Using Multi-Source Nighttime Lights Data to Proxy for County-Level Economic Activity in China from 2012 to 2019

Abstract

The use of nighttime lights (NTL) data to proxy for local economic activity is well established in remote sensing and other disciplines. Validation studies comparing NTL data with traditional economic indicators, such as Gross Domestic Product (GDP), underpin this usage in applied studies. Yet the most widely cited validation studies do not use the latest NTL data products, may not distinguish between time-series and cross-sectional uses of NTL data, and usually are for aggregated units, such as nation-states or the first sub-national level, yet applied studies increasingly focus on smaller and lower-level spatial units. To provide more updated and disaggregated validation results, this study examines relationships between GDP and NTL data for 2657 county-level units in China, observed each year from 2012 to 2019. The NTL data used were from three sources: the Defense Meteorological Satellite Program (DMSP), whose time series was recently extended to 2019; and two sets of Visible Infrared Imaging Radiometer Suite (VIIRS) data products. The first set of VIIRS products is the recently released version 2 (V.2 VNL) annual composites, and the second is the NASA Black Marble annual composites. Contrasts were made between cross-sectional predictions for GDP differences between areas and time-series predictions of economic activity changes over time, and also considered different levels of spatial aggregation.

Remotely too equal: Popular DMSP night‐time lights data understate spatial inequality

“…In support of the assumption that VNL or BM data provide truer values,Vuong (1989) non-nested tests based on the Kullback-Leibler information criterion show that models using VNL or BM data provide results that are closer to the truth than are models using DMSP data(Gibson, 2021;Zhang & Gibson, 2022).…”

Measurement errors in popular night lights data may bias estimated impacts of economic sanctions: Evidence from closing the Kaesong Industrial Zone

“…The data that support the findings of this study are openly available in OpenICPSR at https://doi.org/10.3886/ E192371V3, Gibson et al (2023) Elvidge et al (2013); this second reference notes that VIIRS data are up to 45-times more accurate than DMSP data. More recently, empirical comparisons with non-nested tests show that models using VIIRS data provide results that are closer to the truth than are models using DMSP data (Gibson, 2021;Zhang & Gibson, 2022).…”

“…(2013); this second reference notes that VIIRS data are up to 45‐times more accurate than DMSP data. More recently, empirical comparisons with non‐nested tests show that models using VIIRS data provide results that are closer to the truth than are models using DMSP data (Gibson, 2021; Zhang & Gibson, 2022).…”

“…10 10 Support for the assumption that the VIIRS data can be used as the true value comes from Vuong (1989) tests that show that models using VIIRS data provide results that are closer to the truth than are models using DMSP data (Gibson, 2021; Zhang & Gibson, 2022). …”

Tracking Spatiotemporal Patterns of Rwanda’s Electrification Using Multi-Temporal VIIRS Nighttime Light Imagery

“…The nighttime light data collected by the Day/Night Band (DNB) sensors of the Visible Infrared Imaging Radiometer Suite (VIIRS) on board the Suomi National Polar-Orbiting Partnership (S-NPP) satellite platform were used in this study. Recently, the National Aeronautics and Space Administration (NASA) has developed the Black Marble (VNP46) suite, which plays a vital role in research on urbanization [40], disaster-related power outages [41], proxies for economic activity [42], and the changing conditions in human settlements (i.e., urban energy access, migration, and disaster impact and recovery) [43]. This study employed the all-angle snow-free layer in the Black Marble annual nighttime light product (VNP46A4) from 2012 to 2020 (https://blackmarble.gsfc.nasa.gov/#product, accessed on 9 February 2022).…”