CloudScape ® VR Visidyne, Inc. 5951 Encina Road, Suite 208 Goleta, CA 93117 805-683-4277 (voice) 805-683-5377 (fax) [email protected] (e-mail) September.

CloudScapeCloudScape®® VR VR

Visidyne, Inc.5951 Encina Road, Suite 208

Goleta, CA 93117805-683-4277 (voice) 805-683-5377 (fax)

[email protected] (e-mail)

September 1999

Prepared by

John DeVore, Ken Sartor and Tim Stephens

Applied Physics Group, Visidyne, Inc.

Visidyne, Inc.3322 South Memorial Pkwy, Suite 223

Huntsville, AL 35801256-880-3411 (voice) 256-880-3284 (fax)

[email protected] (e-mail)

A current version of this briefing is available atftp.visidyne.com/cloud

2

Briefing Overview

Who is Visidyne? What is CloudScape® VR?

– API for Vega™ development system

– Sample CloudScape® VR images How does CloudScape® VR Work?

– Rendering

– Databases

– Validation How can I get cloud databases? What are the CloudScape® VR Development Plans? Where can I find technical references?

3

Who is Visidyne?

4

Who is Visidyne?

A small business Headquarters in Burlington, MA

– Offices in Santa Barbara, CA, Huntsville, AL and Parsippany, NJ Founded in 1969 Provides engineering services,

instrumentation and software products Nationally-recognized capability in several key

defense technologies– Atmospheric physics

– Battlefield and battlespace environments

– Optical sensor systems 60+ staff

– Dr. Jack Carpenter, President

– Dr. Tim Stephens, Vice President and Manager, Applied Physics Group

– Dr. John DeVore, Manager, Computer Visualization Program

– Dr. Jim Thompson, Radiative Transfer

– Dr. Ken Sartor, Real-time Visualization

5

Silicon Graphics

Optics

Modeling &

Simulation

Measurements&

Analysis

Instrumentation&

Fabrication

Visidyne Technical Focal Points

6

Visidyne Applied Physics Group Customers

Major U.S. Department of Defense (DoD) Customers– Defense Threat Reduction Agency (formerly DNA)

– Ballistic Missile Defense Organization

– U.S. Air Force Research Laboratory

– U.S. Army Space and Missile Defense Command Major Industrial Customers

– Nichols Research Corporation (Huntsville, AL)

– Photon Research Associates (La Jolla, CA)

– Kaman Sciences Corporation (Colorado Springs, CO) Small Business Innovative Research Program

– Cloud prediction and classification

– Real-time simulation

– DIS / HLA interactive simulation environment

– Ground target BRDF modeling

7

What is CloudScape® VR?

1. API for Vega™ development system

8

CloudScape® VRA Vega™ module for

– weather clouds, – battlefield effects, and

– other atmospheric obscurants

QuantitativePhysics-Based Background and Obscuration Prediction

3 - DFaceted Representation

of Cloud Regions

Real TimeSupport for Simulation

and Virtual Reality

Multi-SpectralVisible, IR and UV Bands

What is CloudScape® VR?

CloudScape® VR is available for purchase from MultiGen-Paradigm, Inc.

9

USAF CloudScape®

High-fidelity Fixed-frame Images

CloudScape® VR

Real-time Simulation and Training

CloudScape® VR built via Technology Transfer

Technology originally developed for US Air Force Research Laboratory to complement CSSM cloud-microphysics predictions

USAF CloudScape– Pixel-by-pixel calculation of background radiance and obscuration– Numerical integration along each line of sight

CloudScape VR uses two approximations for real-time performance– Database of azimuth/elevation samples is interpolated at vertices– Gouraud-shading by graphics engine interpolates between vertices

10

A Vega™ module for radiometric visualization of

diffuse, translucent atmospheric (cloud) phenomena

How do you use CloudScape® VR?

