IN13C-3653:
Improving Cloud-screening for Global Nighttime Lights Composites Using the VIIRS Day/Night Band

Monday, 15 December 2014
Kimberly Baugh1, Christopher Elvidge2, Mikhail N Zhizhin1 and Feng Chi Hsu1, (1)Cooperative Institute for Research in Environmental Sciences, Boulder, CO, United States, (2)NGDC NOAA, Boulder, CO, United States
Abstract:
The Earth Observation Group (EOG) at the National Oceanic Atmospheric Administration's National Geophysical Data Center (NOAA/NGDC) has built a 22 year record of annual global nighttime lights composites using data from the Defense Meteorological Satellite Program's Operational Linescan System (DMSP-OLS). Using the newly available data from the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) allows EOG to extend this record beyond the life of the DMSP-OLS sensors, and create vastly improved products due to the increased spatial resolution and dynamic range of the DNB over the OLS sensor. Creation of high quality nighttime lights composites requires filtering of data degraded by th presence of clouds, heavy aerosols, or other atmospheric conditions that alter the sharpness and brightness of the nighttime light signal. The VIIRS Cloud-Mask (VCM) product is currently being used to define cloud-impacted regions in the creation of VIIRS DNB nighttime lights composites. Artifacts in the VCM, however, allow for some known cloud-impacted regions to be included in the nighttime lights composites, and conversely some known cloud-free regions to be excluded from the nighttime lights composites. This study investigates alternative methods to define regions of the nighttime DNB imagery that are impacted by atmospheric conditions. The first method filters the DNB using a spectral sharpness index. The second method examines shape of peaks in the DNB imagery to assess atmospheric impacts on the nighttime light signal. These two methods are compared against using the VCM for creation of nighttime lights composites using the VIIRS DNB.