A spatiotemporal comprehensive constraint interpolation method for missing values in monthly composite Suomi-NPP VIIRS Cloud Mask images

Abstract

ABSTRACT The Earth’s nighttime light imaging data collected by the Visible Infrared Imaging Radiometer Suite Day/Night Band (VIIRS DNB) are severely lost in summer, hindering the further application of VIIRS DNB data. In this article, a spatiotemporal comprehensive constraint interpolation method (SCIM) that accounts for the relative stationarity of temporal nighttime light, spatial correlations and heterogeneity among pixels is designed. Meanwhile, the method considers the seasonal fluctuations in nighttime light radiation, introduces spatial autocorrelation coefficients and spatial heterogeneity coefficients, and solves the problem of missing VIIRS DNB data. First, 8 widely used temporal interpolation methods (TIM) are selected to perform temporal interpolation for missing image data. Second, the existing VIIRS monthly images are used to construct the constraint relationship between time and space, and the pixels that do not meet the constraint conditions in the TIM simulation image are removed. The inverse distance weight (IDW) method is applied to the removed null pixels to obtain the image simulated by the SCIM. Finally, the accuracy of the TIM and SCIM is evaluated. The results show that the accuracy of the SCIM-simulated image is better than that of the TIM-simulated image. The proposed approach improves the quality of the VIIRS DNB dataset.