Designing an HLA Federation Object Model for Atmospheric Dispersion
Dr. Richard Penney, Dr. John Carson, Dr. Peter Hoare,
Malvern, Worcestershire, WR14 3PS, UK
Dstl Chemical and Biological Sciences
Porton Down, Salisbury, Wiltshire, SP4 0JQ, UK
Dr. Martyn Bull,
St Bartholomews Court, 18 Christmas Street, Bristol, BS1 5BT, UK
Dr. Tom Stark, Carl Simchick
Defense Group Inc,
2034 Eisenhower Ave, Suite 105, Alexandria, VA 22314
Cubic Defense Systems Inc, East Coast Engineering Division
6915 Telegraph Road, Alexandria, Virginia 22310
Defense Threat Reduction Agency
Chemical, Biological, Simulation, HLA, BOM, FOM, EnviroFed, IMPACT, WALTS
ABSTRACT: There is currently considerable interest in modeling the effect of atmospheric dispersion of
chemical/biological agents in military scenarios. Both the US Defense Threat Reduction Agency (DTRA) and UK
MoD have programs which are studying the development and integration of a number of Chemical and Biological
(CB) software models. Two related activities being undertaken by these organizations are the DMSO-sponsored
Environment Federation, which this year includes CB effects and the DTRA sponsored Weapon Effects FOM
(WEFOM) group which is considering the wider use of CB models across US DoD.
The focus of this paper is to discuss how to make best use of DoD's High-Level Architecture (HLA) to achieve the
integration of atmospheric dispersion models, like the UK's UDM and US SCIPUFF model into HLA federations. This
integration involves taking inputs from source release model , e.g. the DTRA sponsored WALTS models, weather data
from meteorological servers, e.g. OASES, and finally feeding outputs into sensor or combat models like JointSAF.
However, this integration process demands that careful thought be given to the protocols by which physical
information about the dispersants is exchanged between the different types of models. This paper is intended to
stimulate discussion of these issues, particularly with a view to ensuring fair consideration of both the physics of the
dispersion process and the information conceptually required to be exchanged between various simulators.
The Environment federation has taken one approach to the integration of dispersion models into an entity level
simulation model, using JointSAF as the ultimate consumer of dispersion information. During the WEFOM meetings
different approaches for describing the transportation of the dispersion were produced which were formulated into
several BOMs[jc2]. There are also differences between the kind of data subscribing models require to visualize
dispersion and those that wish to stimulate detailed sensor models. The trade offs between the different approaches are
discussed in the paper, together with details of the practical experimentation undertaken to participate in the
EnviroFed efforts, and DTRA’s involvement in the
1. Introduction EnviroFed III project is described in Section IV. In
participating in the EnviroFed III project, DTRA
provided expertise in developing a realistic WMD
Operating in its role as the DoD focal point for scenario, as well as the IMPACT and Weapons
weapons of mass destruction (WMD), The Defense Analysis and Lethality Tool Set (WALTS) simulations.
Threat Reduction Agency (DTRA) is the owner and IMPACT, which simulates hazard flow in both open
custodian of a variety of physics-based codes and urban terrains, is described in Section II. The
describing WMD-related phenomena. The use of these WALTS simulation, which simulates the effects of
codes, which have been developed internally or as part weapons on pre-defined targets, is also described in
of collaborative programs with other DoD and NATO section II. The paper concludes with a brief conclusion
organizations, has traditionally been limited to the that includes future directions.
