Upload
jonathan-rice
View
216
Download
1
Tags:
Embed Size (px)
Citation preview
Agent-Based Simulation of Agent-Based Simulation of Biocomplexity: Interactions of Biocomplexity: Interactions of
Natural Organic Matter, Natural Organic Matter, Mineral Surfaces, and Mineral Surfaces, and
MicroorganismsMicroorganisms
Gregory R. MadeyGregory R. MadeyDepartment of Computer Science and Department of Computer Science and EngineeringEngineering
Patricia A. MauricePatricia A. MauriceDepartment of Civil Engineering & Department of Civil Engineering & GeosciencesGeosciences
Center for Environmental Science and Center for Environmental Science and TechnologyTechnology
Natural Organic MatterNatural Organic Matter• Forms primarily from the breakdown of organic Forms primarily from the breakdown of organic
debrisdebris• Consists of a complex mixture of heterogeneous Consists of a complex mixture of heterogeneous
molecules that varies spatially and temporallymolecules that varies spatially and temporally• Is ubiquitous in aquatic and terrestrial environmentsIs ubiquitous in aquatic and terrestrial environments• Serves as a primary C source to ecosystemsServes as a primary C source to ecosystems• Binds metals, radionuclides, organic pollutants and Binds metals, radionuclides, organic pollutants and
helps to control their mobilitieshelps to control their mobilities• Acts as a natural ‘sunblock’ in surface watersActs as a natural ‘sunblock’ in surface waters• Defies simple analysis and deterministic, Defies simple analysis and deterministic, ab-initioab-initio
modeling because of complex, variable structure.modeling because of complex, variable structure.
NOM has complex, NOM has complex, variable structurevariable structure
NOM model (Leenheer)
250 300 350 400 450
excitation wavelength (nm)
300
350
400
450
500
550
600
- irrad - sorption, 8.72 mg/l
Fluorescence EEM, NMR spectra
220 200 180 160 140 120 100 80 60 40 20 0
chemical shift (ppm)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Integration ranges
S2, RO isolate, Spring '97
Aromatic
Acetal/Aromatic
Aliphatic
21.6
12.7
8.425.7
34.1
24.2 7.4
Forest Service Bog (FSB) [DOC] 7 MW 2200
Twomile Creek (TMC) [DOC] 17 MW 1500
Nelson Creek (NLC)[DOC] 79 MW 900
Many processes affect NOM Many processes affect NOM propertiesproperties
• AdsorptionAdsorption• Photodegrad-Photodegrad-
ationation• CoagulationCoagulation• BiodegradationBiodegradation• Primary Primary
productionproduction
This complexity lends itself This complexity lends itself well to ‘biocomplexity’ well to ‘biocomplexity’ modeling, agent-based modeling, agent-based models.models.• System heterogeneity controls reactivitySystem heterogeneity controls reactivity• High degree of spatiotemporal variabilityHigh degree of spatiotemporal variability• Complex and cooperative interactions Complex and cooperative interactions
between different components and processes between different components and processes • Need for scaling between laboratory and field Need for scaling between laboratory and field
experimentsexperiments• Two approaches being used in this model:Two approaches being used in this model:
Composition-based modeling Composition-based modeling Molecular weight (Mw)-based modelingMolecular weight (Mw)-based modeling
NOM concentration, Mw generally NOM concentration, Mw generally decrease from soils into ground decrease from soils into ground waterwater
depthbelowland surface
NOM adsorption to minerals and NOM adsorption to minerals and bacteria decreases with increasing bacteria decreases with increasing pHpH
0
20
40
60
80
100
1 2 3 4 5 6 7 8
pH
% FA Adsorbed
Adsorption fractionates NOMPreferential adsorption of high Mw components
Decreased % sorption---->
Challenges for Research Challenges for Research into Biocomplexityinto Biocomplexity
• Heterogeneity of system componentsHeterogeneity of system components– Component identities only partially knownComponent identities only partially known– Often cannot assume homogeneity, averages, aggregate values, or Often cannot assume homogeneity, averages, aggregate values, or
simple distributions; cannot ignore individual differencessimple distributions; cannot ignore individual differences
• Complex interactions between componentsComplex interactions between components– Processes and signaling pathways only partially knownProcesses and signaling pathways only partially known– Often cannot assume a well mixed solution, spatial independenceOften cannot assume a well mixed solution, spatial independence
• Complex interactions with environmentComplex interactions with environment– Dynamic coupling/feedback between components and systemDynamic coupling/feedback between components and system– Phenomena at different system levelsPhenomena at different system levels
• Limitations ofLimitations of– Equation-based modeling Equation-based modeling – Reductionism (complexity —> emergence —> scaling problems)Reductionism (complexity —> emergence —> scaling problems)– Sensitive dependence to initial conditionsSensitive dependence to initial conditions
New Computer Capabilities New Computer Capabilities —>—>
New Methods for ScienceNew Methods for Science• Faster/cheaper/more CPUs —>Faster/cheaper/more CPUs —>
