Decomposition of the large-scale atmospheric state driving downscaling: a perspective on dynamical downscaling for regional climate study

Abstract

Abstract In this study, we provide a perspective on dynamical downscaling that includes a comprehensive view of multiple downscaling methods and a strategy for achieving better assessment of future regional climates. A regional climate simulation is generally driven by a large-scale atmospheric state obtained by a global climate simulation. We conceptualize the large-scale state based on reconstruction by combining decomposed components of the states, such as climatology and perturbation, in different global simulations. The conceptualization provides a comprehensive view of the downscaling methods of previous studies. We propose a strategy for downscaling regional climate studies based on the concept of covering a wider range of possibilities of large-scale states to account for the uncertainty in global future predictions due to model errors. Furthermore, it also extracts the individual influences of the decomposed components on regional climate change, resulting in better understanding of the cause of the change. We demonstrate a downscaling experiment to highlight the importance of the simultaneous consideration of the individual influences of climatology and perturbation.

Keywords

• Dynamical downscaling
• Regional climate
• Uncertainty
• Model errors