An API library with a LynX™ panel

For use with Vega™

– To generate virtual-reality environments including clouds

– Real-time “out-the-window” view with

Quantitative cloud-obscuration of backgrounds

Semi-quantitative images of cloud-radiance

For optional use with SensorVision™

– To generate radiometric (quantitative) virtual-reality environments

Backgrounds and obscuration

– SensorVision™ for solid objects, CloudScape® for diffuse objects

11

CloudScape® VR Features

3-D faceted cloud descriptions

Calibrated fog function

Full bi-directional-radiance computation

– Arbitrary solar (lunar) illumination angle

– Arbitrary viewing geometry

Any waveband (visible, infrared, ultraviolet)

12

CloudScape® VR LynX™ Panel

13

CloudScape® VR

SensorVision™

Vega™

Performer™

OpenGL®

User Application

SGI IRIX™

Planned

Windows NT®

OpenGL®

Vega™ NT

CloudScape® VR Libraries in Simulation Hierarchy

14

What is CloudScape® VR?

2. Sample CloudScape® VR Images

15

105 mm Shell

1000 LB. BombStratus Layer

Vega™ Performer Town Simulation

CloudScape® VR Sample ImageGround-Based Scenario

16

CloudScape® VR Sample ImagesGround Fog

Images with fog layer

Images without fog layer

17

7 am 11 am 5 pm

Viewing toward Northwest

Sun behind viewerClouds in frame are

shadowed by other clouds

Sun nearly overhead Sun at left front

Solar-Angle Effects on Stratocumulus Clouds

18

Sample CloudScape® VR Imageusing CSSM* Particulate Database

*Cloud Scene Simulation Model by U.S. Air Force Research Laboratory

19

Cloud TexturingTime-of-Day Dependence

7:00 am 12:00 nTextured

Untextured

20

Cloud TexturingViewing-Angle Dependence

Sun at Viewer Side

Sun In Front of Viewer Sun Behind Viewer

21

Cloud TexturingBroken Versus Variable-Thickness Clouds

Broken Cloud Layer Variable-Thickness Cloud Layer

22

Cloud TexturingBroken Stratus Viewed From Above

Time-of-Day Dependence

7:00 am 12:00 n

23

Cloud Shadowing of Terrain

Solar Illumination

Direction

24

Dust Cloud from 105-mm Tank Round Buried in Wet Cohesive Soil

1 second 3 seconds 10 seconds 30 seconds

Time after tank round detonation

25

CloudScape® VR Sample ImageSooty Smoke Plume

26

Vega Mode vs. SensorVision (quantitative) Mode

CloudScape® VR with Vega™

CloudScape® VR with SensorVision™

27

SensorVision (quantitative) Mode

Vega window showing observer’s view of a tank and a large munitions cloud

SAOimage window showing quantitative values from the “grabbed” image

28

Multi-Spectral Rendering

Visible

MWIR3 to 5 microns

LWIR8 to 12 microns

Night Vision

29

CloudScape® VR Iso-Surface VisualizationHazardous Material Plume Example

30

How Does CloudScape® VR Work?

1. Rendering

31

CloudScape® VR Rendering Approach

Rendered Dust CloudCloud FacetizationVertex Radiance and Transmission

Parameterized by Angle

Vertex geometry, radiance and transmission information is contained in OpenCloud database

CloudScape® VR renders cloud in real time

32

CloudScape® VR Rendering Equations

Cloud Radiance

Composite Scene

Transmitted Background

+ =

Cloud Transmission

Scene Background

+ =

Transmitted Background

Background is attenuated by cloud opacity

Cloud radiance (brightness) is

added

33

CloudScape® VR Rendering Timelines

Facetized cloud model is combined with terrain and other models and rendered at frame rate (typically 30 to 60 Hz).

Cloud radiative properties are interpolated asynchronously for viewer location and aspect from tabulated bi-directional values at each vertex (typically a few Hz).

Cloud facet geometry and vertex radiative properties are obtained from a CloudScape® VR database at “key frame” times (typically every few seconds).

