The VIIRS-DNB radiance product is insufficient to assess the effect of "cool pavement" materials on nighttime radiances of treated areas

TL;DR

This study evaluates whether VIIRS-DNB radiance data can reliably detect the impact of cool pavement treatments on nighttime radiance, finding the data insufficient for conclusive assessment due to measurement limitations.

Contribution

The paper demonstrates the limitations of VIIRS-DNB data in detecting radiance changes from cool pavement treatments in urban environments.

Findings

Radiance increases from CPT are not conclusively detected.

VIIRS-DNB data limitations hinder robust assessment of CPT effects.

No significant radiance change beyond 14% was observed.

Abstract

In warmer climates, cities face the prospect of higher air temperatures in summer compared to historical averages due to the urban heat island effect. An approach intended to address this problem is the application of "cool pavement" treatments (CPT) to city streets to make them more reflective to sunlight. Raising the albedo of roadways for this purpose may also have the effect of increasing the amount of street light that is reflected into the night sky. The simplest hypothesis explaining the relationship between CPT application and upward radiance is that CPT applied to road surfaces in areas where street lighting is dominant should increase the upward radiance of neighborhoods where the treatments are applied. A simple model predicted radiance increases of 2-6% immediately after CPT application. To test the hypothesis and model predictions, we looked for radiance changes coinciding with the application of CPT in residential neighborhoods of Phoenix, U.S., since 2020. We obtained time series radiance measurements from Visible Infrared Imaging Radiometer Suite Day-Night Band (VIIRS-DNB) data from Radiance Light Trends for Phoenix neighborhoods receiving CPT and nearby "control" neighborhoods not receiving CPT. At the 95% confidence level, we found that any increases in nighttime radiances from treated neighborhoods did not exceed about 14%. As a consequence, we cannot rule out either the expected radiance increases from our model or the possibility that CPT application yielded no change in radiance. We therefore cannot draw robust conclusions about the potential influence of CPT deployment on skyglow given the limitations of the DNB as a data source.