Exploring the Drivers of Variations in Daily Nighttime Light Time Series From the Perspective of Periodic Factors

Periodic factors, including lunar phase cycle, sensor viewing angle, and seasonality, synergistically affect daily nighttime light (NTL) observations, which bring difficulties to the correction and application of daily NTL data. Here, assuming Lhasa is an ideal observation area, we quantified the periodic characteristics of NTL data with the normalized difference between hotspot and darkspot (NDHD) NTL radiance and further utilized the econometric models to verify the individual and synergistic effects. Our results showed periodic factors differentially affected NTL observations. Sensor viewing angle and lunar phase cycle mainly dominated the variations in NTL time series at the pixel and urban scale, respectively. Then, the econometric model results indicated different periodic factors had a significant individual effect on NTL observations. Lunar phase cycles positively increased observed NTL radiance; for the sensor viewing angle, the sensor zenith angle (VZA) had a nonlinear link with observed NTL radiance, which was regulated by the sensor azimuth angle (VAA). Spring and summer had a negative impact on NTL observations, whereas the opposite was true for autumn and winter. Moreover, a positive synergistic effect existed between the lunar phase cycle and sensor viewing angle, while the synergistic effects of seasonality and other periodic factors were heterogeneous. Our analysis provides a more comprehensive understanding of the temporal changes in the daily NTL time series.