As PV systems proliferate and become an important part of the global energy mix, it is increasingly important to forecast their energy output in order to ensure a safe and reliable integration of their variable output into electric power grids. One of the main prerequisites for that is the detailed recording and interpolation of the actual irradiance in a spatially resolved way. Such 2D irradiance maps would also allow the assessment of the performance of the many PV systems that do not have irradiance sensors installed at the site. The maps are
ideally based on a dense network of irradiance sensors; however, in many cases the costs of high-precision pyranometers, real-time monitoring and frequent maintenance are prohibitive for such operational forecasting systems. On the other hand, many PV installations are in fact equipped with reference cells in the plane of array (POA) for evaluating and monitoring the performance of the systems. Adding this network of reference cells to existing pyranometer networks (from meteorological services or research institutes) would substantially help
in improving the accuracy of the irradiance maps. This paper introduces an irradiance conversion technique that allows POA irradiance measurements from an on-site reference cell to be converted to global horizontal irradiance data, which can then collectively be used to generate large-area irradiance maps.