A common conclusion from wind integration studies is the benefit of spatial diversification of Wind Power Plants for power systems. However, few of these studies quantify the benefit that may be apparent from different wind power portfolios. To quantify that benefit, temporally and spatially accurate models of wind power are required. A wind power model is constructed starting with wind speed time-series extracted from the ECMWF-interim reanalysis model. The wind speed time-series are interpolated, scaled, and imputed such that they are representative of the wind incident on the Wind Power Plants. Imputation is performed using a Wavelet Multi-Resolution Analysis approach that ensures temporally consistent correlations while accommodating heteroskedasticity. The wind speed time-series are transformed to power by applying wind power plant power curves, low pass filters, and a Markov Chain model for operational efficiency. Simulated wind power time-series are validated using a set of measurements made at Wind Power Plants in New Zealand. The wind power model is used to simulate power time-series for 2 GW portfolios of wind power plants representing compact, disperse, diverse, and Business As Usual portfolios. Metrics for dependability, variability, and predictability are used to quantify the benefits of spatial diversification.