A Review of JSIMS Environmental Tailoring Services JETS with Focus on Sensor Impacts

1
A Review of JSIMS Environmental Tailoring
Services (JETS) with Focus on Sensor Impacts

• 15 September 1999
• Peter S. Dailey, Ph. D.
• psdailey_at_tasc.com

55 Walkers Brook Drive Reading, MA 01867
[phone]
2
Agenda

• SNE TAILORING PAST AND PRESENT
• JETS PFM ALGORITHM
• Correlation component
• Constraint component
• SENSOR IMPACTS DEMONSTRATION
• Feed SNE data through target acquisition sensor
  model
• Edit a tactically significant variable within the
  SNE
• Feed edited SNE data through sensor model
• Edit same variable using JETS PFM algorithm
• Compare results
• CONCULSIONS

3
SNE Tailoring Past and Present
4
JETS Vision
To provide JSIMS-like simulation designers,
controllers and participants an ability to
influence the exercise/training experience
through modification of the SNE
focusing on tools enabling new levels of realism
for simulations (exercises/training/rehearsals)

• Applications
• Provide flexible, more efficient tools for
  initial construction of environmental scenarios
  using authoritative data sources and tailoring
• Enable run-time tailoring of environmental data
  to steer the simulation at appropriate levels of
  fidelity

5
Versatile ApplicationAcross the Simulation
Spectrum
JETS algorithms can be applied to environmental
data at multiple resolutions
JETS algorithms are independent of data source
(gridded fields from any model or system)
DATA SOURCE
RESOLUTION

• ESG
• GlobalDB
• various models

• low resolution
• high resolution

REPRESENTATION

• Constructive
• -- to --
• Entity-based 3D Visualizations

SNE data
JETS algorithms work on base state fields and can
be applied to any level representation
atmospheric data
6
Importance of Tailoring

• Meteorology/Oceanography (METOC) especially
  difficult to maintain cross- and intra-domain
  correlation in space and time (JETS focuses on
  meteorology only)
• Correlation critical to realistic simulation of
  all sensor/intel assets in JSIMS

Requires consistent stimulation
SF Simulation
Environmental Ground Truth
Passive Sensors
Behaviors (e.g.) - Scout - March -
Occupy - Target - Fire - Search /
Find - Localize - Track - ..
Data (e.g.) - Terrain (e.g. surface, hydro,
veg, ..) - Atmosphere (e.g. aerosols,
clouds, ..) - Ocean (e.g. sea state, SVP,
..) - Space (e.g. particle flux, fields,
..) - Cultural (e.g. roads, structures,
..) - Military (e.g. engineering works, ..)
Effects (e.g.) - Propagation (e.g. intervis
/ TL) - Mobility (e.g. corridor)
Active Sensors
Internal Dynamics
Weapons Countermeasures
Impacts (e.g.) - Obscurants/Energy (e.g.
smoke, chaff, noise, ..) - Damage (e.g.
structural, combat engineering, craters,
..)
Units / Platforms
7
JETS Fundamentals

• quick on-the-fly computation
• greatly simplified physics

• multi-variate, multi-resolution
• atmosphere/adaptable to space/ocean
• pre-computed/run-time

• appropriate levels of fidelity
• steer models using environmental tailoring

8
Where are we going?
9
JETS PFM Algorithm
10
JETS Project Tasks

• Task 1 Tailoring Requirements Analysis
• Identify requirements for atmosphere-ocean
  tailoring to support the training needs of JSIMS
  federates
• Task 2 SNE Representation Frameworks to Support
  Tailoring
• Develop representations of atmospheric features,
  effects, models, and data
• Identify schemes for representing scenarios and
  environment-sensor-entity interactions important
  to environmental tailoring
• Task 3 Numerical Algorithm Development and
  Tradeoff Studies
• Develop and test alternative algorithms for
  environmental tailoring
• Identify conditions under which each is most
  appropriate
• Optional Tasks
• Develop stochastic model generation algorithm
• Ocean and surf-zone numerical tailoring
• Geo-temporal constraint analysis
• Platform-level and Unit/Platform-level proof of
  concept demo

