Development and Evaluation of an Urban Atmospheric Inverse Modeling Framework for the Washington, DC and Baltimore, MD Metropolitan Area: Initial Results From an Inert Tracer Case Study

Abstract

Accurate quantification of hazardous material releases in cities is critical for emergency managers aiming to mitigate health impacts and mortality in the population. Here, a study on the development and evaluation of a high‐resolution (1‐km, 1‐hr) urban‐scale atmospheric inverse modeling system for the Washington, DC, and Baltimore, MD, metropolitan area (DCBA) is presented to advance surface flux quantification in urban and industrial areas where the potential for a hazardous release is high. The system employs an inert tracer as a proof of concept for January 2019 and builds upon HYSPLIT atmospheric transport and dispersion simulations driven by high‐resolution Weather Research and Forecasting (WRF) model simulations that ingest urban meteorological observations, coupled with the CarbonTracker‐Lagrange (CT‐L) inverse model tailored to account for hourly spatiotemporal fluxes. The system assimilates tower‐based tracer observations from the National Institute of Standards and Technology (NIST) Northeast Corridor Urban Test Bed and utilizes the 1‐km Vulcan and 4‐km GRA2PES bottom‐up inventories as prior knowledge of the tracer fluxes. Numerical experiments using synthetic and actual data reveal that the system significantly improves the quantification of flux strength (mean flux) and total mass released across the urban cores of Washington, DC, and Baltimore, MD. Hourly flux variations are fairly well resolved when originating from a broad territory (city scale), while the accuracy of source retrieval improves when flux estimates are aggregated to coarser spatial resolutions (e.g., 4 km). This work has important implications for estimating hazardous releases from urban areas for emergency response and hazard analysis.