Aeris is currently partnered with Lawrence Berkeley National Laboratory (LBNL) to provide subject matter expertise, modeling and simulation support, and technical assistance required to evaluate and further enhance the current operational suite of urban/indoor contaminant dispersion modeling capabilities developed by the Defense Threat Reduction Agency (DTRA). For this effort, the Aeris team is supporting the development and demonstration of a next-generation synthetic modeling environment generator—a GPU-based large eddy simulation (LES) model with an in-line tracer capability—that can synthesize environments with physically realistic peak/mean concentrations and spatio-temporal gradients for an urban location. The synthetic datasets are calculated up to 150x faster than those computed by a CPU-based system.
Aeris has performed extensive verification and validation (V&V) of the tracer capability employed by the GPU-LES for an open-terrain synthetic environment. Recently, Aeris’ partner Wiffle, Ltd. incorporated a multi-grid capability into the GPU-LES which will allow for the flow near the boundaries to be unaffected by the buildings (see figure below). Aeris is in the process of applying the same methods used in an open-terrain environment for urban environments using observations from the Mock Urban Setting Test (MUST) and Joint Urban 2003 field experiments. Aeris will be submitting a journal article on the open-terrain V&V effort for unstable, neutral, and stable boundary layer scenarios in late 2018. A follow-on publication for the urban V&V is expected to be submitted at the beginning of 2019.
The synthetic datasets derived from the GPU-LES will be utilized to evaluate the performance and identify limitations of the urban/indoor contaminant modeling capabilities contained within the Hazard Prediction and Assessment Capability (HPAC). HPAC is an operational and forward-deployable decision aid, which allows operators and analysts to estimate the consequences from a Chemical, Biological, Radiological, and Nuclear (CBRN) release event. It utilizes advanced source term models for properly estimating the amount of material released into the atmosphere for a variety of release mechanisms, which are linked to advanced array of physics-based transport and dispersion models to predict the subsequent contaminant transport in both rural and urban environments. Finally, downwind exposure levels and impacts to the population are estimated based on human effect and physiological models. In recent years, an initial capability to further model the infiltration to and exfiltration from physical structures has been added to the HPAC system, but needs more comprehensive V&V before being made available for operational utilization.
Illustration of two notional contaminant releases simulated within the urban synthetic environment generator using buildings from Chicago, IL, USA.