direct support of the warfighting CINCs for the purpose
of, for example, counter-terrorism response, planning
strikes on WMD-related targets, estimation of collateral 2. Simulations
effects, and weaponeering. However, spurred by the
maturation of these codes and the prevalence of WMD- 2.1 IMPACT
related matters in current DoD and domestic response
planning, DTRA is leading an effort to make these The Integrated Modeling Platform for Advanced
codes available for re-use. This effort is expected to Computational Technologies (IMPACT) [MB1] is
make validated, realistic WMD environments available being produced by the Defense Science and
across the gamut of DoD simulation domains, and help Technology Laboratory (Dstl) for the United Kingdom
define requirements for future Chemical and Biological Ministry of Defense (MoD). IMPACT is being
Defense M&S capabilities. developed as the CB modeling calculation engine for
the UK MoD’s Nuclear Chemical and Biological
This paper describes a snapshot of DTRA’s on-going Battlefield Information System Application (NBC
efforts toward the goal of WMD simulation re-use. BISA), which will be the future warning and reporting
The term ‘weapons of mass destruction’ generally system for the UK armed forces. An IMPACT
refers to chemical, biological, nuclear, or radiological prototype has been developed as a proof-of-concept
weapons, or weapons involving high explosives. It demonstrator to show that the system can incorporate
therefore follows logically that in order to simulate state-of-the-art physics-based models and still provide
WMD effects, it is necessary to simulate the natural real-time CB dispersion estimates. An example
environment, and the changes to the environment that scenario modeled by IMPACT and visualized in a
are induced by these weapons. Thus, an important part separate application is shown in figure 1.
of this effort has been to develop a reference Federation
Object Model (FOM), called the Weapons Effects
FOM (WEFOM, described in Section III), that seeks to
standardize the representation of the WMD-related
environment. DTRA’s work in this area has followed
that of the DMSO-led EnviroFed project, which has
utilized the SEDRIS Environmental Data Coding
Standard (EDCS) in order to provide a common
terminology set for describing the environment. This
has led to a natural synergy between DTRA and the
Software Agent base class allows IMPACT to take
advantage of the additional power of multi-processor
hardware configurations where possible. Since all
updates to the Dynamic Object Database are recorded
to file, the system can restore its previous state easily
after a system crash, and resume modeling. A rollback
facility is also included, allowing the system to be
returned to its state at any specified time.
IMPACT incorporates the IMPACT Meteorological
Preprocessor (IMP), which provides the local scale
meteorological and wind field for the dispersion
modeling. IMP can ingest gridded mesoscale national
weather prediction (NWP) data or live weather station
data or a combination of the two. IMP incorporates
algorithms for calculating atmospheric boundary layer
Figure 1: IMPACT Dispersion Visualization parameters required by the dispersion models. These
are an implementation of the Environment Protection
The IMPACT architectural model grew out of the Agency (EPA) sponsored AERMET algorithms. IMP
requirement to integrate diverse collections of existing also includes the Flow Across Complex Terrain Solver
modeling components together to create modeling and (FACTS) [IHG2], a fast wind flow model that is an
simulation applications, using a generic object-oriented implementation of the Danish LINCOM model [IHG3].
architecture. The internal architecture chosen was The model solves linearised forms of the fundamental
based on the Blackboard Architecture, which is used in governing equations for turbulent flows in the planetary
certain types of data fusion applications. In boundary layer. It conserves both mass and momentum
conventional blackboard architectures, several software and in a matter of seconds it can predict a wind field
“agents” (autonomous modules) collaborate to solve a that incorporates the effects of terrain, morphology
problem by reading information from a “blackboard” (land usage) and atmospheric thermal conditions. The
database, and writing their conclusions back to the two main components of IMP are able to satisfy the
database. In the IMPACT architecture, a set of meteorological requirements of the dispersion models
modeling and gateway connectivity Software Agents within IMPACT.
collaborate to model dispersion, assess downwind
hazard and exchange information with other systems The IMPACT architecture is versatile enough to enable
(via HLA or otherwise). The software agents function the inclusion of a selection of different dispersion
autonomously, exchanging information via the models. The current IMPACT can include DTRA’s
Dynamic Object Database. Figure 2 provides a high- Hazard Prediction Assessment Capability (HPAC)
level overview of the IMPACT architecture. Second-order Closure Integrated Puff model
(SCIPUFF) [IHG4], Dstl’s Urban Dispersion Model
Software Agent (UDM) [IHG5] or an integrated version of the two
dispersion models. (This integrated model was
previously called CUSP, the Coupled UDM and
Impact Dispersion Mean and Variance Concentration Hazard Assessment
SCIPUFF, but this has recently been superseded by an
and HLA Gateway
urban version of HPAC that incorporates the UDM.)
