Seasonality and decomposition of monthly VIIRS-DNB image composites

Authors: Naizhuo Zhao*, Texas Tech University
Topics: Remote Sensing, Urban Geography, United States
Keywords: Nighttime lights imagery, VIIRS, time series, seasonality
Session Type: Paper
Day: 4/12/2018
Start / End Time: 8:00 AM / 9:40 AM
Room: Lafayette, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded

The new generation nighttime lights (NTL) image products, namely the Visible Infrared Imaging Radiometer Suite’s (VIIRS) Day/Night Band (DNB) image composites, are superior to the traditional Defense Meteorological Satellite Program’s (DMSP) Operational Linescan System (OLS) NTL image products due to the larger quantitative range and finer spatial and temporal resolutions. Moreover, with an onboard calibration, pixel values represent radiance and are theoretically compatible across different VIIRS-DNB images. However, the radiance of the VIIRS-DNB images is considerably affected by the albedo of land surface that often demonstrates apparent seasonal trends. In wintertime, for example, snow and ice cover dramatically increases backscatter of NTL, which leads to the brightness of NTL observed by the satellite abnormally larger than that in the summertime. In this study, we stacked all the monthly VIIRS-DNB images for the United States from January of 2014 to June of 2017 and then decomposed each of the time-series images into three individual images corresponding to seasonal, trend, and remainder components. We found that with the seasonal component removed from the original VIIRS-DNB images, radiance is more proportional to personal incomes at the county level. Additionally, a state’s changing trend of radiance more agrees with that of its GDP. These findings suggest that the seasonal decomposition is necessary before the current monthly VIIRS-DNB image products are used in practical socioeconomic studies.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login