Satellite-based Cloud Remote Sensing: Fast Radiative Transfer Modeling and Inter-Comparison of Single-/Multi-Layer Cloud Retrievals with VIIRS

Abstract

This dissertation consists of three parts, each of them, progressively, contributing to the problem of great importance that satellite-based remote sensing of clouds.

In the first section, we develop a fast radiative transfer model specialized for Visible Infrared Imaging Radiometer Suite (VIIRS), based on the band-average technique. VIIRS, is a passive sensor flying aboard the NOAA’s Suomi National Polar-orbiting Partnership (NPP) spacecraft. This model successfully simulates VIIRS solar and infrared bands, in both moderate (M-bands) and imagery (I-bands) spatial resolutions. Besides, the model is two orders of magnitude faster than Line-by-line & discrete ordinate transfer (DISORT) method with a great accuracy.

The second and third parts are going to investigate the retrieval of single-/multi- layer cloud optical properties, especially, cloud optical thickness (τ) and cloud effective particle size (De) with different methods. By presenting the comparison between results derived from VIIRS measurements and benchmark products, potential applications of Bayesian and OE retrieval methods for cloud property retrieval are discussed. It has proved that Bayesian method is more suitable for single-layer scenarios with fewer variables with fast speed, while Optimal Estimation method is superior to Bayesian method for more complicated multi-layer scenarios.