DTIC ADA545199: SCIPUFF Tangent-Linear/Adjoint Model for Release Source Location from Observational Data

1.2 MB·English

by  Defense Technical Information Center

REPORT DOCUMENTATION PAGE Form Approved OMB NO. 0704-0188
The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions,
searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments
regarding this burden estimate or any other aspect of this collection of information, including suggesstions for reducing this burden, to Washington
Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA, 22202-4302.
Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any oenalty for failing to comply with a collection of
information if it does not display a currently valid OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
18-01-2011 Final Report 8-Sep-2006 - 7-Sep-2010
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
SCIPUFF Tangent-Linear/Adjoint Model for Release Source
Location from Observational Data
5b. GRANT NUMBER
W911NF-06-C-0161
5c. PROGRAM ELEMENT NUMBER
206023
6. AUTHORS 5d. PROJECT NUMBER
O. O. (“Luwi”) Oluwole, S.E. Albo, Richard C. Miake-Lye
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAMES AND ADDRESSES 8. PERFORMING ORGANIZATION REPORT
NUMBER
Aerodyne Research, Inc.
45 Manning Road
Billerica, MA 01821 -3976
9. SPONSORING/MONITORING AGENCY NAME(S) AND 10. SPONSOR/MONITOR'S ACRONYM(S)
ADDRESS(ES) ARO
U.S. Army Research Office 11. SPONSOR/MONITOR'S REPORT
P.O. Box 12211 NUMBER(S)
Research Triangle Park, NC 27709-2211 48892-CH-CDP.1
12. DISTRIBUTION AVAILIBILITY STATEMENT
Approved for Public Release; Distribution Unlimited
13. SUPPLEMENTARY NOTES
The views, opinions and/or findings contained in this report are those of the author(s) and should not contrued as an official Department
of the Army position, policy or decision, unless so designated by other documentation.
14. ABSTRACT
The threat of atmospheric contamination by hazardous materials remains a high national security concern. There is
a strong need for the development of emerging technologies which can significantly advance risk assessment and
response capabilities.
In this project, Aerodyne Research Inc. (ARI) has developed and demonstrated an
15. SUBJECT TERMS
Release Source Location, SCIPUFF, Inverse modeling, Threat detection, Source term estimation, Atmospheric dispersion modeling
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 15. NUMBER 19a. NAME OF RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE ABSTRACT OF PAGES Richard Miake-Lye
UU UU UU UU 19b. TELEPHONE NUMBER
978-663-9500
Standard Form 298 (Rev 8/98)
Prescribed by ANSI Std. Z39.18
Report Title
SCIPUFF Tangent-Linear/Adjoint Model for Release Source Location from Observational Data
ABSTRACT
The threat of atmospheric contamination by hazardous materials remains a high national security concern. There is a strong need for the
development of emerging technologies which can significantly advance risk assessment and response capabilities.
In this project, Aerodyne Research Inc. (ARI) has developed and demonstrated an
algorithm for source estimation, called AIMS (“Aerodyne Inverse Modeling System”). AIMS takes as input all available observational data
and optionally any prior knowledge of the source parameters. The output is the set of source parameters that best describes the observations,
including number of sources, emission rates, locations, start and end times. AIMS is also designed to include an a posteriori assessment of
its solution quality, providing useful feedback on how much confidence to put in a particular solution and in what ways the solution quality
might be improved. A novel feature in AIMS is the ability to integrate multiple observation types in order to maximize information content
for source estimation. This capability has been demonstrated for datasets from stationary and mobile sensors.
AIMS provides a significantly improved capability to protect the warfighter and civilians through successful identification and
quantification of unknown hazardous atmospheric releases.
List of papers submitted or published that acknowledge ARO support during this reporting
period. List the papers, including journal references, in the following categories:
