Interpolation of daily solar radiation for New Zealand using a satellite data-derived cloud cover surface
Daily values of global solar radiation are important inputs for pasture growth models which are often run for rural locations (e.g. where detailed pasture growth data have been collected). However, it is possible that the nearest climate station with daily solar radiation data is located tens of kilometres away from these pasture measurement sites. This paper describes a method of interpolating daily global solar radiation data recorded or estimated at climate stations throughout New Zealand onto a regular 0.05° latitude / longitude (approximately 4 km) grid using a thin plate smoothing spline model. An analysis of the mean annual solar radiation prediction standard error from several spline model runs shows that the lowest prediction error is obtained using two positional variables and a satellite data-derived cloud cover surface. For the daily interpolations, maximum use is made of observational data by combining daily solar radiation and sunshine measurements collected over the period 1972 – 2006. An analysis of the daily solar radiation interpolation error at 20 validation sites shows that the average root mean square error varies between about 4.1 MJ/m2/day in the summer months to about 1.5 MJ/m2/day in winter (average relative error of between 20 and 30%). Time series data derived from the daily gridded solar radiation data are currently being used in a pasture growth model for New Zealand.