Abstract:
A systematic investigation of the bulk radiative characteristics of homogeneous water clouds at solar wavelengths has been made using extensive computer simulation involving the application of Mie theory and the Monte Carlo method of studying radiative transfer. Practical parameters for describing cloud for radiative calculations have been identified and their relationship to the bulk radiative characteristics investigated. By fitting polynomials to the many relationships examined a model has been developed that enables rapid determination of bulk radiative characteristics from the limited information that is available in many situations. The model includes means of accounting for cloud shape and absorption by both water droplets and water vapour. Where possible, results from the model have been compared with those in published work. In all cases the agreement has been good.
Droplet size distribution is well characterised by droplet number, cross-sectional area and volume, unless large droplets are included. Quantitative assessment shows that, contrary to previous researchers expectations, if large droplets are present further information is required to adequately account for droplet size distribution. Cloud shape is well characterised by the dimensions of the cloud together with a single parameter, the mean depth in the direction of incident radiation. Modelling of non-conservative transfer is facilitated by the introduction of absorptances for reflected and transmitted photons and by their relationship to variables describing averaged photon pathlengths. This method may enable extension of the model to broken cloud layers.
The empirical model may find wide application. Its speed makes it a particularly useful investigative tool and this, together with its structure, makes it suitable for the inverse problem of determining cloud features from radiative measurements. As the model is based on practical parameters it is ideal for obtaining estimates of bulk radiative characteristics from observed or predicted variables.
