Building-resolving large-eddy simulations for entire Berlin (Germany) – first results using the high-performance urban microscale model PALM-4U

Abstract

Due to the increasing number of people living and/or working in dense urban environments, the importance of city planning in consideration of human health and comfort has been continuously growing. Health and comfort factors such as thermal comfort, air quality, ventilation and UV exposure, must be considered in a future-oriented development of urban regions. For decision support, urban climate models (UCM) are applied to model the effects of existing and planned building distributions, facade and city greening, etc., based on the above mentioned factors. A highly-efficient microscale UCM, PALM-4U, has been developed allowing simulations of large cities with grid-resolved buildings and vegetation canopy, which consider a large variety of processes important for urban environments. PALM-4U is part of the PALM model system (http://palm-model.org), which is based on the large-eddy simulation code PALM. While there exist numerous UCMs that have been used for over two decades and which are well established in the scientific community, they are difficult to adapt to state-of-the-art parallel computer systems and thus often have limitations in either performance and/or possible number of grid points. PALM-4U is able to compute entire city environments like Berlin (about 1 700 km²) at building-resolving grid spacing (here 10 m) on massively parallel computers, where limitations are mainly imposed by the available computational resources. It offers several features required in urban environments, such as an energy balance solver for urban and natural surfaces, radiative transfer in the urban canopy layer, chemical reactions, biometeorological analysis products, and self-nesting to allow high resolution (e.g. 1 m) in regions of special interest. In this presentation we will focus on an overview of PALM-4U's current and planned capabilities for application in urban environments. Besides, we will demonstrate PALM-4U's performance and features based on microscale building-resolving large-eddy simulation of entire Berlin (Germany, 1700 km²) at a grid spacing of 10 m, with a nested domain of size of 1 km² at a grid spacing of 1 m. The simulation spans a simulated period of a full diurnal cycle during a selected heat wave period and is characterized by low geostrophic winds and a strong solar forcing during daytime. General features of the simulation will be visualized. This presentation is intended to be the first in a series of presentations that all have a more specific focus on single features of PALM-4U and for which abstracts are submitted separately. In this overview talk we will thus focus on the more general features of the simulation.