Gap-Filled Daily Global Water Transparency Products from Multi-VIIRS Measurements

Global Secchi depth (ZSD) data, a key indicator of water transparency, have recently been derived from the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the NOAA-21, NOAA-20, and Suomi National Polar-orbiting Partnership (SNPP) satellites. However, there is significant data loss in a daily ZSD image from a single sensor due to clouds, extreme satellite (viewing) angles, sun glint contamination, and other adverse conditions. In this study, we develop a global gap-filled ZSD dataset by integrating multi-sensor measurements through data fusion. Specifically, three-sensor (VIIRS) merged Level-3 global ZSD are generated, and then the Data Interpolating Empirical Orthogonal Function (DINEOF) method is applied on the three-sensor merged data to reconstruct missing values within the valid processing domain and generate a global Level-4 gap-filled ZSD dataset. It is found that a three-sensor integrated daily global ZSD image contains ~18% more satellite measurements than a two-sensor image. In terms of evaluation of reconstructed ZSD against the original data, three-sensor gap-filled data demonstrate a good reconstruction agreement with small overall bias, i.e., mean and standard deviation (STD) of reconstructed/original ratio are ~0.995 and ~0.135, respectively. High-resolution (2 km) global daily gap-filled ZSD are also derived by integrating observations from the three VIIRS sensors, and they can reveal high-frequency variations and fine features in global coastal oceans and inland lakes. As more data from future VIIRSs and other sensors are integrated, it is highly expected that this will further improve the performance of reconstructing ZSD values at higher spatial resolutions.