2. Study Area and Data
We selected eight globally distributed regions from different land cover and land use types, and urban development transitions, corresponding to various NASA Black Marble Level-3 VIIRS tiles (latitude/longitude extent of 10°10° per tile): h10v04, h13v11, h18v04, h17v08, h29v05, h32v12, h11v07, and h21v05 (Fig. 1). The analysis employed all available 15-arc-second spatial resolution daily DNB atmospheric- and Lunar-BRDF-corrected and Top of Atmosphere (TOA) Black Marble NTL radiance collected between 2013 and 2020.
We also manually interpreted 610 calibration samples that have undergone various kinds of changes, such as construction actions, economic growth, gathering events, armed conflict, power outages, and new streetlights, as well as stable samples without any NTL changes, based on the opportunistic strategy around the major cities and metropolitan areas within the study area (Table 1). To explore the complex angular effects, the calibration samples were selected from areas with typical types of local geometry conditions, such as downtown areas with skyscrapers, single- and multi-story residential regions, and areas with dense vegetation canopy. For each calibration sample, we recorded the interval of each change event for the period between 2013 and 2020. Note that certain NTL changes, such as transition changes, defined as changes in long-term trend, can last for more than a year.