(a) Papers published in peer-reviewed journals (N/A for none)
Number of Papers published in peer-reviewed journals: 0.00
(b) Papers published in non-peer-reviewed journals or in conference proceedings (N/A for none)
Number of Papers published in non peer-reviewed journals: 0.00
(c) Presentations
1. SCIPUFF Tangent-Linear Adjoint Model for Release Source Location from Observational Data. Luwi O. Oluwole and Richard C.
Miake-Lye, Chem-Bio Information Systems conference (Data Assimilation and Tactical Applications), 10 January 2007; Austin, TX
2. Source Estimation using SCIPUFF Tangent-Linear Adjoint Model, Luwi O. Oluwole and Richard C. Miake-Lye, 11th Annual GMU
Atmospheric Transport and Dispersion modeling conference, 10 July 2007; Fairfax, VA
3. SCIPUFF Tangent-Linear and Adjoint Models for Release Source Estimation from Observational Data, Simón E. Albo, Luwi O.
Oluwole
and Richard C. Miake-Lye, 12th Annual GMU Atmospheric Transport and Dispersion modeling conference, 8 July 2008; Fairfax, VA
4. SCIPUFF-TLA : A plume inversion tool for identifying and quantifying pollutant emission sources from multi-measurement data,
Oluwayemisi O. (“Luwi”) Oluwole, Simón E. Albo, Richard C. Miake-Lye, 13th GMU Conference, July 15 2009; Fairfax, VA
5. New developments in SCIPUFF Tangent-Linear Adjoint, Simón E. Albo, Oluwayemisi O. (“Luwi”) Oluwole and Richard C.
Miake-Lye, 13th Annual GMU Atmospheric Transport and Dispersion modeling conference, 14 July 2009; Fairfax, VA
Number of Presentations: 5.00
Non Peer-Reviewed Conference Proceeding publications (other than abstracts):
Oluwole, O.O., S.E. Abo and R. C. Miake-Lye, 2008; “Source Estimation Using SCIPUFF Tangent-Linear or Adjoint,” CBD Physical
Science and Technology conference proceedings.
Number of Non Peer-Reviewed Conference Proceeding publications (other than abstracts): 1
Peer-Reviewed Conference Proceeding publications (other than abstracts):
Number of Peer-Reviewed Conference Proceeding publications (other than abstracts): 0
(d) Manuscripts
Number of Manuscripts: 0.00
Patents Submitted
Patents Awarded
Awards
Graduate Students
NAME PERCENT_SUPPORTED
FTE Equivalent:
Total Number:
Names of Post Doctorates
NAME PERCENT_SUPPORTED
FTE Equivalent:
Total Number:
Names of Faculty Supported
NAME PERCENT_SUPPORTED
FTE Equivalent:
Total Number:
Names of Under Graduate students supported
NAME PERCENT_SUPPORTED
FTE Equivalent:
Total Number:
Student Metrics
This section only applies to graduating undergraduates supported by this agreement in this reporting period
The number of undergraduates funded by this agreement who graduated during this period:...... 0.00
The number of undergraduates funded by this agreement who graduated during this period with a degree in
science, mathematics, engineering, or technology fields:...... 0.00
The number of undergraduates funded by your agreement who graduated during this period and will continue
to pursue a graduate or Ph.D. degree in science, mathematics, engineering, or technology fields:...... 0.00
Number of graduating undergraduates who achieved a 3.5 GPA to 4.0 (4.0 max scale):...... 0.00
Number of graduating undergraduates funded by a DoD funded Center of Excellence grant for
Education, Research and Engineering:...... 0.00
The number of undergraduates funded by your agreement who graduated during this period and intend to
work for the Department of Defense...... 0.00
The number of undergraduates funded by your agreement who graduated during this period and will receive
scholarships or fellowships for further studies in science, mathematics, engineering or technology fields:...... 0.00
Names of Personnel receiving masters degrees
NAME
Total Number:
Names of personnel receiving PHDs
NAME
Total Number:
Names of other research staff
NAME PERCENT_SUPPORTED
FTE Equivalent:
Total Number:
Sub Contractors (DD882)
1 a. Sage Management 1 b. 6731 Columbia Gateway Drive
Suite 150
Columbia MD 21046
Sub Contractor Numbers (c):
Patent Clause Number (d-1):
Patent Date (d-2):
Work Description (e): Developers of SCIPUFF. Provided technical assistance for implementation of inverse modeling techniques in SCIPUFF.
Sub Contract Award Date (f-1): 10/1/2006 12:00:00AM
Sub Contract Est Completion Date(f-2): 9/7/2009 12:00:00AM
Inventions (DD882)

