Forecasting is an inseparable part of modern renewable energy generation systems including wind, solar and wave. In order to effectively manage integration of grid-connected energy resources and energy storage systems, the energy generation potential of these resources needs to be estimated ahead in time. To this aim, this paper applies a simple and linear statistical forecasting technique, the Autoregressive Integrated Moving Average (ARIMA) on hourly global horizontal irradiance (GHI W/m2) data. The objective is to test the capabilities of ARIMA forecasting on high resolution solar time series for the state of Abu Dhabi, the United Arab Emirates and lay down the foundations for further in-depth analysis. The performance of the model is assessed using commonly used statistical metrics, coefficient of determination (R2) and root mean square error (RMSE). The model is trained and tested on hourly GHI from March, 2016. The R2 and RMSE values for the best fit model are found to be 88.63% and 72.88 W/m2, respectively.