Intercomparison of Radiative Transfer Models for TOA Radiance Simulation: Application to a CubeSat-Based Multispectral Sensor

Abstract

Satellite remote sensing plays a key role in the monitoring of inland water bodies, where accurate observations are essential for assessing processes, such as eutrophication, water quality degradation, and algal blooms. The accuracy of these observations depends critically on calibration and validation protocols, which must be adapted to the technical specifications of each sensor; and on the correct representation of atmospheric effects. A comparative analysis of two radiative transfer models (second simulation of a satellite signal in the solar spectrum-vector and SOS) was carried out for a ten-band multispectral camera designed to monitor continental waters, using a simulation-based framework supported by satellite and in situ data. This study focused on the influence of aerosol characterization and aerosol optical thickness sources on radiative transfer models. Hyperspectral products from the Precursore Iperspettrale Della Missione Applicativa mission, together with in situ reflectance measurements, were used to simulate Top of Atmosphere radiance of the instrument's bands over two Italian lakes: Lake Garda and Lake Trasimeno. Results show that for the two radiative transfer codes in this study, model performance is primarily driven by atmospheric inputs and atmospheric models rather than by the choice of the radiative transfer code itself, highlighting the importance of accurate atmospheric characterization for inland water applications. This approach can be transferable to other optical payloads and provides a practical framework for optimizing radiative transfer analysis strategies in small satellite missions.