Developing an urban canopy model for improved weather forecasts in cities

世界上大多数人口现在都经历着城市版的极端天气和气候变化。城市天气和空气质量的准确预报依赖于正确地表示地表和上层大气之间湍流交换的物理学。每个建筑物产生一个复杂的流动，但通过处理他们的影响集体作为一个“城市冠层”，相对简单的模型可以制定。问题在于理解城市异质性如何影响流动过程：高层建筑与小型建筑的热通量如何变化？如果太阳只照到街道的一边而不照到另一边呢？一个单独的高层建筑物会支配表面阻力吗？该项目将探索建筑阵列周围的流动过程，为城市冠层模型的开发提供信息。该项目将采用为植被冠层开发的建模方法1，将城市冠层模型2扩展到非中性稳定状态。现有的数据（风洞，计算流体动力学，现场）将用于测试湍流交换过程的相似性。将设计和进行新的风洞实验，以研究更真实的城市冠层形态的非中性流。该项目获得了英国气象局的案例奖。统一模型中的当前城市表面方案（MORUSES）在单一层面上与模型耦合：实际上，建筑物是平坦的。城市地区包含越来越大的建筑物，可以占据边界层的很大一部分，单一的，表面预测是不够的，不明确的。因此，英国气象局城市建模策略的一个关键部分是开发一个垂直分布的计划，捕捉整个城市树冠层深度的动量和标量交换。MORUSES 3是一种简化的城市冠层热通量模型。该方案背后的概念基于2D街道峡谷，包括建筑物尾流和“非尾流”区域。MORUSE内的湍流交换方案是针对中性流态开发的，并使用街道峡谷热通量的风洞模型进行验证4。从2D框架转向3D框架是对真实的城市区域进行建模的必要步骤。对加热建筑物的风洞实验将在萨里大学的EnFlo进行。模型面-街道，墙壁，屋顶-将被加热，并使用快速响应传感器测量热通量。通过在中等流动中将模型加热到相对较低的温度，热充当被动标量。目的是探索不同的配置（三维建筑布局，加热模式）对湍流交换的影响。这种方法已经被用于研究植被冠层和街道峡谷5。将模型加热到更高的温度解决了热量是活跃标量的情况6。风洞入口条件也可以改变，以模拟建筑物上方的对流和稳定边界层。1 Harman and Finnigan 2007 BLM2 Coceal and Belcher 2004 QJRMS 3 Porson et al 2010 QJRMS 4 Barlow et al 2004 BLM5 Nogueira-Neto，2015 PhD（阅读）6 Marucci and Carpentieri 2019 BandE

Most of the world's population now experiences an urban version of weather extremes and climate change. Accurate forecasting of weather and air quality in cities relies on correctly representing the physics of turbulent exchange between the surface and the overlying atmosphere. Each building produces a complex flow but by treating their effect collectively as an "urban canopy", relatively simple models can be formulated. The problem lies in understanding how urban heterogeneity affects flow processes: how does the heat flux change from tall vs small buildings? What if sun heats one side of the street and not the other? Does a single tall building dominate surface drag? This project will explore flow processes around building arrays to inform development of an urban canopy model. The project would adapt a modelling approach1 developed for vegetation canopies to extend an urban canopy model2 to non-neutral stability regimes. Existing data (wind tunnel, CFD, field) would be used to test similarity of turbulent exchange processes. New wind tunnel experiments would be designed and conducted to investigate non-neutral flows for more realistic urban canopy morphologies. This project has a CASE award from the UK Met Office. The current urban surface scheme (MORUSES) in the Unified Model is coupled to the model at a single level: effectively, the buildings are flat. Urban areas contain increasingly large buildings that can occupy a significant fraction of the boundary layer, for which a single, surface prediction is inadequate and ill-defined. A key part of the Met Office urban modelling strategy is thus to develop a vertically distributed scheme that captures momentum and scalar exchange throughout the depth of the urban canopy. MORUSES3 was developed to represent urban canopy heat fluxes in a simplified way. The concept behind the scheme is based on a 2D street canyon, consisting of building wake and "non-wake" areas of the flow. The turbulent exchange scheme within MORUSES was developed for a neutral flow regime and validated using wind tunnel modelling of heat fluxes from street canyons4. Moving from a 2D to a 3D framework is a necessary step to modelling real urban areas. Wind-tunnel experiments with heated buildings would be done at EnFlo at the University of Surrey. Model facets - streets, walls, roofs - would be heated and heat fluxes measured using fast response sensors. By heating the models to relatively low temperatures in moderate flows, heat acts as a passive scalar. The objective would be to explore the impact of different configurations (3D building layout, heating patterns) on turbulent exchange. This methodology was already used to study vegetation canopies and street canyons5. Heating the models to higher temperatures addresses the regime where heat is an active scalar6. Wind-tunnel inlet conditions can also be varied to simulate convective and stable boundary layers above the buildings. 1 Harman and Finnigan 2007 BLM2 Coceal and Belcher 2004 QJRMS3 Porson et al 2010 QJRMS4 Barlow et al 2004 BLM5 Nogueira-Neto, 2015 PhD (Reading)6 Marucci and Carpentieri 2019 BandE