11
PFM AlgorithmTypes of edits

• Variable edit the edit
  performed by the user (at the edit time ti)
• Correlated edit edit to a
  variable correlated with the user edit (at ti)
• Temporal edit edit to edited
  variable and correlated variables at times
  preceding and following edit times
  (ti-m,,tim)

ti is edit time
Horizontal slices of
edited variable
time
t
t
t
X
X
X
i-1
i
1
i
1
1
1
1
X
X
n
n
y
x
Variable edit
Temporal blend
Temporal blend
t
X
i
t
i
X
X
t
n
i
Horizontal slices of
unedited variables
Correlated edit
12
PFM AlgorithmCorrelation Component
13
PFM AlgorithmConstraint Component

• Atmospheric constraints limit range of acceptable
  edits
• climatological extremes
• reasonable spatial and temporal gradients
• physical laws (thermodynamics, moisture, )
• Constrained Domain Analysis provides
• guidance to user as to allowable range of edits
• scheme for blending edits in the vertical
  direction

14
Example of PFM Cloud Edit

• MM5 NWP model used to generate a consistent SNE
• NOGAPS 1º x 1º grid used to initialize and bound
  MM5
• December 98 scenario over the contiguous U. S.
• Desired edit place a cloud mask at t 12 h
  centered at 35º latitude with 1500 km radial
  coverage

MM5 SNE geographic coverage
15
Example of PFM Cloud Edit
16
Sensor Impacts Demonstration
17
TAWS Sensor Model

• TAWS (Target Acquisition Weather Software) used
  to demonstrate impact of SNE editing
• predicts the performance of air-to-ground weapons
  and direct view optics
• includes models for Vis/IR/Laser
• based on environmental and tactical information
• TAWS Proof-of-concept used to demonstrate...
• significant side effects of edits to the SNE
• importance of correlating SNE edits properly
  (application of JETS PFM algorithm)

18
TAWS Demo - Control and Edit Runs
RUN 1 CONTROL
RUN 2 EDITED SNE
RUN 3 CORRELATED (JETS) SNE
19
TAWS Demo - The Scenario

• MM5 run over China Jan 2, 1999 at 12Z to produce
  24 hour SNE scenario
• TAWS run with MM5 weather data input to find
  sensor range detection
• TARGET Scud Mobile Missile Launcher
• SENSOR IR at 2500 ft., TOT 03Z (edit time)
• BACKGROUND Vegetative

MM5 model domain
20
Application of PFM Algorithm
03Z
03Z
Surface Temperature
03Z
03Z
Surface Relative Humidity
Unedited
Edited
21
TAWS DemonstrationModel Results Delta T and
Target Detection
CONTROL
CONTROL
JETS EDIT
EDIT
EDIT
JETS EDIT
DT (Target vs. Background)
Range Detection
22
TAWS DemonstrationModel Results Target
Detection over Azimuth
Time over Target 0500
Time over Target 0500
EDIT
EDIT
JETS EDIT
JETS EDIT
270
270
CONTROL
2B
2A
CONTROL
Range Detection
DT (Target vs. Background)
23
Sensor Model Results

• Impact of SNE editing on sensor models can be
  substantial if not critical
• Without proper correlation of the SNE, sensor
  impacts are not physically correct
• demonstration emphasizes correlation of T and RH
• inconsistent result may bring about poor training
• With application of JETS, tailored scenario is
  correlated
• sensor impacts more realistic and properly
  synchronized with the environment
• overall training experience is improved

24
CONCLUSIONS
25
Conclusions

• JETS PFM algorithm used to provide SNE editing
  capability
• correlation component used to assure changes to
  SNE are physically consistent across time and
  space
• constraint component used to provide guidance as
  to appropriate range of acceptable edits
• TAWS Sensor Model demonstration
• edited SNE run through TAWS without JETS PFM
  results in improper representation of the
  environment
• edit using PFM results in more physically
  consistent SNE and more accurate sensor response