Dynamic Object Database
SCIPUFF is a Gaussian puff model capable of
Time Series Source
providing estimates of both the concentration mean and
Wind Flow Output
Realisations Locations variance and probabilistic estimates that include the
uncertainties of meteorology and small-scale
Figure 2: IMPACT Architecture Overview atmospheric turbulence. The model can provide
estimates of dosage and deposition as well as the
The architecture is highly modular, allowing software concentration field.
agents to be developed independently and tested in
isolation, before being integrated together with relative The UDM is a Gaussian puff model designed to handle
ease. The support of multi-threading within the dispersion within urban environments. It is a semi-
empirical model based on a large number of urban
experiments conducted in a boundary layer wind including the rapid fluctuations and intermittency
tunnel. The model is computationally very fast, capable resulting from the turbulence in the wind field. At
of providing dispersion estimates of releases in urban present, a Markov Chain Monte Carlo approach is
areas consisting of tens of thousands of buildings utilized, but current work involves the use of a spectral
significantly faster than real time. It provides estimates synthesis approach, which will allow sets of time series
of the concentration and dosage fields and can also with physically realistic spatial and temporal
estimate the concentration within buildings through an correlation to be generated.
integrated building infiltration model. Figure 3 shows
the results from an example release in an urban area 7.00E-04
modeled by the UDM. The UDM is currently the 6.00E-04
subject of a major validation effort (e.g. [IHG6]) and it
Concentration (kg m-3)
was used together with SCIPUFF as part of DTRA’s 4.00E-04
research support for the 2001 US Presidential
-145 -130 -115 -100 -85 -70 -55 -40 -25 -10
Relative Tim e (s)
Figure 4: Mean Concentration Time Series
Concentration (Kg m-3)
-60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0
Figure 3: UDM Dispersion Output for Urban Area Relative Time (s)
IMPACT also includes functionality to utilize the
dispersion model output to calculate concentration time Figure 5: Concentration Realization Time Series
series for sample locations within the simulation space.
These sample points may be fixed or moving. A two An effects contouring module has been included,
stage approach is utilized: primarily to provide a simple data construct for other
federates to visualize. The effects contouring module
In the first phase, discrete time series for ensemble first makes a rapid assessment of the spatial region
mean concentration and concentration variance are currently affected by the dispersing plume. It then
calculated for each sample location. The gaussian puff calculates a two-dimensional grid of concentration
output data from the dispersion models is utilized for values over the affected region. Finally, an appropriate
this calculation. An output sample interval of 5 seconds concentration threshold is applied, and the shape of the
is usually chosen. The output data is stored in the region experiencing concentrations at this or higher
Dynamic Object Database, ready to be passed to the levels is converted to a vector representation, then
HLA federation. passed to the HLA federation.
In the second optional phase, the discrete concentration The HLA gateway is implemented as a Software Agent,
mean and variance time series from step 1 are utilized whose responsibility is to communicate with the HLA
to synthesize time series “realizations”, with smaller federation, to add releases, meteorology and sample
sample rates, typically 5Hz. These time series locations notified by the federation to the IMPACT
realizations are designed to resemble typical Dynamic Object Database, and to report any output
concentration time series observed in field experiments, from the dispersion modeling and assessment modules
which appears in the Dynamic Object Database back to
the federation. In order to facilitate the application of DTRA’s core
weapon effects and WMD environment simulation
The HLA gateway uses a configuration file to control capability into a wide variety of application areas,
interactions with other federates and includes the DTRA initiated the development of a reference
concept of implementations, together these allow Weapons Effects Federation Object Model (WEFOM).
IMPACT to be quickly tailored to the requirements of a
particular federation. DTRA assembled a large WEFOM development team.
