PREEVENTS Track 2: A fast-response wildland fire modeling framework for prediction and risk assessment

野火在美国越来越普遍，对人口稠密地区产生了重大影响。目前的快速反应管理是基于40多年前开发的经验或半经验模型。该项目的重点是创建一个多阶段野火研究和预测系统（MWRPS），该系统连接了几个现有的社区，开发了开源模型。这个新模型有可能改变火灾的研究方式，并显著改善野火和烟雾的预测。MWRPS将为需要确定野火和烟雾的社会和生态影响的消防专业人员、城市和环境规划者以及灾害管理人员提供工具。作为一种社区模式，MWRPS将向公众开放，以审查各种与灾害有关的问题。除了研究生和博士后培训外，该项目还包括通过Hi-GEAR计划进行K-12推广的计划。火线处水流的突然变化对极端火灾行为和相关危险至关重要。由于火线上的气流可以显著影响火灾的蔓延和随后的所有野火行为，因此从天气尺度到火线尺度，耦合大气-火气流的预测和模拟必须准确。准确的预测需要能够模拟当地风的快速天气驱动变化，真实地呈现复杂地形中的流量，并捕捉火灾本身对当地天气的影响。为了应对这一挑战，多阶段野火研究和预测系统（MWRPS）将基于基本流体动力学原理开发多尺度模型，3D模型。MWRPS将能够解析建筑物、树木和土地覆盖，结合复杂地形、不同植被类型和几何形状的影响，分散烟雾，并表示野火环境中的辐射、感热和潜热。该模型预测的野火特性将包括火灾周长和强度的高分辨率时空演变；表面和顶部火灾的行为；在荒地城市界面（WUI）或通过树冠的烟雾产生和扩散（对于规划拟议的规定燃烧至关重要）；以及烟雾浓度和热通量对安全区和WUI结构的影响。除了模型开发之外，还将讨论极端野火行为的三个方面：(1)燃料异质性，(2)复杂地形和(3)火灾相互作用的作用。将为大气火灾模型开发一个数据驱动系统，以统计合理的方式引导来自多种来源的模拟，包括：天气数据、传感器、机载火灾图像和卫星遥感。

Wildfires are increasingly common throughout the US and have significant impacts on populated areas. Current rapid-response management is based on empirical or semi-empirical models developed more than 40 years ago. The focus of this project is to create a Multistage Wildfire Research and Prediction System (MWRPS) that links several existing community, open source models developed. This new model has the potential to change how fire is studied, and significantly improve operational wildfire and smoke forecasting. MWRPS will serve as tool for fire professionals, urban and environmental planners, and disaster managers who need to determine the societal and ecological impacts of wildfire and smoke. As a community model, MWRPS will be available to the public to examine a variety of hazard-related issues. In addition to graduate student and postdoctoral training, the project includes a plan for K-12 outreach through the Hi-GEAR program.Sudden changes in flow at the fire line are crucial to extreme fire behavior and associated hazards. Because flow at the fire line can significantly impact fire spread and subsequently all wildfire behavior, prediction and simulation of coupled atmosphere-fire flow must be accurate from synoptic down to fire line scales. Accurate prediction requires the ability to model rapid synoptically-driven changes in local winds, realistically render flow in complex terrain, and capture the impacts of the fire itself on local weather. To meet this challenge, the Multistage Wildfire Research and Prediction System (MWRPS), a multi-scale model, 3D model will be developed based on fundamental fluid dynamical principles. MWRPS will have the ability to resolve buildings, trees, and land cover, to incorporate the effects of complex terrain, different vegetation types and geometries, to disperse smoke, and to represent radiation, sensible, and latent heating in the wildfire environment. Predicted wildfire properties from this model will include high resolution temporal and spatial evolution of the fire perimeter and intensity; behavior for both surface and crown fires; smoke production and dispersion in the Wildland Urban Interface (WUI) or through tree canopies (crucial for planning a proposed prescribed burn); and impacts of smoke concentrations and of heat flux in safety zones and on WUI structures. In addition to the model development, three aspects of extreme wildfire behavior will be addressed: the roles of (1) fuel heterogeneity, (2) complex topography, and (3) fire interactions. A data-driven system will be developed for atmosphere-fire models to steer simulations from a multitude of sources including: weather data, sensors, airborne fire images, and satellite remote sensing in a statistically sound manner.