNB sensor on VIIRS will also be insufficient for this type of analysis, unless you have a priori information about the spectrum of the lamps.

    If you’d like to actually measure skyglow, there are a number of ways to do it. I wrote about three in another IYL blog piece: https://light2015blog.org/2015/02/09/help-measure-how-the-night-sky-is-changing/

  2. Hi Chris,

    The image source was the Version 4 DMSP-OLS Nighttime Lights Time Series. Which was then inter-satellite and inter-year calibrated based on the coefficient from Chris Elvidge’s manuscript located below. The national trends data set is based on inter calibrated DMSP stable lights, an annual cloud-free composite of average digital brightness value for the detected lights, filtered to remove ephemeral lights and background noise.

    DMSP OLS is sensitive in the 500 – 900nm region. It is true that LEDs have a strong peak in the 450nm, but the integral emission above 500nm is still larger and it gets larger with lower LED temperature. Already at temperature of 4000K the integral emission below 500nm only accounts for 20% of total. You can see the graph here: http://agi32.com/blog/wp-content/uploads/2014/12/Photosynthesis-4000K-LED.jpg

    What is the percentage of the 4000K, 5000K, 7000K LEDs out there? LEDs are not the only light sources that emit light below 500nm and some of them we have been using for decades.

    The 2012 manuscript from the National Trends in Satellite Observed Lighting (link below) adds weight to the fact there’s been a reduction:

    “In the antipole group are some of the richest and most well developed countries in the world – Canada, UK, Japan, USA, Netherlands and Belgium. In these countries aggregate lighting has declined despite expansions in population numbers and GDP. Our interpretation of this decline in lighting is that ongoing improvements in lighting efficiency are offsetting the growth lit infrastructure. Over the past two decades there has been a proliferation of state and local regulations designed to conserve energy and protect the night sky from light pollution. This includes banning incandescent lights and improving shielding to limit the quantity of light which shines directly into the sky.”

    There are problems with the methods suggested for measuring sky glow as a lot of people would be needed for the whole planet and second, the actual sky brightness varies by as much as 0.5 mag/sq. arcsec (that’s close to 100% of a value since the scale is logarithmic) due to atmospheric conditions which means you again need a lot of measurements.

    Satellite data is the only consistent source of data we have and unfortunately nothing else comes close to it.

    http://ngdc.noaa.gov/eog/dmsp/download_national_trend.html
    http://ngdc.noaa.gov/eog/dmsp/downloadV4composites.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s