Key-frame rates depend on model; asynchronous and rendered frame rates depend on processor performance.

34

CloudScape® VR Use of Level of Detail (LOD)

Range = 45 m 115 m 175 m 300 m 500 m

# Facets = 1218 608 206 60 12

Re

du

ce

d L

OD

a

t lo

ng

ra

ng

eL

OD

-eff

ec

t a

t c

on

sta

nt

ran

ge

Wir

e F

ram

e

Ima

ge

35

Munitions Cloud Dust LoadingExamples show importance of physics-based solution

Hei

gh

t o

r D

epth

of

Bu

rst

Above Ground

Below Ground

Dry cohesive soil 0.031

Dry sandy soil 0.094

Moist sandy soil 0.22

Wet sand 0.44

Wet cohesive soil 1.0

Rock 0.025

Mass Loading FactorWET COHESIVE SOIL DRY SANDY SOIL

SURFACE

BURIED

Mass Loading Factor

F =1F =1

F = 0.1F = 0.1 F = 0.01F = 0.01

F = 0.1F = 0.1

1 second following the detonation of a 105-mm tank round

36

CloudScape® VR Visualization Steps

CloudGen™

Cloud 3-D Particulate Database

Select a cloud type

CloudRad™Select waveband(s),

solar illumination angle(s) and atmospheric state

OpenCloud™ Radiometric Database

Execute simulation CloudScape® VR

Real-time Cloud Imagery

Pre

pro

cess

ing

Ru

n T

ime

37

How Does CloudScape® VR Work?

2. Databases

38

OpenCloud™ Database Tools

CSSMdatabase

COMBICdatabase

CloudGen™ / Weather Toolkit• CloudGen™ / Weather• CloudScape® HF• AtmosRad™• CloudRad™• CloudTex™

OpenCloud database

CloudGen™ / Battlefield Toolkit

Optional

Optional Broken box indicates future development

MOSART

Optional

MOSART

Optional

39

Radiative Processes Modeled in CloudRad®

ReflectedSunshine

TransmittedEarthshine

ThermalEmission

ReflectedEarthshine

EarthshineShadows

CloudAttenuation

Skyshine

Reflected Skyshine

Path Radiance

Path Absorption

Sunshine / Moonshine

40

Cloud Radiance CalculationUsed in CloudRad™

Cloud illumination– Solar (or lunar) radiation

– Sky shine

– Earth shine Can include illumination from city lights

Self-shadowing of cloud– Shadowing of ground by clouds is also recorded

Cloud scattering– Single-particle scattering

– Multiple-particle scattering Build-up factor approximation is used

– Narrow angle forward scattering Thermal emission Obscuration

– Total direct beam loss

– Wide-angle beam loss

41

Details of CloudRad™ Calculation

Conversion of mass density to effective extinction coefficient– Robust , efficient Mie scattering model (developed by Dr. Thompson)

– Variety of particle size distributions

– Complex indices of refraction Water, ice, dust, and smoke Ultraviolet, visible, and infrared

Cloud surface facetization– Marching Cubes algorithm

– Vertex culling for levels of detail Tabulation of bi-directional radiance and transmission at vertices

– Fast, approximation to integral form of radiative transfer equation Exact in single scattering limit Validated multiple scattering approximation Variety of sources: sun, moon, earthshine and skyshine

– MOSART (U.S. Air Force Research Laboratory) parametric databases Atmospheric path attenuation and radiance Up- and down-welling atmospheric diffuse radiation

42

How Does CloudScape® VR Work?

3. Validation

43

San Luis Obispo, CA

Meas 5 x 10-

6 Calc 5 x

10-6

Meas 4 x 10-6

Calc 5 x 10-6

Meas 30 x 10-6

Calc 40 x 10-6

Early MorningUniform Stratus LayerSWIR Window Region

Meas 8 x 10-6

Calc 7 x 10-6

Data Comparisoncalculated vs. measured cloud radiance (W/cm2/sr)

44

Data ComparisonSolar Scattering Angle Dependence

1 Zachor, A. S., J. A. Holzer, and F. G. Smith, 1979, IR Signature Study, AFAL-TR-79-1012, Air Force Avionics Laboratory, Wright-Patterson AFB, OH.

