Meteorological stations are often non-representative over complex topography, heterogeneous landscapes and urbanized areas. Access to increasing computing resources and advanced numerical models made practical to use statistical-dynamical downscaling (SDD) [1] to obtain local climate information. The SDD computes high-resolution, dynamically self-consistent meteorological fields for typical regional climate regimes. There are usually 50-150 of such regimes [2], which could be found by a statistical analysis or other methods. The added value (as compared to global/regional climate model results) is in the SDD capturing essential meteorological phenomena induced by local surface features and forces. A model resolving impact of such features on the atmosphere have an advantage over a model based on parameterizations. In this sense, a limited area turbulence-resolving large-eddy simulation (LES) model is a prospective tool for the services.
A turbulence-resolving model or LES [3] is a numerical model solving dynamical and transport equations at resolution (1-10 m) permitting explicit computation of the energy-containing non-universal part of the turbulence spectrum. Because of this high resolution, LES utilize only very simple parameterizations (a LES error scales as D in the power (-2/3), where D is the grid size) but remain computationally expensive. LES provide high-frequency data on the air-surface interaction, meso-scale atmospheric flows and turbulence in regimes where our theoretical understanding in insufficient for a less expensive statistical description. In this work, the Parallelized Atmospheric Large-eddy Model (PALM) of Institute for Meteorology and Climatology (IMUK) at University of Hannover is used (www.palm.hannover.de) [4]
What can LES provide for climate services? As defined, LES resolves the atmospheric turbulence, meteorological meso-scale flows and turbulent interactions with heterogeneous surfaces. It is strong in providing essentially turbulence-dependent quantities. This study presents three prospective directions of the LES application: (a) Local wind fields, wind gusts and wind load vulnerability – a direct use for the regional climate services; (b) Geographically-locked correlation analysis – an indirect use to support geo-statistical interpolation and micro-climate mapping at high-resolutions; (c) Scenario simulations – an application to support strategic planning simulating scenarios of the micro-climate response, e.g. scenarios for urban air quality. We will refer to these LES applications as an Ultimate Climate Downscaling (UCD). Studies of the types (a) and (c) have been reported [5,6], whereas the type (b) study is a novel approach under development at NERSC.
Elements of the Ultimate Climate Downscaling: (i) Classification splits climate records (reanalysis/gridded data) in typical climate regimes; (ii) Each regime is simulated with the LES model; concrete situations within the regime could be simulated with the model nudging or other types of data assimilation; (iii) The model results are used to assess the spatial variability and correlations of the fields; if observations are available, the model output is corrected with observed data; (iv) Statistics of the regimes is combined with spatial maps and high-frequency extreme assessment to obtain vulnerability maps and scenarios of micro-climate changes.
Due to stochastic nature of turbulence, deficit of high-resolution observations and computational cost, the turbulence-resolving simulations or LES cannot be utilized for the regional climate services in a traditional way as a simulation of a succession of meteorological fields. Nevertheless, LES can be utilized in frameworks of the statistical-dynamical downscaling approach or UCD where they provide valuable information for vulnerability and scenario assessments. Moreover, LES open a possibility for a new climate services – a high-resolution micro-climate mapping of the complex territories sparsely covered with observation network.
References.
[1] Frey-Bunnes F, Heimann D, Saussen R. (1995): Theor. Appl. Climatol., 50, 117-131
[2] Badger J., et al. (2014): J. Appl. Meteor. Climatol., 53, 1901–1919
[3] Esau I. (2004): Environmental Fluid Mechanics, 4, 273-303
[4] Letzel M., Raasch S., et al. (2012): Meteorologische Zeitschrift, 21(6), 575-589
[5] Ishihara T., Yamaguchi, A., Fujino, Y. (2006): CWE2006, 533-536
[6]Park S.-B., et al. (2012): J. Appl. Meteorol. Climatol., 51(5), 829-841