VIIRS Day-Night Band Comparisons of Northeastern City Lights

As clouds are starting to dissipate with the slow demise of Sandy, we are getting some beneficial peeks through the clouds with the VIIRS Day-Night Band (DNB).  As we demonstrated with Isaac in the New Orleans area, the DNB is great for inferring quite a bit about power outages and even cloud cover through interactions of emitted light versus reflected moonlight.  We have assembled two examples of the DNB imagery over the northeastern United States.  The first image is from the morning of October 26.  Although some low clouds were present across several states, they were not thick enough to completely mask the cities below, resulting in a vibrant display of light (though diffuse) throughout the major cities extending from Washington D.C./Baltimore through Boston.

This morning, November 1, we obtained some additional city light views (and some cloud cover remnants from Post-Tropical Cyclone Sandy) across the area.  As you toggle between the first two slides, your eyes will pick up some loss of city lights along the coastal areas of New Jersey, Long Island, and portions of southern Maine.  News reports confirm power outages in many of these areas.  Unfortunately, some of the lingering cloud cover makes it a bit difficult to make strong claims about power outages in more areas, as some of the missing city lights may be cloud-obscured.  However, these coastal areas appear to be cloud free given that the city lights that *are* present are appearing as fine, small points of bright light that are not being diffused or scattered by clouds at higher altitude.  We can highlight areas where we feel pretty confident that the changes are due to power loss, not cloud cover, but it can be tricky given that even thin clouds can obscure the lights below – similar to viewing cities from above in an airplane.  There may be other areas where city lights are out but not fully viewable in the difference image – for example, portions of eastern Massachusetts would likely have had similar impacts, but this morning’s cloud cover makes it difficult to determine.  In these images, the clouds that are present are visible from reflected moonlight given that we are still at or near to the full moon phase.

As skies continue to clear, there may also be other means of watching the recovery by looking for city lights that reappear.  This is just another great example of VIIRS applications in response to disasters.  SPoRT provides this type of imagery via our website and to NWS partnering offices and National Centers that we collaborate with through NOAA’s Joint Polar Satellite System (JPSS) Proving Ground.  The final imagery here are derived from data obtained from the Cooperative Institute for Meteorological Satellite Studies (CIMSS) at the University of Wisconsin Madison and are processed using a lunar reflectance algorithm provided to us by colleagues at the Cooperative Institute for Research in the Atmosphere (CIRA) at Colorado State University in Ft. Collins, Colorado.

Day night band reflectance imagery prior to the approach of Sandy, observed on October 26. City lights show up as bright white areas since they emit far more light than would be reflected by moonlight alone.

Day night band reflectance imagery a few days after the landfall of Sandy, seen here on the morning of November 1. Many city light areas are missing, though some are also obscured by cloud cover remnants from the storm.


Annotated image helping to identify some areas where power outages may be occurring, inferred from loss of bright city lights in the VIIRS day-night band. Note that other areas such as eastern Massachusetts may also have some outages, but remaining clouds obstruct our view.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s