Representatives from DTRA proper, Defense Modeling
2.2 WALTS and Simulation Office; Simulation, Training, &
Instrumentation Command; Naval Surface Warfare
The Weapon Analysis and Lethality Tool Set Center, Ballistic Missile Defense Organization,
(WALTS) is a Defense Threat Reduction Agency Dugway Proving Grounds, Sandia National Laboratory,
(DTRA) simulation to model weapons effects. This John Hopkins University/Applied Physics Laboratory,
HLA compliant simulation allows credible physics- Institute for Defense Analysis; Defense Science and
based weapons effects for virtual targets to be Technology Laboratory (UK), to name but a few; along
distributed among an HLA Federation or offline with a host of support contractors, own this effort.
analysts. Some of the key software components of Their needs span the spectrum from use in operations,
WALTS are the munitions definitions, physics engine training, analysis, experimentation, and acquisition
calculations, tools which allow CB contaminants to be environments, and from the platform / single entity
integrated into the target, and the visualization of data level through fully aggregated simulations.
relevant to the damaged target.
The structure of the Weapon Effects Federation is
WALTS munitions database includes over 50 of the much like that of the Environment Federation. Both
most common JMEMS structural munition definitions. federations have terrain and environment servers that
Among the physics engine calculations currently in use simulate the dynamic changes to the terrain and
are cratering, penetration, internal blast, first order environment, and then publish this information to the
fragmentation, external blast and contaminant other federates. In addition, both federations utilize the
expulsion. Configuration tools for the virtual target Synthetic Environment Data Representation
allow the user to “place” CB contaminants in the 3D Interchange Specification (SEDRIS) standard to foster
model on a per room basis. Any resulting internal blast FOM agility. The WEFOM currently uses only the
code run on this model will then determine if an agent SEDRIS Environmental Data Coding Standard (EDCS)
release occurs. Based on the calculated effects, a because it offers a standardized code for specifying
damaged model is dynamically generated for 3D terrain and environmental objects, attributes,
visualization. In addition to this 3D model, pressure vs. interactions, and parameters. The other pieces of the
time and temperature vs. time information is also SEDRIS specification, such as, the Data Representation
available to the other federates or the analyst. Model (DRM) or the Spatial Reference Model (SRM),
are not being used at this time.
3. The Weapon Effects FOM
The actual construction of the WEFOM is based on the
3.1 WEFOM Introduction Base Object Model (BOM) concept. The idea behind
the BOM concept is to produce object model
As stated earlier in the paper, the Defense Threat components that are reusable. This enables FOM and
Reduction Agency was chartered to focus and integrate SOM developers to leverage pre-built simulation OM
the DoD’s capabilities to combat the emerging components for rapid development. Typically, BOM
Weapons of Mass Destruction threat. In addition to components consist of object classes, interaction
developing the high fidelity, physics based codes, classes, and any associated attributes, parameters, and
necessary to model the phenomena; DTRA has been meta-data. The BOM concept is currently being
actively adapting these codes for use by other studied within the Simulation Interoperability
simulations within DoD’s High Level Architecture Standards Organization (SISO).
environment. Within this simulation world, DTRA’s
goal was to match their weapons effects modeling and DTRA’s participation in last years EnviroFed III
simulation capabilities with the end users needs, demonstration was the first test of the WEFOM
emphasizing weapons of mass destruction. development effort. That version of the WEFOM
contained but a small subset of all the available BOMs
from DSTL (UK), DTRA, and OASES. The success of In reality, releases can take a variety of forms, e.g.
that initial demonstration confirmed the development instantaneous, constant, atmospheric or ground based.
path taken by the WEFOM team. It also highlighted In producing the BOM the decision was made to have a
areas where more work is needed. single interaction which could be specialized to
represent all forms of release rather than a hierarchy of
Future efforts will focus on the development of new release interactions which could potentially lead to
BOMs and improving upon those already developed, dangerous ambiguities.
assisting with the SEDRIS development effort to ensure
that it is capable of adequately supporting rapid, 3.2.2 Sensor BOMs
dynamic terrain changes, providing input to SEDRIS
EDCS to allow for a fuller representation of the hazard Sensor BOMs provide federates with the capability to
environment, optimizing the hazard geometry for monitor the effects of the dispersion cloud and receive
improved publication/subscription, and scalability reports on contamination and dosage. Subscribing