• AFGL aircraft at 6.0 kft• Stratus at 3.0 kft• Solar elevation = 8.88°• LOS depression = 5.0°

• Background Measurement Program1

• Case 726/7/FB43• San Luis Obispo, CA• Spectra from 1800 to 3600 cm-1

• CloudScape® calculations• 3.70-3.85 m band• 8 m mean droplet size

10-4

10-5

10-6

CloudScape®

Measurements

45

Data ComparisonSpectral Details

Wavelength (m)

Forward (data)

Forward

Backward (data)

Backward

Side (data)

Side

3.25 3.5 3.75 4.00 4.25

10-4

10-6

10-5

10-7In-B

and

Rad

ianc

e (W

/cm

2 /sr

)/(2

5 cm

-1)

46

How Can I Get Cloud Databases?

47

Ways to Obtain OpenCloud™ Databasesfor use with CloudScape® VR

Sample databases are included with CloudScape® VR product Database packages will be available from Visidyne

– Weather Cloud Database

– Battlefield Cloud Database Database generation toolkits will be available from Visidyne

– CloudGen™/Weather Toolkit (release planned for Summer 1999) Beta demonstration version available

– CloudGen™/Battlefield Toolkit (in development) Engineering services available from Visidyne, Inc.

– Custom databases for clouds and atmospheric obscurants

– Customer specified: Sensor bands Cloud types Illumination conditions

OpenCloud™ database format specifications are available upon request

48

CloudGen™/Weather Toolkit

CloudGen™/Weather– Generates 3-D particulate database– Alternatively, CSSM* databases can be used

AtmosRad– Generates optional atmospheric propagation database with MOSART*– Sample databases provided for standard atmospheric states and bands

CloudScape® HF– High-fidelity pixel-by-pixel calculation of cloud radiance and transmission– FITS file output

CloudRad™– Generates OpenCloud radiance database for use in CloudScape® VR

CloudTex™– Generates texture file for use in CloudScape® VR

Sample cases– Input and output files for weather-cloud databases

*Use of MOSART and CSSM are optional; these codes may be obtained directly from U.S. Air Force Research Laboratory (Phillips Site)

49

CloudGen™/Weather ToolkitMain Screen

50

Relationship of Tools and DatabasesCloudScape® VR and CloudGen™/Weather Toolkit

Vega™

CloudRad™

Particulate Database

CloudGen™/Weather

Radiometric DatabaseCloudTex

Texture Database Purchased Databases

Developer Application

SensorVision™

CloudScape® VR

API Library Calls

CSSM

CSSM Database

Atmospheric Database

MOSART

ORMAT™AtmosRad™

51

CloudGen™/Weather

User input

– Number of cloud layers (currently 1)

– For each layer Horizontal extent, resolution and orientation Ceiling, cloud-top altitude, vertical resolution Atmospheric mean vertical profile (sounding data) Percent cover Wind speed and direction Random number seed Cloud type, OR

Particulate type Particle size distribution parameters Structure parameters

Alternatively, a CSSM*-generated database can be used

* Air Force Research Laboratory model

52

* Army Research Laboratory model (distribution restrictions apply)

CloudGen™/Battlefield

User input: generic model mode– Non-buoyant burst

– User input Soil type (6 categories) TNT-equivalent yield of munitions Switch between surface or shallow-buried burst Wind speed and direction (vertical profile)

– Unrestricted distribution COMBIC* model mode

– Reads user-supplied COMBIC-generated cloud databases

– Adds spatial structure and wind effects

53

What are the CloudScape® VR Development Plans?

