Periodic reminder that any effective policy for light pollution control (not just 'mitigation') unavoidably requires limiting artificial light emissions.

Trust, but verify: Reducing artificial light emissions and monitoring compliance
In order to reduce light pollution we have to reduce the overall amount of artificial light emissions. This is a consequence of the basic rules governing the propagation of light in the terrestrial environment. In this work I revisit the physical laws causing that (i) "all" artificial light emitted outdoors is lost or pollutant, (ii) the negative effects of light pollution depend monotonically on the local concentration of photons of anthropogenic origin, and (iii) this concentration depends linearly on the total artificial light emissions, weighted by the light pollution propagation functions. Setting total emission limits becomes necessary in order to ensure that the negative effects of light pollution do not surpass red-lines of unacceptable degradation of the natural night. Once these red-lines are socially agreed, monitoring compliance becomes a relevant task. Several complementary methods for assessing total light emissions are being used nowadays, including public inventories of installed lights, direct radiance measurements from ground or low Earth orbit satellites, and scattered radiance measurements using ground based detectors (night sky brightness monitoring). While updated administrative inventories could in principle be trusted, independent verification is a must. The required measurements pose in turn significant challenges: we discuss here how the variability of the terrestrial atmosphere sets a lower limit on the minimum emission changes that can be reliably detected by measurements, and propose some ways to improve this performance.