Solar radiation model and an analysis of synoptic solar radiation regimes in British Columbia

Interest in the spatial and temporal distribution of solar radiation has increased recently, primarily as a consequence of the current interest in the potential utilization of solar energy. Since the existing solar radiation monitoring network in most areas is sparse and largely inadequate for solar energy feasibility studies, numerical modelling of solar radiation including its direct and diffuse components must be used to provide supplemental data. In addition to information on actual radiative values, it may be useful to attain knowledge about the temporal and spatial behaviour of solar radiation and a synoptic meteorological framework presents one possible approach. A cloudless sky solar radiation model based on the original work of Houghton (1954) was tested for the three Canadian locations of Goose, Nfld., Edmonton, Alta. and Port Hardy, B.C. In order to obtain better estimates of the direct and diffuse components of the total solar radiation, revised values of the aerosol parameter and forward- and back-scattering factors were determined. The model was also verified for other locations in British Columbia. The effects of cloud were incorporated through the use of a new cloud layer - sunshine (CLS) model which calculates the direct and diffuse components of solar radiation separately. When compared to a previously developed cloud layer model, the CLS model provided better daily estimates of the total solar radiation for the five Canadian locations of Goose, Nfld., Edmonton, Alta., Vancouver, B.C., Sandspit, B.C. and Summerland, B.C. The potential applicability of the CLS model to Canada is discussed. In order to study solar radiation from a synoptic viewpoint, synoptic weather types were established for British Columbia and the adjacent areas of the Pacific Ocean. An objective correlation classification technique was used for this purpose. The synoptic types were described and their frequency, persistency and sequence analysed. Since precipitable water is one of the required input parameters for the solar radiation model, these synoptic types were initially used to help specify the precipitable water fields for the region. The CLS model was applied to the British Columbia area for the year 1972 and the annual and seasonal solar radiation distributions were discussed. An assessment of the spatial variability of solar radiation showed that there is a considerable advantage to numerically modelling solar radiation in order to supplement the existing measurement network. The solar radiation distributions, based on both calculated and observed data, were related to the synoptic weather types and the variance of solar radiation between synoptic types was shown to be generally greater than the within type variance. This allowed the establishment of synoptic solar radiation regimes for the British Columbia area. However, it was shown that knowledge of synoptic weather types neither precludes the need for measuring solar radiation nor negates the advantage of numerically modelling solar radiation. Although some useful potential applications for a synoptic approach to the study of solar radiation were noted, it requires additional study to overcome the problems of seasonal variability, spatial scale and resolution in the synoptic - solar radiation relationships.Arts, Faculty ofGeography, Department ofGraduat