The proposed work will evaluate the ability of operational and experimental versions of the High Resolution Rapid Refresh (HRRR) modeling system for the continental United State and Alaska to forecast the characteristics of mesoscale atmospheric boundaries arising from thunderstorm outflows, gust fronts, and downburst winds (referred collectively as convective outflows). The objective is to lead to enhanced situational awareness within the operational fire weather community of the ability of the HRRR model and predictive tools that rely on its output to nowcast and forecast convective outflows. Expected benefits from this project include: 1) improved understanding of the ability of HRRR data assimilation/forecast systems to forecast convectively-driven changes in wind, temperature, and moisture that influence fire behavior, which will assist fire weather forecasters in their ability to incorporate high-resolution model output in probabilistic convective outflow forecasts on time scales < 24 h. 2) improved understanding of the impacts of operational model uncertainties in convective outflow characteristics on fire spread scenarios. This will be conducted using: (1) a fire behavior tool that will rely on HRRR output combined with landscape characteristics and fuelbed flammability and (2) simulations of selected case studies using a coupled fire-atmosphere model, WRF-SFIRE. 3) improved communication methodologies for fire weather forecasters, incident meteorologists, and fire behavior analysts to inform alerts and warnings of potential and imminent risks to fire managers and fireline personnel who will be most affected by abrupt changes in weather conditions near wildfires. The research team led by the University of Utah and University of Alaska-Fairbanks has extensive experience assessing the benefits and limitations of weather forecasts for fire management applications. We will classify and examine many cases in which convective outflow boundaries were identified by on-site fire personnel to have affected fire behavior in complex terrain. We will also objectively identify convective-outflow signatures in the vicinity of wildfires using a wide-array of meteorological resources to identify their critical characteristics, e.g., duration, intensity, speed, and seasonal and time-of-day dependencies. Using an archive at the University of Utah of HRRR forecasts in the continental United States and Alaska from 2017-2018, we will then assess the ability of HRRR forecasts to detect the presence of outflow boundaries and the fidelity of the forecast guidance available from that model system to forecast them on a range of temporal and spatial scales. We hypothesize that model guidance will be most useful for: (1) nowcasts of longer-lived (< ~6 h) outflows that develop typically within preferred synoptic-mesoscale situations and (2) forecasts that provide improved situational awareness for the potential for outflows at lead times from 6-24 h rather than accurate depiction of each outflow. Validation of the numerical analyses and forecasts will be completed using a mix of spatial statistical techniques and subjective evaluation emphasizing the overall situation from the perspective of what a forecaster or fire behavior analyst would be able to establish from ~24 hours prior to until immediately before any major changes in the atmospheric state that would affect fire behavior. The proposed study will equip incident meteorologists and fire behavior analysts assigned to major fires and NWS WFO forecasters and predictive service personnel in the continental United States and Alaska with improved situational awareness of the potential for outflows and their impacts on fire behavior when they examine HRRR output operationally. In order to do so, operational personnel will help evaluate how to interpret the results of the validation studies and design procedures and best practices to inform the wider community.