Original paper

An approach for generating synthetic fine temporal resolution solar radiation time series from hourly gridded datasets

Perry, Matthew; Troccoli, Alberto

Meteorologische Zeitschrift Vol. 26 No. 3 (2017), p. 265 - 276

12 references

published: Jun 14, 2017
published online: Nov 23, 2016
manuscript accepted: Feb 12, 2016
manuscript revision received: Feb 6, 2016
manuscript revision requested: Dec 21, 2015
manuscript received: Sep 15, 2015

DOI: 10.1127/metz/2016/0746

BibTeX file


Open Access (paper can be downloaded for free)

Download paper for free


A tool has been developed to statistically increase the temporal resolution of solar irradiance time series. Fine temporal resolution time series are an important input into the planning process for solar power plants, and lead to increased understanding of the likely short-term variability of solar energy. The approach makes use of the spatial variability of hourly gridded datasets around a location of interest to make inferences about the temporal variability within the hour. The unique characteristics of solar irradiance data are modelled by classifying each hour into a typical weather situation. Low variability situations are modelled using an autoregressive process which is applied to ramps of clear-sky index. High variability situations are modelled as a transition between states of clear sky conditions and different levels of cloud opacity. The methods have been calibrated to Australian conditions using 1 min data from four ground stations for a 10 year period. These stations, together with an independent dataset, have also been used to verify the quality of the results using a number of relevant metrics. The results show that the method generates realistic fine resolution synthetic time series. The synthetic time series correlate well with observed data on monthly and annual timescales as they are constrained to the nearest grid-point value on each hour. The probability distributions of the synthetic and observed global irradiance data are similar, with Kolmogorov-Smirnov test statistic less than 0.04 at each station. The tool could be useful for the estimation of solar power output for integration studies.


solar irradiancesolar powersynthetictime seriesspatial variabilityfine temporal resolution