SCIPUFF Tangent-Linear/Adjoint Model for Release
Source Location from Observational Data

Final Report
ARO Contract No. W911NF-06-C-0161

September 8 2006 – September 7 2010

Prepared by

O. O. (“Luwi”) Oluwole, S.E. Albo, Richard C. Miake-Lye1
Aerodyne Research, Inc.
45 Manning Road, Billerica MA 01821-397

Submitted to

US Army Research Office
PO Box 12211
Research Triangle Park, NC 27709-2211

1 Principal Investigator; E-mail: [email protected]; Phone: [phone] x251
Aerodyne Research, Inc. W911NF-06-C-0161

THIS PAGE INTENTIONALLY BLANK
2
1/17/2011 Aerodyne Research, Inc. W911NF-06-C-0161
Table of Contents

Project Objective.................................................................................................................3
Background.........................................................................................................................4
SCIPUFF Atmospheric Transport and Dispersion Model..............................................5
Prior Development..........................................................................................................6
Benefit to Warfighter..................................................................................................8
Maturity of the Technology Prior to this Project........................................................8
Project Accomplishments: The Aerodyne Inverse Modeling System (AIMS)..................9
Overview.........................................................................................................................9
The Cost Function...........................................................................................................9
Automatic Differentiation.............................................................................................11
“Forward” or “Tangent-Linear” Sensitivity..............................................................12
“Reverse” or “Adjoint” Sensitivity...........................................................................12
Source Estimation Algorithm Heuristics......................................................................14
Quantifying Source Estimate Uncertainty....................................................................15
1. Solution Quality Index......................................................................................15
2. Data Information Content Index.......................................................................16
Modifications to SCIPUFF: Modeling Continuous Releases.......................................17
Source Location Model Testing and Validation...............................................................19
Application to Ideal (model-generated) Data...............................................................19
AIMS Application to Field Data...................................................................................28
Summary...........................................................................................................................33
Recommendations for Future Research & Development.................................................34
References.........................................................................................................................35

Project Objective
The threat of atmospheric contamination by hazardous materials remains a high national
security concern. There is a strong need for the development of emerging technologies
which can significantly advance risk assessment and response capabilities. Thus, the
objective of this project was the development and validation of SCIPUFF tangent-linear
or adjoint model as a counterproliferation and counterforce tool for characterizing the
source of a nuclear, biological, chemical (NBC) release using observational data. A
fundamental aspect of gradient-based inverse modeling methods are the sensitivities of
the ‘mismatch’ between numerical simulation and measurement for observables of a
hazardous release to model parameters and inputs. Thus, another objective of this
research project was to use the SCIPUFF tangent-linear or adjoint model to further the
current understanding of the relationships between field observations, the characteristics
of the release source, and the transport and dispersion of the hazardous cloud.

In this project, Aerodyne Research Inc. (ARI) has developed and demonstrated an
algorithm for source estimation, called AIMS (“Aerodyne Inverse Modeling System”).
AIMS takes as input all available observational data and optionally any prior knowledge
of the source parameters. The output is the set of source parameters that best describes
3
1/17/2011 Aerodyne Research, Inc. W911NF-06-C-0161
the observations, including number of sources, emission rates, locations, start and end
times. AIMS is also designed to include an a posteriori assessment of its solution quality,
providing useful feedback on how much confidence to put in a particular solution and in
what ways the solution quality might be improved. A novel feature in AIMS is the
ability to integrate multiple observation types in order to maximize information content
for source estimation. This capability has been demonstrated for datasets from stationary
and mobile sensors.

AIMS provides a significantly improved capability to protect the warfighter and civilians
through successful identification and quantification of unknown hazardous atmospheric
releases.

Background
The standard modeling approach for the tracking of atmospheric plumes falls in the
category of forward modeling: given an initial state (such as the atmospheric state, an
initial tracer concentration), and, possibly time-dependent boundary values (such as
space-/time-dependent tracer emissions and, in case of an offline meteorological
background the time-evolving meteorological state), a transport model is stepped forward
in time to produce a field of tracer concentrations at subsequent times.