54

CloudScape® Family Software

CloudGen™ / Weather Toolkit

– Fall 1999

Weather Cloud Database Set

– Fall 1999

CloudGen™ / Battlefield Toolkit and Databases

– Early 2000

CloudScape® VR

– Version 1.1 currently shipping

– Version 1.2 release: fall 1999

55

CloudScape® VR 1.2 Enhancements

Support for CGWT databases with 2-D texture

Cloud Shadows

Tiled-clouds

– (for large spatial extent)

Corrections of coding errors in 1.1

Updated LynX panel

56

Effect of Cloud Texture

CloudScape® VR Version 1.1Untextured

CloudScape® VR

Version 1.2Textured

57

Effect of Cloud Shadows

Vega simulation without CloudScape® VR

No Clouds

CloudScape® VR Version 1.1

Clouds, no Shadows

CloudScape® VR

Version 1.2Clouds and Shadows

58

Phenomena which can be Supported by OpenCloud Database Format

Candidate Modules for Future CloudGen™ Offerings

Rain: steady-state rainfall consistent with weather cloud model Vehicular: clouds from vehicular land traffic Fire: flame and smoke from steady-state ground fires Exhaust: heated exhaust gas from vehicles Muzzle: flash from guns Plume: from missiles and aircraft Stack: steady-state steam/smoke from ground sources Effluent: transient, non-instantaneous expulsion of material into the atmosphere (e.g.

underground bunker destruction) Obscurant: battlefield smoke Flare: IR battlefield countermeasure Splash: underwater detonations Chaff: clouds of RF reflectors Volcano: dynamic particulate and gaseous high-altitude clouds

59

Where Can I Find Technical References?

60

References

DeVore, J.G., J.H. Thompson, and R.J. Thornburg, 1995, “CloudScape®: Stochastic Cloud Visualization from Volumetric Descriptions:, in Cloud Impacts on DoD Operations and Systems 1995 Conference, D. Grantham, editor, Phillips Laboratory/GPAA, Hanscom AFB, MA, 41-44.

Schlueter, W. A., and J. G. DeVore , 1995, “Radiometric Atmospheric Dust Environments for Distributed Interactive Simulations”, in Proceedings, Sixth Annual Ground Target Modeling and Validation Conference, Volume I, K.R. Johnson, editor, Keweenaw Research Center, Houghton, MI, 19-27.

DeVore, J.G., J.H. Thompson, and R.J. Thornburg, 1996, “Physics-based Background Visualization from Volumetric Cloud Descriptions Using CloudScape®”, Proceedings of the IRIS Specialty Group on Targets, Backgrounds, and Discrimination, Sandia National Laboratory, Albuquerque, NM, 30 January - 1 February.

DeVore, J. G., J. H. Thompson, K. W. Sartor, T. L. Stephens, and R. J. Thornburg, “CloudScape® VR: Radiometric Visualization of Clouds for Interactive Training and Simulation”, Proceedings of the Cloud Impacts on DoD Operations and Systems Conference, PL-TR-97-2112, Phillips Laboratory, Hanscom AFB, MA, 1997.

Thornburg, R. J., J. G. DeVore, J. H. Thompson, R. J. Jordano, and T. L. Stephens, “Validation of CloudScape ® AF”, Proceedings of the Cloud Impacts on DoD Operations and Systems Conference , PL-TR-97-2112, Phillips Laboratory, Hanscom AFB, MA, 1997.

The references listed above are available in Adobe’s PDF format by anonymous ftp to the cloud/papers directory at ftp.visidyne.com.

61

Additional Information

Technical Contact

– Dr. John G. DeVoreVisidyne, Inc.5951 Encina Road #208 Goleta, CA 93117-2211805-683-4277 (voice) 805-683-5377 (fax)e-mail: [email protected]

Marketing Contact

– Ms. Kristin DavisMultiGen-Paradigm, Inc.14900 Landmark Blvd. #400Dallas, TX 75240972-960-2301 (voice) 972-960-2303 (fax)e-mail: [email protected]

Web Sites

– Visidyne: www.visidyne.com

– MultiGen-Paradigm: www.multigen.com