Content area
Full text
1. Introduction
Wind energy is an increasingly important part of the energy portfolios as the amount of installed wind power production grows each year. The U.S. Department of Energy aspires to have 30% of power coming from wind by 2030 while several states have set their own standards that call for significant increases in renewable power production over the next decade (U.S. Department of Energy 2008). A major challenge to the efficient integration and utilization of increased wind power production is the variable nature of the wind resource. Rapid changes in wind speed, known as wind ramps, produce fluctuations in wind power generation that pose a challenge to grid operators who must ensure that sufficient power is on the grid to satisfy demand and to individual wind farms that need to anticipate the amount of power that they can sell. Variations in wind speed are therefore a concern throughout the wind energy industry and an improved understanding of wind ramps and their predictability is essential for increasing wind energy’s presence on the grid (Shaw et al. 2009; Banta et al. 2013).
One approach to investigating wind ramps and their predictability is ensemble sensitivity analysis (ESA), which relates a scalar forecast metric to an atmospheric variable at an earlier time (Hakim and Torn 2008; Ancell and Hakim 2007a; Torn and Hakim 2008). Several studies have successfully used ESA to understand the atmospheric features relevant to the evolution of different high-impact phenomena. Applications of ESA to the minimum surface pressure of midlatitude cyclones over the midwestern United States showed a sensitivity to the surface cold front and highlighted the upper-level trough at 500 hPa (Hakim and Torn 2008). For landfalling midlatitude cyclones, sea level pressure in western Washington State was shown to be sensitive to synoptic features throughout the depth of the troposphere and frequently exhibited a sensitivity to the geopotential height field associated with the developing cyclone (Ancell and Hakim 2007a; Torn and Hakim 2008). For tropical cyclones, the minimum sea level pressure associated with two typhoons undergoing extratropical transition was found to possess sensitivity to the 500-hPa geopotential height that likely reflected both position and intensity changes (Torn and Hakim 2009). Zack et al. (2010) applied ESA to wind power forecasts and found that an...