Clouds are an important modulator of the global radiation budget and yet representing their formation, evolution and interaction with aerosols still remains as one of the largest uncertainties in modelling future climate. An important requirement to understanding the processes which govern clouds is accurate measurement of their global distribution and microphysical properties over a wide range of spatial and temporal scales which can only be satisfied by passive remote sensing measurements from satellite platforms. As such the development and validation of cloud remote sensing techniques is an important ongoing task. Of particular radiative importance are marine stratocumulus clouds, due to their large global extent and high solar reflectance. This thesis uses a range of in situ and remote sensing observations of marine stratocumulus over the South East Pacific taken during the Variability of the American Monsoon Systems (VAMOS) Ocean-Cloud-Atmosphere-Land Study Regional Experiment (VOCALS-REx) to investigate some outstanding issues relating to passive remote sensing. In particular answers to two questions are sought: 1) Do measurements of solar reflectance at multiple wavelengths with different absorption properties allow information about the vertical structure of the cloud to be derived? 2) Is there a high bias in passive retrievals of droplet effective radius? A unique airborne hyperspectral data set is evaluated for its potential to provide insight into these problems but through extensive comparison to collocated in situ and satellite observations along with an analysis of historical calibrations, it is concluded that the calibration quality of this dataset is not sufficient to meet its scientific objectives. A theoretical study into the information content of multi-wavelength measurements to retrieve the vertical variation of droplet size is presented. Measurements from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite instrument are shown to contain little information related to the vertical structure of typical marine stratocumulus. The information content of hyperspectral measurements is shown to be significantly larger, indicating the potential to perform profile retrievals from future measurements. A comparison of in situ profile measurements to collocated MODIS cloud retrievals adds to the existing body of evidence that passive retrievals of the droplet effective radius of marine stratocumulus are high biased when compared to other measurement sources.Potential sources of this bias are investigated and many of the previously postulated reasons behind the bias are ruled out. It is also shown that the differences between MODIS retrievals of effective radius performed at different wavelengths bear no relation to the in situ observed vertical structure of the cloud.