Developing Understanding of Ecological Economic Systems - PowerPoint PPT Presentation

1 / 39
About This Presentation
Title:

Developing Understanding of Ecological Economic Systems

Description:

Simulation Module Markup Language. XML-Based Declarative Language ... Module: Reusable component. Variable: Simulation atomic object. ... – PowerPoint PPT presentation

Number of Views:47
Avg rating:3.0/5.0
Slides: 40
Provided by: thomasp89
Category:

less

Transcript and Presenter's Notes

Title: Developing Understanding of Ecological Economic Systems


1
Developing Understanding of Ecological Economic
Systems
  • Thomas Maxwell
  • Robert Costanza
  • University of Maryland

2
Motivation
  • Unbridled expansion of human enterprise.
  • Depletion of natural life support systems.
  • Resource depletion -gt global tensions.
  • Interacting complex systems.
  • Tremendous uncertainty.
  • Potentially disastrous consequences.

3
Integrated Problem Solving
  • Vision
  • State of the world.
  • Possible future worlds (postitive negative).
  • What to tweak?
  • Expected outcomes of policy adjustments.
  • Methodology
  • Hard problem science.
  • Adaptive management.

4
Science in Service of Society
  • Comprehensive systems approach
  • Conceptual pluralism
  • Problem driver
  • Multiscale
  • Integrated modeling
  • Links with policy
  • Modeling as consensus building tool
  • Communicating uncertainty

5
Collaborative (Visible) Modeling
  • Realistic models require multiple teams
  • Modelers typically not computer scientists
  • Stake holders must be included
  • Communication to a wide audience

6
Three Stage Modeling Process
  • Scoping models
  • Consensus building
  • Research models
  • Understanding dynamics
  • Management models
  • Exploring scenarios

7
Modeling Collaboratory
  • Constructivist learning.
  • Paradigm expansion.
  • (narrow,linear,static) -gt
  • (broad,nonlinear,dynamic)
  • Conflict resolution.
  • Consensus building.
  • Collective decision making.
  • Develop management scenarios.

8
Supporting Collaborative Modeling
  • Graphical modeling tools
  • Modular model development
  • Transparent high performance computing
  • Integrated data access
  • Integrated visualization
  • Variety of formalisms and frames

9
  • Graphical Modeling
  • Model viewed and manipulated graphically.
  • Opens model development to non-programmers.
  • Facilitates rapid development of models.
  • Enforces modeling standards.
  • Facilitates collaboration in model development.
  • Graphical representation serves as a blackboard.

10
STELLA Model
11
Spatial Modeling Framework
12
Environmental Modeling Workbench
Spatial Modeling Environment
Inputs to multiple models
Coupled Bio-Hydro Simulation
Integrated wireless Sensor web
CavernSoft Collaborative Environment
Environmental Hydrology Applications Team
13
Two types of modules
  • Ecological Modules
  • No general theory.
  • Primary focus on modeling.
  • Examples
  • Macrophytes, Epiphytes, Consumers, Phytoplankton
  • Modules developed in Stella/SME.
  • Physical Modules
  • Theory well known (e.g. Navier Stokes).
  • Primary focus on computation.
  • Examples
  • hydrodynamics, atmospheric dynamics.
  • Modules developed externally and linked to SME.

14
  • Spatial Modeling Environment
  • Collaborative Spatial Modeling Workbench
  • Includes integrated support for
  • Icon-based unit module development
  • Module archiving and reuse
  • Integration of multiple spatial representations
  • Distributed computing
  • Web-based modeling simulation
  • Configuration, control, and visualization of
    remote simulations.
  • Data access and visualization
  • Real-time links to other apps (e.g. Swarm).

15
Spatial Modeling Environment
Module Repository
Module Builder
Simulation Driver
Unit model
Spatial model
Graphical modeling
HPC
STELLA
Code Generator
PowerSim
Module Constructor
SME Module Editor
Java Portal
SMML Module Library
16
  • Module Specification Language
  • Declarative
  • Modular
  • Fully visible structure dynamics
  • Supports encapsulation and specialization
  • Separate universal specs / site-specific configs
  • Platform and operating system independent
  • Facilitates extensive simulation services

17
  • Simulation Module Markup Language
  • XML-Based Declarative Language
  • Simulation Module Specification
  • Major Classes
  • Module Reusable component.
  • Variable Simulation atomic object.
  • Action Performs computation or data IO.
  • Event Orders the execution of Actions.
  • Frame Defines a spatial topology.

18
SMML Example
ltatom name"CONS_BIOM" id"CONS_BIOM"
status"private" type"state" gt
ltport type"input" name"CONS_INGEST" /gt
ltport type"input" name"CONS_EGEST" /gt
ltport type"input"
name"CONS_MORT_BIOM" /gt ltport
type"input" name"CONS_RESPIRATION" /gt
ltdynamic event"integrate" type"code"
gt ltcodegt ( ( (
CONS_INGEST-CONS_EGEST )-CONS_MORT_BIOM ) )
CONS_RESPIRATION )
lt/codegt lt/dynamicgt
ltport type"input" name"P1_CONS_IC" /gt
ltport type"input" name"CELL_SIZE" /gt
ltdynamic event"init" type"code"
gt ltdocgt
CARBON BIOMASS OF AN AGGREGATED CONSUMER. (KGC).
CONSUMERS EXCLUDE THE
MICRO ORGANISMS WHICH ARE
ACCOUNTED FOR IN THE
RESPIRATION FLUXES
lt/docgt ltcodegt ( (
P1_CONS_IC0.001 )CELL_SIZE )
lt/codegt lt/dynamicgt
lt/atomgt
19
SMML Example