issues. federates can then use these data to drive damage
models or gauge performance degradation for entities
3.2 Chemical Biological Dispersion in WEFOM exposed to the gas cloud.
As part of our work in WEFOM a number of Base One central problem with developing a FOM or BOM
Object Models have been developed to provide HLA for gas dispersion is how to represent the physical
federates with the capabilities of modeling chemical or characteristics of a gas cloud in HLA terms. While the
biological warfare in a synthetic environment. HLA is very good at capturing discrete values such as
vehicle locations and velocities, it is less well suited to
The BOMs were derived from a gas dispersion FOM modeling the characteristics of an evolving dispersion
[jc1] developed to originally connect IMPACT to other plume. Arguably gas dispersion is best represented
simulations. BOMs were developed in accordance with mathematically as continuous fields whose description
the guidelines presented in [jc3] and developed using a in discrete variables is neither well-defined nor
set of in house tools. necessarily independent of the evolution of that field.
The BOMs have been categorized into three separate In developing these BOMs it was decided that the
areas: 1) To allow the description of the characteristics simulation modeling gas dispersion should not publish
of a chemical or biological agent release into the its own internal representation of the gas cloud, as it is
environment. 2) Allowing federates such as Stealth principally an internal design decision that determines
Viewers to visualize the resulting gas plume 3) how the simulation should translate this discrete
Providing a mechanism for federates to measure representation into the properties of the gas at a
contamination or dosage levels at particular points in particular point of interest.
the virtual space.
To this end a number of sensor BOMs are defined
In the following sections we describe these different which correspond broadly to the types of sensor found
categories of BOMs in more detail. in operational usage, such as a point or grid based
sensors, or stand off sensors, such as a laser based
3.2.1 Agent Release BOM system. In this hierarchy of sensors entities such as
personnel or vehicles can be considered as a
The Agent Release BOM defines an interaction and set specialized case of point sensor. The Visualization
of Complex Types to capture the properties of a BOM, discussed below, also corresponds to the sensor
dispersant agent as it enters the atmosphere. model. Federates that wish to visualize the cloud
produce a pseudo-sensor which is updated with puff
The interaction allows a federate to specify a variety of data as the cloud evolves.
factors such as the location of the release, the type of
agent involved, the mass, release dynamics as well as The various types of sensor are modeled as a hierarchy
particle sizes. with Complex Types that allow the characteristics of a
particular sensor, for instance sensitivity to particle
An Agent Release interaction is used as a cue to size, to be specified. It should be noted that the sensor
IMPACT to start modeling the dispersion of the agent. BOMs do not attempt to model real world sensor
systems, the readings returned to subscribing federates
are “perfect” in so far that they represent an insight into EnviroFed III consisted of a number of excursions
what the dispersion model determines to be the status centered around an assault on Camp Pendleton in San
of the cloud. Federates that wish to model the Diego County. The scenario envisaged a sea assault by
characteristics of real world sensors can subscribe to blue forces against red forces defending the airfield.
these data and then manipulate the data to model The various excursions allow the impact of
probability of detection and so forth. environmental changes to be assessed on mission
3.2.3 Visualization BOMs
In one particular excursion a JSAF F-18 identified a
Most dispersion plumes resulting from a CB release are bunker close to the airfield as a target of opportunity.
invisible to the human eye but in the realm of synthetic Unknown to blue forces the bunker contained a
environments there is often a need to visualize the chemical warfare agent. As the excursion progressed,
cloud to determine ground truth and drive Stealth an incoming round penetrated the storage bunker,
displays. causing expulsion of the chemical agent. The data
flows between the various federates in simulating this
The Visualization BOM defines a set of Object Classes excursion is described below.
and Complex Types to allow geometrical
representations of the gas cloud to be transmitted to IMPACT’s federate interactions are summarized in
subscribing federates. figure 6:
The visualization is represented as a set of Gaussian
puffs heuristically derived from the simulation’s OASES
internal representation, which may typically be a puff- Data
based dispersal model. Each published puff can be
associated with a set of bounding planes to amputate Dosage Reports
parts of the puff; this functionality allows us to model JSAF Concentration IMPACT
not only puffs but also the dispersant wakes that can Contour
persist in the lee of buildings.
