Abstract

This paper describes a new computationally efficient and statistically robust sampling method for generating dynamically downscaled climatologies. It is based on a Monte Carlo method coupled with stratified sampling. A small yet representative set of "case days" is selected with guidance from a l... Show moreThis paper describes a new computationally efficient and statistically robust sampling method for generating dynamically downscaled climatologies. It is based on a Monte Carlo method coupled with stratified sampling. A small yet representative set of "case days" is selected with guidance from a large-scale reanalysis. When downscaled, the sample closely approximates the long-term meteorological record at a location, in terms of the probability density function. The method is demonstrated for the creation of wind maps to help determine the suitability of potential sites for wind energy farms. Turbine hub-height measurements at five U.S. and European tall tower sites are used as a proxy for regional climate model (RCM) downscaled winds to validate the technique. The tower-measured winds provide an independent test of the technique, since RCM-based downscaled winds exhibit an inherent dependence upon the large-scale reanalysis fields from which the case days are sampled; these same reanalysis fields would provide the boundary conditions to the RCM. The new sampling method is compared with the current approach widely used within the wind energy industry for creating wind resource maps, which is to randomly select 365 case days for downscaling, with each day in the calendar year being represented. The new method provides a more accurate and repeatable estimate of the long-term record of winds at each tower location. Additionally, the new method can closely approximate the accuracy of the current (365 day) industry approach using only a 180-day sample, which may render climate downscaling more tractable for those with limited computing resources. Show less