lt?xml version"1.0"?gt lt!DOCTYPE compound SYSTEM
"http//iee.umces.edu/SME/dtd/smml.dtd"gt ltcompoun
d id"PLMD_module" name"PLMD_module" gt
ltcompound id"CONSUMERS_module"
name"CONSUMERS_module" location"CONSUMERS_
module.xml" gt ltport
type"output" name"CONS_EGEST" /gt
ltport type"output" name"CONS_MORT_BIOM" /gt
ltport type"output" name"CI_DETR"
/gt ... lt/compoundgt
... ltlink name"c__0" origin
"GLOBALS_module.CELL_SIZE" destination
"CONSUMERS_module.CELL_SIZE" /gt ltlink
name"c__1" origin "DETRITUS_module.DET_AVAIL
" destination "CONSUMERS_module.DET_AVAIL"
/gt ltlink name"c__2" origin
"DOM_module.DOM_C_AVAIL" destination
"CONSUMERS_module.DOM_C_AVAIL" /gt lt/compoundgt
20
Typical State Variables
  • Examples of some typical state variables
  • (Dissolved Inorganic) Nitrogen, Phosphorus
  • Water (Saturated, Unsaturated, Surface, Snow)
  • Detritus
  • Macrophyte (Non)Photosynthetic Biomass
  • Consumers
  • Deposited Organic Matter
  • Phytoplankton
  • Epiphytes

21
Agent Based Modeling in SME
  • Swarm agents can populate SME landscapes.
  • SME-Swarm integration
  • http//iee.umces.edu/villa/swarmsme
  • Swarm classes serve as wrappers for
  • SME model.
  • SME grid layers.
  • SME spatial variables.
  • Two-way remote data transfer.
  • Built on SNI simulation server architecture
  • http//iee.umces.edu/villa/sni

22
Multi-Grid Library
  • Integrates multiple spatial representations
  • Implements space in SME
  • Major Components include
  • Cell Spatially referenced area (or volume)
    element.
  • Grid Distributed set of Cells links.
  • Frame Hierarchy of distributed Grids.
  • Link Connection between Cells.
  • Intra-Grid spatial contiguity.
  • Inter-grid scaling relations or mappings.
  • Activation Layer Subset of Cells in a Frame.
  • Coverage Mapping Activation Layer -gt floats.

23
Distributed Processing
  • Spatial grid
  • partitioned over processors
  • Highly parallel application
  • Recursive N-section
  • excellent load balancing
  • Fully transparent to user

24
Model Calibration toolkit
  • Built on MPE toolkit
  • http//iee.umces.edu/villa/svp/
  • Calculate performance measure (MPE)
  • Estimate of match between model system.
  • Weighted sum of tests (Bounds, Theil, Freq, etc).
  • Search parameter space to maximize MPE.
  • Evolutionary and gradient searches.
  • Params, tests, searches configured in SME.

25
SME Java Portal
  • Desktop access to remote supercomputing resources
  • Web-enabled ( using java servlets )
  • Grid enabled ( using globus gram utility )
  • Java applet lt-gt Java servlet lt-gt C apps
  • Portal interfaces include
  • Workspace management
  • Module development
  • Model configuration
  • Simulation initialization, control,
    visualization

26
WorkSpace Manager
27
Configuration Manager
Documentation Panel Documentation of selected
command Model Panel Hierarchical View of model
objects Associated commands as boxes Command
Panel Structure of selected command Property
Panel Command Arguments
28
Parameter Editor
Edit Simulation Parameters Spreadsheet format
29
Simulation Control
Control Execution View Model Structure Trace
Dependencies View Model Equations Configure
Visualization
30
ViewServer Control Panel
  • Associates DataSets
  • with Viewers
  • Creates Viewers
  • Manages DataSets

31
2D Animation Viewer
  • 2D Animation Control
  • Dynamic and manual rescaling
  • ColorMap editor
  • Data viewer (point/spreadsheet)
  • Export as GIF or JPG

32
3D Animation Viewer
  • Dynamic Landscapes
  • Variable1 -gt Altitude
  • Variable2 -gt Color
  • Mouse controlled navigation

33
Image Spreadsheet
  • Simultaneous display of variables at multiple
    timesteps
  • Useful for time series comparisons
  • Configure start time, time step, magnification,
    scaling, etc.

34
Numerical Spreadsheet
  • View spatial data
  • Attach to vis panels
  • Follows animation
  • Export to Stat packages.

35
Chesapeake Bay Model
  • Links components
  • Circulation (OM3)
  • Ecology (SME)
  • Atmospheric coupling

Environmental Hydrology Applications Team
36
Collaborative Virtual Environment
Chesapeake Bay data in CVE with Cave5D/Virtual
Director
Environmental Hydrology Applications Team
37
Example Applications
  • Everglades Landscape Model
  • http//www.sfwmd.gov/org/erd/esr/elm/intro/welcome
    .htm
  • Patuxent Landscape Model
  • http//iee.umces.edu/PLM
  • Baltimore Ecosystem Study
  • http//baltimore.umbc.edu/lter
  • Illinois TES Models
  • http//blizzard.gis.uiuc.edu/

38
Environmental Hydrology

Environmental Hydrology Applications Team
39
SME Distribution
  • The SME home page
  • http//iee.umces.edu/SME3
  • Includes
  • Overview.
  • Technical documentation.
  • Publications.
  • Source code (C and java).
  • Links
Write a Comment
User Comments (0)
About PowerShow.com