Built-up Areas Mapping from Luojia 1-01 Nighttime Light Imagery with Considering Observation Number

Abstract

As the first global professional nighttime light (NTL) remote sensing satellite, the Luojia 1-01 (LJ1) has a finer resolution and an earlier visiting local time, compared with the Visible Infrared Imaging Radiometer Suite (VIIRS) sensor on the Suomi National Polar-orbiting Partnership (NPP) satellite. NTL imagery has been proved to be effective for mapping builtup areas (BUA). Concerning the performance of built-up areas mapping from LJ1, previous researches were all conducted with NPP data of the same period as benchmark. We found that LJ1 is significantly less effective than NPP in all but the mega/large cities. However, it should be mentioned that the LJ1 dataset was a spatial mosaic of multiple LJ1 observations of a single date, while NPP was a composite of the average observations of half a year. Therefore, it is of great necessity to study and analyze the influence of the number of NTL observations on the effect comparison of LJ1 and NPP extraction of BUA. In this paper, representative cities in different levels were sampled to consider the influence by the two following aspects: 1) the multi-date observations of LJ1 images and 2) the single-day NPP product (VNP46A1). Results show that the gap between LJ1 and NPP gradually narrows by increasing the number of NTL images for LJ1, and even LJ1 surpasses NPP. Moreover, the disadvantage of LJ1 over NPP will be improved in less developed regions, and the advantage can be further expanded in developed regions (e.g., the megacity and large city), when LJ1 images are available in large amounts. These findings have supplied important knowledge for the properties of the new LJ1 NTL images and guidance to its future research.