Night lights

At night most of the light a satellite sees is ours, which makes it a measure of human activity.

The basics

At night, much of the Earth's light comes from us. A satellite camera sensitive enough to see street lamps, gas flares, fishing lights and fires gives a picture of human activity that does not rely on a survey or a census. Brighter and wider lit areas go with larger populations and bigger economies, a power cut shows as a dark patch, and a new settlement or industrial site appears as a new glow. The method is coarse, since one pixel reports the combined light of a neighbourhood, but it treats every country in the same way and it updates every month.

Why it matters

Night lights give the same view of every country, every month. They are used to estimate population and economic activity where statistics are thin, to see where the power has gone out, and to follow new settlements, industry and light pollution.

Brighter and wider glows go with more people and more activity, and a power cut shows as a dark patch. One pixel reports the light of a whole neighbourhood, so the view is coarse.

In depth

Night-lights sensors are low-light radiometers. They report radiance, the power emitted per unit projected area per unit solid angle, where $\Phi$ is power, $A$ the emitting area, $\Omega$ the solid angle and $\vartheta$ the angle from the surface normal.

L = \frac{\mathrm{d}^{2}\Phi}{\mathrm{d}A\,\cos\,\vartheta\,\mathrm{d}\Omega}

(1)

where $L$ is the radiance, integrated over the band of the sensor, in $\mathrm{W\,m^{-2}\,sr^{-1}}$ ; $\Phi$ is the radiant power, in $\mathrm{W}$ ; $A$ is the emitting area, in $\mathrm{m^{2}}$ ; $\Omega$ is the solid angle, in $\mathrm{sr}$ ; and $\vartheta$ is the angle between the direction of emission and the surface normal.

The unit is usually nanowatts per square centimetre per steradian, and one such unit equals $10^{-5}\ \mathrm{W\,m^{-2}\,sr^{-1}}$ . The instrument that made global night-time radiometry routine is the day/night band of the Visible Infrared Imaging Radiometer Suite, which gave the first quantitative measurements of nocturnal visible and near-infrared light from orbit and showed that moonlight, airglow and cloud change the background everywhere [1]. Figure 1 sketches what the sensor sees.

Cities and roads
Gas flares and industry
Fishing fleets
Moonlight, airglow and cloud change the background and are corrected for or screened out

Figure 1. What a low-light sensor sees at night. Lights from cities, flares and fishing fleets reach the sensor from the dark side of the Earth, on top of a background from the Moon, airglow and cloud.

A useful product needs substantial cleaning. Monthly and annual global composites are produced by excluding background noise, solar and lunar contamination, cloud-affected data, and sources unrelated to electric lighting such as fires, flares and volcanoes [2]. Studies of human activity then summarise the cleaned radiance over a region of interest $A$ , for example by a sum of lights over its pixels $i$ .

S = \sum_{i\in A} L_i

(2)

where $S$ is the sum of lights over the region, in the radiance unit of the product, usually $\mathrm{nW\,cm^{-2}\,sr^{-1}}$ ; $A$ is here the set of pixels in the region of interest; $i$ indexes those pixels; and $L_i$ is the cleaned radiance of pixel $i$ .

Early work showed that lit area and brightness track population, economic activity and electric power consumption [3], and later work used the sum of lights to estimate economic growth where official statistics are weak [4]. Night lights also feed maps of artificial sky brightness and light pollution [5]. The review sets out the limitations of current sensors and identifies the worldwide change to LED lighting as a particular challenge [6].

References

Miller, S., Straka, W., Mills, S. et al. (2013) Illuminating the Capabilities of the Suomi National Polar-Orbiting Partnership (NPP) Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band. Remote Sensing, 5(12), 6717-6766. https://doi.org/10.3390/rs5126717
Elvidge, C. D., Baugh, K., Zhizhin, M., Hsu, F. C. and Ghosh, T. (2017) VIIRS night-time lights. International Journal of Remote Sensing, 38(21), 5860-5879. https://doi.org/10.1080/01431161.2017.1342050
Elvidge, C. D., Baugh, K. E., Kihn, E. A. et al. (1997) Relation between satellite observed visible-near infrared emissions, population, economic activity and electric power consumption. International Journal of Remote Sensing, 18(6), 1373-1379. https://doi.org/10.1080/014311697218485
Henderson, J. V., Storeygard, A. and Weil, D. N. (2012) Measuring Economic Growth from Outer Space. American Economic Review, 102(2), 994-1028. https://doi.org/10.1257/aer.102.2.994
Falchi, F., Cinzano, P., Duriscoe, D. et al. (2016) The new world atlas of artificial night sky brightness. Science Advances, 2(6). https://doi.org/10.1126/sciadv.1600377
Levin, N., Kyba, C. C., Zhang, Q. et al. (2020) Remote sensing of night lights: A review and an outlook for the future. Remote Sensing of Environment, 237, 111443. https://doi.org/10.1016/j.rse.2019.111443