IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 57, NO. 4, APRIL 2019 2161 Improving Nighttime Light Imagery With Location-Based Social Media Data Naizhuo Zhao , Wei Zhang, Ying Liu, Eric L. Samson, Yong Chen, and Guofeng Cao Abstract— Location-based social media have been extensively utilized in the concept of “social sensing” to exploit dynamic information about human activities, yet joint uses of social sensing and remote sensing images are underdeveloped at present. In this paper, the close relationship between the number of Twitter users and brightness of nighttime lights (NTL) over the contiguous United States is calculated and geotagged tweets are then used to upsample a stable light image for 2013. An associated outcome of the upsampling process is the solution of two major problems existing in the NTL image, pixel saturation, and blooming effects. Compared with the original stable light image, digital num- ber (DN) values of the upsampled stable light image have larger correlation coefficients with gridded population (0.47 versus 0.09) and DN values of the new generation NTL image product (0.56 versus 0.52), i.e., the Visible Infrared Imaging Radiometer Suite day/night band image composite. In addition, total personal incomes of states are disaggregated to each pixel in proportion to the DN value of the pixel in the NTL images and then aggregate by counties. Personal incomes distributed by the upsampled NTL image are closer to the official demographic data than those distributed by the original stable light image. All of these results explore the potential of geotagged tweets to improve the quality of NTL images for more accurately estimating or mapping socioeconomic factors. Index Terms— Location-based social media, nighttime light (NTL) imagery, United States demographics, upsampling. I. I NTRODUCTION N IGHTTIME lights (NTL) imagery has an extensive record of utilization in human systems research. Since Croft first recognized its potential to monitor emissions of waste gas from oil fields in 1973 [1], NTL imagery has been used to estimate and disaggregate socioeconomic fac- tors (see [2]–[13]), map urban expansion (see [14]–[19]), determine urban forms/patterns (see [20], [21]), and detect impacts of human activities on natural systems (see [22]–[25]). The Defense Meteorological Satellite Program’s Operational Linescan System (DMSP-OLS) annual stable light image com- posites are presently the most widely used NTL image data Manuscript received September 28, 2017; revised March 10, 2018 and June 18, 2018; accepted August 27, 2018. Date of publication October 24, 2018; date of current version March 25, 2019. This work was supported by the Office of the Vice President for Research and the College of Arts and Sciences at Texas Tech University. (Corresponding author: Guofeng Cao.) N. Zhao, Y. Liu, and G. Cao are with the Center for Geospatial Tech- nology, Texas Tech University, Lubbock, TX 79409 USA, and also with the Department of Geosciences, Texas Tech University, Lubbock, TX 79409 USA (e-mail: guofeng.cao@ttu.edu). W.Zhang and Y. Chen are with the Department of Computer Science, Texas Tech University, Lubbock, TX 79409 USA. E. L. Samson is with the Mayan Esteem Project, Farmington, CT 06032 USA. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TGRS.2018.2871788 because of its global coverage with relatively long temporal span. Ephemeral lights, such as fires and lightning, have been removed from the stable light image composites, which greatly increases the utility and dependability of the NTL images as a measure of socioeconomic factors [26]. However, the saturated pixels in the stable light image products hinder the applica- tions for areas with highly dense anthropogenic activities [7]. In addition, urban peripheries and water bodies inside cities tend to be extensively brightened by urban lights, resulting in lit areas much larger than actual urban/developed areas [14]. This distinctive phenomenon in NTL imagery is usually called blooming [15]. The existence of saturated pixels and blooming effects in DMSP-OLS stable light image products leads to large under-estimations in urban areas and over-estimations in suburban and rural areas when brightness of NTL is selected as a proxy to estimate or spatially disaggregate socioeconomic factors [13]. In 2013, the National Oceanic and Atmospheric Adminis- tration’s (NOAA) National Centers for Environmental Infor- mation (NCEI) (formerly National Geophysical Data Center) released a new generation NTL image products named the Visible Infrared Imaging Radiometer Suite (VIIRS) day/ night band (DNB) image composites collected by the Suomi National Polar-orbiting Partnership (SNPP) satellite. Compared with the DMSP-OLS stable light image prod- ucts, the VIIRS-DNB composites have a larger quantiza- tion range (14 bit versus 6 bit) to avoid the issue of pixel saturation. However, blooming effects, inherent to NTL, still exist in the VIIRS-DNB images. In addition, VIIRS-DNB image composites have a finer spatial resolu- tion (0.00416° versus 0.00833°) than the DMSP-OLS NTL images [27]. The difference in spatial resolution raises another issue of NTL images: how to upsample the DMSP-OLS NTL image products to make their spatial resolution con- sistent with the VIIRS-DNB composites for joint uses in practice. To overcome the problems of saturation and blooming and to improve spatial resolutions of the NTL images, auxiliary data that have the following two characteristics are often needed: 1) close correlation to brightness of NTL and 2) the ability to be formatted to raster layers with finer spatial resolutions than the NTL images [13]. Previous studies have demonstrated that the brightness of NTL strongly correlates too many socioeconomic variables (e.g., population, gross domestic product, electric power consumption, and fossil fuel carbon dioxide emissions) [2]–[13], [28], [29]. However, data of these variables are mostly survey or census based and 0196-2892 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. Authorized licensed use limited to: Texas Tech University. Downloaded on July 18,2020 at 19:03:28 UTC from IEEE Xplore. Restrictions apply.