Target WALTS Release
4. Participation in EnviroFed III Impact
Part of DMSO’s Integrated Natural Environment
Program, the Environment Federation (EnviroFed) is a Figure 6: IMPACT Federate Interactions
project in the Environmental Integration/Experiments
Technology Area. From its start in 1999 EnviroFed has JSAF is responsible for modeling the aircraft and
had a number of key aims[jc5]: reporting the weapon detonation of a 2000 pound bomb
on the top of the storage bunker.
• Demonstrate the state-of-the-art with regard to the
This MunitionDetonation interaction is mapped by the
representation of the natural environment in DoD
Agile FOM Interface (AFI) to a WeaponTargetImpact
interaction and is received by WALTS which begins to
• Demonstrate the application of SEDRIS products
model the effects of the detonation on the structure of
and HLA tools and processes to combat
the bunker. Upon receipt of this interaction WALTS
simulations and environment simulations
begins to model the effects of the penetration and blast
• Facilitate reuse of resulting environmental
on the bunker. WALTS takes into account the type of
munition, the fuse setting, the impact coordinates and
velocity vector of the weapon that has detonated to
IMPACT recently participated in the EnviroFed III
calculate the damage caused to the building. Figure 7
exercise which was helpful in the development and
below shows a 3D model of the damage inflicted on the
refinement of the BOMs discussed above. As a member
building by the detonation calculated by WALTS. This
of this federation IMPACT interacted with WALTS,
model was dynamically generated by WALTS as a
OASES[jc4] and JSAF to augment the environmental
result of the calculated damage.
model with chemical and biological effects during
several scenario excursions.
[jc2] P. L. Gustavson, J. P. Hancock, M.
McAuliffe: “Base Object Models (BOMs):
Reusable Component Objects for Federation
Development” Paper 98F-SIW-034 Simulation
Interoperability Workshop Fall 1998.
[MB1] M. Bull: “Impact Design Study” Dstl Internal
Document, 18 December 2000.
[IHG2] F. Dunkerley: “FACTS LINCOM Technical
Documentation (Draft)” Risoe National Laboratory
Figure 7: Collateral Damage to the Bunker report prepared for Dstl, May 2001.
[IHG3] P. Astrup, N. O. Jensen, T. Mikkelsen:
The WALTS physics engine internal blast code models “Surface Roughness Model for LINCOM” Risoe
damage to the bunker and determines that the CB National Laboratory, 1996.
contaminants stored inside would have been released [IHG4] R. I. Sykes et al: “PC-SCIPUFF Version 1.0
into the atmosphere. WALTS communicates this to Technical Documentation” Titan Corporation
IMPACT using the AgentRelease BOM, detailing the report prepared for DTRA, 1997.
type of agent released, for example, VX nerve gas, and [IHG5] D. J. Hall, A. M. Spanton, I. H. Griffiths, M.
environmental factors such as the velocity of the Hargrave, S. Walker: “The UDM – A Model for
release. Estimating Dispersion in Urban Areas” Envirobods
Ltd report prepared for Dstl, August 2000.
On receipt of the AgentRelease interaction IMPACT [IHG6] I. H. Griffiths, N. V. Beck, C. John, D. J. Hall,
begins modeling the evolution of the gas plume A. M. Spanton: “The Results of an Initial
resulting from the release. Factors such as Validation Study of an Urban Dispersion Model”
meteorological conditions, terrain type and interactions American Met. Soc. Third Symposium on the
with surrounding buildings are all used as inputs into Urban Environment, Davis, CA, August 2000.
the modeling process. [IHG7] J. Pace, I. H. Griffiths, R. I. Sykes, M.
