COMPARISON BETWEEN NIGHTTIME LIGHT AND SOCIOECONOMIC INDICATORS ON AN INTERNATIONAL SCALE USING VHRS DAY-NIGHT BAND

摘要

Observation of nighttime light is the representative method for monitoring human activities using satellite remote sensing. The Visible Infrared Imaging Radiometer Suite (VIIRS) sensor on the Suomi National Polar-orbiting Partnership (NPP) Satellite observes nighttime light at higher spatial resolution and with higher radiometric accuracy than the Defense Meteorological Satellite Program's Operational Linescan System (DMSP-OLS), and it increased the potential to utilize nighttime light data for urban monitoring. This study aims to evaluate the ability of the VIIRS nighttime light data to estimate socioeconomic indicators on an international scale. Firstly, the VIIRS nighttime light data is corrected by combining with the DMSP data for removing lights attributed to ephemeral events, and the total nighttime light (TNL) is calculated for each country. Then the TNL is compared with six indicators; GDP, population, electric power consumption, energy use, total road network and motor vehicles. As a result, the determinant coefficient (R~2) value between the TNL and GDP is 0.88 and this is almost the same value as that of the previous study in provincial units in China. The R2 value between the TNL and electric power consumption is 0.81, which is smaller than that in provincial unit of China. This result shows the electricity use efficiency differed from country to country. The R2 value of the TNL and population is 0.61, and the TNL values in developing regions are overall small for their population and those in developed regions are overall large for their population. This result is considered to show that as the economy grows, the lights emitted per person become bright. The R~2 value between the TNL and total road network is 0.69, and the R~2 value between the TNL and motor vehicles is 0.91. It is found that the TNL has a very good performance in estimating the number of motor vehicles.