Assessment of Two Types of Interior Nudging for High-Resolution Simulations during Oahu, Hawaii's 40-Days and 40-Nights Extreme Precipitation Event

Abstract:

Dynamical Downscaling for climate length simulations is a process where large-scale atmospheric fields are input to a regional climate model (RCM) to explicitly simulate regional and local- scale climate features that are not captured by the large-scale model. One such example of features not being captured occurs over the Central North Pacific's Hawaiian Island chain, where most large-scale models poorly resolve the small islands and their complex topographies, which have a significant influence on the regions winds, temperature, and precipitation. For the dynamical downscaling procedure to realistically resolve the small Hawaiian Islands we must use three nested, where the innermost domain centered over Oahu, Hawaii has a grid resolution of 1.1 km. Also during the dynamical downscaling process RCMs interior solution tends to drift away from the large-scale driving fields resulting in significant RCM temperature and precipitation biases especially during extreme events. Two possible solutions developed to allow the RCM to retain the large-scale features, yet still generate the small scale variabilities are analysis and spectral nudging. Here, we examine the performance of both analysis and spectral nudging in the downscaling of NCEPII reanalysis data during the extreme 2-month wet period known as the 40-days and 40-nights of rain over the Hawaiian Islands using the Advanced Research Weather and Forecasting (WRF-ARW) model. The simulations are compared against land based observations acquired from the National Climate Data Center (NCDC) to show the differences for 2-m temperature and precipitation between the nudging techniques and observations from Oahu, Hawaii.