Phillips, J. Hodge, K. Kim: “Capabilities of DTRA
The meteorological data subscribed to by IMPACT Urban Modeling during Presidential Inauguration
includes wind speed and direction as well as data Support” 5th Annual George Mason University
regarding cloud coverage and ground surface Transport and Dispersion Modeling Workshop,
temperature. These data are processed by IMPACT’s July 2001.
meteorological model to build a description of the wind [jc1] R. Penney, P. Hoare: “Towards an HLA
conditions over the terrain and transports the plume Federation Object Model for Atmospheric
downwind. Dispersion Modeling” Internal QinetiQ report,
IMPACT subscribes to the locations of platform [jc3] Base Object Model (BOM) Study Group:
entities controlled by JSAF. As the plume evolves “BOM Methodology Strawman (BMS)
dosages are calculated and reported back for any entity Specification Version 0.7”
that is exposed to the hazard. This data is used by JSAF http://www.sisostds.org/doclib/doclib.cfm
to drive damage models and degrade the performance ?SISO_RID_1002220
of units exposed or in MOPP gear. [jc5] R. Lutz et al.: “Envirofed Phase III
Demonstration Slides” November 8th 2001
JSAF can also request a contamination contour from [jc4] R. A. Reynolds, H. Iskenderian, S. O. Ouzts:
IMPACT, essentially a footprint of the plume, for “The Ocean, Atmosphere and Space
display on the PVD. This capability was not provided Environmental Services (OASES) System” Paper
in the original Gas Dispersion BOMs but added to 01S-SIW-047 Simulation Interoperability
support EnviroFed. Workshop Spring 2001.
IAN GRIFFITHS has worked as a scientist at Dstl
Chemical and Biological Sciences for six years. He led
the research teams that developed the UDM and to joining DTRA, Mr. Zimmers served 23 years in the
FACTS. He now heads the team that is producing US Army as a helicopter pilot. He flew numerous
IMPACT as well as teams integrating the UDM with combat and support missions in the Vietnam conflict
HPAC, producing operational hazard assessment where he was awarded a Bronze Star, Purple Heart, and
systems and researching wind flow modeling 19 Air Medals.
particularly, most recently, at the land-sea interface.
RICHARD PENNEY is a Senior Mathematician at
MARTYN BULL is a senior software consultant at QinetiQ Malvern UK, and has research interests that
RiskAware Ltd. Through his consultancy work with cover mathematical modeling, signal processing and
Dstl Chemical and Biological Sciences, he has been automatic routing. He studied physics at Oxford
involved with the development and integration of University, and completed a DPhil there in theoretical
mathematical models for the CB theatre for five years. physics on spin-glasses and neural networks. Since
He was the software architect for the IMPACT joining QinetiQ (formerly DERA) in 1995, he has
prototype. worked on problems encompassing synthetic
environments, submarine sonar-evasion, radar signal
PETER HOARE is a Principal Scientist in the processing, and gas dispersion.
Distributed Technology Group at QinetiQ Malvern. He
received a BSc(Hons) and PhD degrees from Royal DONALD WARF is a Software Engineer with Cubic
Holloway, University of London in 1988 and 1993 Defense Systems, Inc., Alexandria, VA, providing
respectively. He has been working in parallel and support to DTRA. He has been involved in the
distributed simulation since 1993 and is the technical integration of WALTS and IMPACT into the Envirofed
lead of several programmes in the UK researching the Federation.
use of the DoD High Level Architecture. Also was also
a Drafting member of the HLA Standards Development TOM S. STARK is a Senior Scientist at Defense
group developing the IEEE 1516 standards documents. Group Inc. in Alexandria, VA. In this position, Dr.
Stark provides modeling and simulation, operations
JOHN CARSON is a scientist at QinetiQ Malvern. research, and information technology support to the
Since joining QinetiQ (formerly DERA) in 1999 he has s
Defense Threat Reduction Agency' Technology
worked on Command Information Systems and the Development and Chemical and Biological Defense
HLA Interface to IMPACT. He has a PhD in parallel Directorates. Prior to working for Defense Group Inc.,
architectures and languages and has worked in the field Dr. Stark worked for Science Applications
of Synthetic Environments and prototyping underwater International Corp. as part of the DMSO support team.
vehicles via simulation. Dr. Stark holds Bachelors, Masters, and Doctoral
degrees in Physics, with a specialization in lasers and
WALTER ZIMMERS is Chief of DTRA’s Weapons quantum electronics.
of Mass Destruction Assessment and Analysis Center,
Counterproliferation Support and Operations
Directorate. He has served in several operational
support positions in his 16 years at DTRA and
predecessor organizations, the Defense Special
Weapons Agency and Defense Nuclear Agency. Prior