Indicators of Electric Power Instability from Satellite Observed Nighttime Lights

Electric power services are fundamental to prosperity and economic development. Disruptions in the electricity power service can range from minutes to days. Such events are common in many developing economies, where the power generation and delivery infrastructure is often insufficient to meet demand and operational challenges. Yet, despite the large impacts, poor data availability has meant that relatively little is known about the spatial and temporal patterns of electric power reliability. Here, we explore the expressions of electric power instability recorded in temporal profiles of satellite observed surface lighting collected by the Visible Infrared Imaging Radiometer Suite (VIIRS) low light imaging day/night band (DNB). The nightly temporal profiles span from 2012 through to mid-2020 and contain more than 3000 observations, each from a total of 16 test sites from Africa, Asia, and North America. We present our findings in terms of various novel indicators. The preprocessing steps included radiometric adjustments designed to reduce variance due to the view angle and lunar illumination differences. The residual variance after the radiometric adjustments suggests the presence of a previously unidentified source of variability in the DNB observations of surface lighting. We believe that the short dwell time of the DNB pixel collections results in the vast under-sampling of the alternating current lighting flicker cycles. We tested 12 separate indices and looked for evidence of power instability. The key characteristic of lights in cities with developing electric power services is that they are quite dim, typically 5 to 10 times dimmer for the same population level as in Organization for Economic Co-operation and Development (OECD) countries. In fact, the radiances for developing cities are just slightly above the detection limit, in the range of 1 to 10 nanowatts. The clearest indicator for power loss is the percent outage. Indicators for supply adequacy include the radiance per person and the percent of population with detectable lights. The best indicator for load-shedding is annual cycling, which was found in more than half of the grid cells in two Northern India cities. Cities with frequent upward or downward radiance spikes can have anomalously high levels of variance, skew, and kurtosis. A final observation is that, barring war or catastrophic events, the year-on-year changes in lighting are quite small. Most cities are either largely stable over time, or are gradually increasing in indices such as the mean, variance, and lift, indicating a trajectory that proceeds across multiple years.

[1]  Stephen V. Stehman,et al.  Selecting and interpreting measures of thematic classification accuracy , 1997 .

[2]  M. Delignette-Muller,et al.  fitdistrplus: An R Package for Fitting Distributions , 2015 .

[3]  Changyong Cao,et al.  Radiometric Inter-Consistency of VIIRS DNB on Suomi NPP and NOAA-20 from Observations of Reflected Lunar Lights over Deep Convective Clouds , 2019, Remote. Sens..

[4]  Thomas Barnebeck Andersen,et al.  Power outages and economic growth in Africa , 2013 .

[5]  C. Elvidge,et al.  Limiting the impact of light pollution on human health, environment and stellar visibility. , 2011, Journal of environmental management.

[6]  B. Selby,et al.  The index of dispersion as a test statistic , 1965 .

[7]  Steven D. Miller,et al.  Illuminating the Capabilities of the Suomi National Polar-Orbiting Partnership (NPP) Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band , 2013, Remote. Sens..

[8]  T. Croft Nighttime Images of the Earth from Space , 1978 .

[9]  R. Lund,et al.  Changepoint Detection in Periodic and Autocorrelated Time Series , 2007 .

[10]  Massoud Amin North America's Electricity Infrastructure: Are We Ready for More Perfect Storms? , 2003, IEEE Secur. Priv..

[11]  C. Elvidge,et al.  VIIRS night-time lights , 2017, Remote Sensing of Night-time Light.

[12]  Xi Shao,et al.  Detecting Light Outages After Severe Storms Using the S-NPP/VIIRS Day/Night Band Radiances , 2013, IEEE Geoscience and Remote Sensing Letters.

[13]  Paulina Jaramillo,et al.  Sustainability implications of electricity outages in sub-Saharan Africa , 2018, Nature Sustainability.

[14]  C. Elvidge,et al.  Early damaged area estimation system using DMSP-OLS night-time imagery , 2004 .

[15]  Richard A. Frey,et al.  The VIIRS Cloud Mask: Progress in the first year of S‐NPP toward a common cloud detection scheme , 2014 .

[16]  J. O’Loughlin,et al.  Detecting the Effects of Wars in the Caucasus Regions of Russia and Georgia Using Radiometrically Normalized DMSP-OLS Nighttime Lights Imagery , 2011 .

[17]  V. Murray,et al.  Power Outages, Extreme Events and Health: a Systematic Review of the Literature from 2011-2012 , 2014, PLoS currents.

[18]  D. Moyer,et al.  VIIRS day-night band gain and offset determination and performance , 2012, Optics & Photonics - Optical Engineering + Applications.

[19]  R. Maini Study and Comparison of Various Image Edge Detection Techniques , 2004 .

[20]  Morgan Bazilian,et al.  Measuring “Reasonably Reliable” access to electricity services , 2020 .

[21]  R. Mead,et al.  On the power of the index of dispersion test to detect spatial pattern , 1979 .

[22]  N. Altman An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression , 1992 .

[23]  Arun Malik,et al.  Using VIIRS Day/Night Band to Measure Electricity Supply Reliability: Preliminary Results from Maharashtra, India , 2016, Remote. Sens..