FFoorrwwaarrdd aattmmoosspphheerriicc
ttrraannssppoorrtt && ddiissppeerrssiioonn mmooddeell

PPrreeddiicctteedd
KKnnoowwnn ggaass ((SSCCIIPPUUFFFF))
ccoonncceennttrraattiioonn
ssoouurrcceess
ffiieelldd

SSoouurrccee llooccaattiioonnss,, MMeeaassuurreedd

mmaaggnniittuuddeess,, ccoonncceennttrraattiioonn
IInnvveerrssee mmooddeell
ttiimmeess?? bbaasseedd oonn SSCCIIPPUUFFFF ffiieelldd
((AAIIMMSS))
Figure 1. An illustration of atmospheric dispersion modeling in forward and inverse modes. In the
forward mode, the gas source characteristics are well-known and the goal is to predict concentrations
at a later time and downwind of the emission sources. In the inverse mode, downwind concentrations
are known from measurements and the goal is to determine the source characteristics.

In source estimation, one wishes to compare the time-evolved tracer concentrations with
a set of measurements of these quantities. This sort of problem falls in the category of
inverse modeling – the goal is to determine the initial and/or boundary values that lead to
a model trajectory most consistent with the observations. So, given the measurement
data, one wishes to identify and quantify the emission sources. Among other benefits,
inverse modeling enables more accurate assessment of the impacts of emissions events:
unknown inputs (e.g. emission rates/quantities) required for the forward modeling can be
automatically determined from measurement data using inverse modeling. Inverse
atmospheric dispersion modeling is an area of active research [Bowen and Stratton
4
1/17/2011 Aerodyne Research, Inc. W911NF-06-C-0161
(1981), Daescu and Carmichael (2003), Errico (1997), Long et al. (2010), Lushi and
Stockie (2010), Mavriplis (2007)].

The forward model of interest in this work is SCIPUFF (Second-Order Closure Integrated
PUFF, [Sykes et al. (1985)]), which is a state-of-the-art atmospheric transport and
dispersion model adopted by the Defense Threat Reduction Agency (DTRA). As
described in the next section, SCIPUFF is equipped to model a variety of emission
scenarios involving varying characteristics (locations, rates and times) and under wind
conditions that vary spatially and temporally.
SCIPUFF Atmospheric Transport and Dispersion Model
The atmospheric dispersion model is the Second-order Closure Integrated Puff
(SCIPUFF) model [Sykes et al., (1986, 1995, 1999)] which is the dispersion model
employed in the Hazard Prediction and Assessment Code Capability (HPAC) [SAIC,
2001]. While SCIPUFF initially treated gaseous puffs evolving in the atmosphere,
interactions with other materials are included in current versions. This includes solid
particles, liquid droplets, and nuclear material and the ability to treat evaporation,
deposition, and precipitation. Release sources include both continuous plumes and
instantaneous puffs. A single calculation allows for any combination of continuous and
instantaneous releases at any time throughout the temporal domain. Continuous releases
are modeled as a series of single puffs releases one after another
The concentration field is modeled using the Gaussian puff method [Bass (1980)]. The
concentration field is given by the summation of individual puffs, and the concentration
for each puff, at a given time, is
(1)

(cid:1)
In Equation 1, Q is the puff mass, x is the puff centroid, and σ is the puff spread. The
dynamic relations are derived by moment methods, where the conservation equations are
integration over all space. Turbulent diffusion is modeled by the second-order schemes of
Donaldson [1973] and Lewellen [1977] where the second order fluctuation terms are
solved for by a corresponding transport equation. The closure scheme provides a
statistical variance of the concentration field, and the variance provides probability
information. Thus, the SCIPUFF calculations provide mean concentration values together
with a corresponding uncertainty estimate. Practical applications of the model on 50 km
and 3000 km length scales have been performed [Sykes et al., (1986, 1995, 1999)].

Numerical techniques used include puff splitting, merging, and adaptive time steps.
Because the puff centroid evolves based on the mean flow velocity at the centroid
location, the Gaussian representation becomes less accurate as the puff grows. In order to
overcome this problem, if a puff reaches a critical width, the puff is split into smaller
puffs. The summation of small puffs can be used to accurately represent a generalized
concentration field. As the smaller puffs evolve, if the puffs have a region of considerable
overlap, then they are recombined or merged to a single Gaussian distribution. As
different puffs evolve, each is exposed to a different dynamic environment, and as such,
5