Individual-based/Agent-based ModelingIndividual-based/Agent-based Modeling- Stochastic modeling- Stochastic modeling- Discrete event simulation- Discrete event simulation
• Bigger/cheaper/more Disk Drives —>Bigger/cheaper/more Disk Drives —>Data warehouses/Data miningData warehouses/Data mining
- Sensor nets- Sensor nets- High dimensional, merged data sets- High dimensional, merged data sets- Data from simulations- Data from simulations- Computer-assisted discovery- Computer-assisted discovery
Agent-based Agent-based Modeling and Modeling and SimulationSimulation• Individual-based Individual-based
modeling (IBM)modeling (IBM)
• Discrete event Discrete event simulationsimulation
• Stochastic birth-Stochastic birth-death models (SBD)death models (SBD)
• Cellular automata Cellular automata (CA)(CA)
• Artificial Life (AL)Artificial Life (AL)
Focus of our NOM Focus of our NOM ModelingModeling
Soil
Ground Water
Surface WaterNOM—Microbes
Inputs
Outputs
BackgroundBackground• Prior modeling work often too simplistic to represent Prior modeling work often too simplistic to represent
NOM heterogeneity and its complex behaviors in NOM heterogeneity and its complex behaviors in ecosystems (e.g., carbon cycling models, nitrogen ecosystems (e.g., carbon cycling models, nitrogen cycling models)cycling models)
• Prior modeling work often too compute-intensive to be Prior modeling work often too compute-intensive to be useful for large-scale environmental simulations (e.g., useful for large-scale environmental simulations (e.g., molecular models employing connectivity maps or molecular models employing connectivity maps or electron densities)electron densities)
• Hence, a Hence, a Middle Computational ApproachMiddle Computational Approach is taken … is taken …
Copyright 1998, Thomas M. Terry, The University of Conn
Elemental Cycling Connectivity MapsAgent-based Approach
ModelingModeling• Molecules and microbes are objectsMolecules and microbes are objects• Molecules and microbes have attributesMolecules and microbes have attributes
– Heterogeneous, distributionsHeterogeneous, distributions– Currently 1,000 objects, testing 10,000 and moreCurrently 1,000 objects, testing 10,000 and more
• Molecules have behaviors (reactions)Molecules have behaviors (reactions)– Molecules in simulation are a representative sample of Molecules in simulation are a representative sample of
the larger populationthe larger population– Behaviors are stochastically determinedBehaviors are stochastically determined– Dependent on the:Dependent on the:
• Attributes (intrinsic parameters)Attributes (intrinsic parameters)• Reaction ratesReaction rates• Environment (extrinsic parameters)Environment (extrinsic parameters)
Modeling (cont)Modeling (cont)• Objects of interestObjects of interest
– Macromolecular precursorsMacromolecular precursors•PolysaccharidesPolysaccharides•ProteinsProteins•Polynucleotide, tannin, lignin, polyterpene, cutinPolynucleotide, tannin, lignin, polyterpene, cutin
– Smaller moleculesSmaller molecules•PhospholipidsPhospholipids•SugarsSugars•Amino acidsAmino acids•FlavonoidsFlavonoids•QuinonesQuinones
– Microbes Microbes
Modeling (cont)Modeling (cont)• AttributesAttributes
– More specific than “percent carbon” but less detailed than More specific than “percent carbon” but less detailed than a molecular connectivity mapa molecular connectivity map
– Elemental compositionElemental composition• Number of Number of C, H, O, N, SC, H, O, N, S and and PP atoms in molecule atoms in molecule
– Functional group countsFunctional group counts• Double-bondsDouble-bonds• Ring structuresRing structures• Phenyl groupsPhenyl groups• AlcoholsAlcohols• Phenols, ethers, esters, ketones, aldehydes, acids, aryl acids, Phenols, ethers, esters, ketones, aldehydes, acids, aryl acids,
amines, amides, thioethers, thiols, phosphoesters, phosphatesamines, amides, thioethers, thiols, phosphoesters, phosphates– The time the molecule entered the systemThe time the molecule entered the system– Precursor type of moleculePrecursor type of molecule
Modeling (cont)Modeling (cont)
• Behaviors (reactions and processes)Behaviors (reactions and processes)– Physical reactionsPhysical reactions
• Adsorption to mineral surfacesAdsorption to mineral surfaces– Initial adsorptionInitial adsorption– Surface migration to high-energy sitesSurface migration to high-energy sites– Hemi-micelle formation at high coverage (cooperative, Hemi-micelle formation at high coverage (cooperative,
hydrophobicity dependent)hydrophobicity dependent)
• Aggregation/micelle formation (e.g., metal cation-Aggregation/micelle formation (e.g., metal cation-induced aggregation) - flocsinduced aggregation) - flocs
• Transport downstream (surface water)Transport downstream (surface water)• Transport through porous mediaTransport through porous media• VolatilizationVolatilization
Modeling (cont)Modeling (cont)• Behaviors (reactions and processes)Behaviors (reactions and processes)
– Chemical reactionsChemical reactions• Abiotic bulk reactionsAbiotic bulk reactions
– HydrolysisHydrolysis– HydrationHydration– Ester condensationEster condensation– Thermal decarboxylationThermal decarboxylation
• Abiotic surface reactionsAbiotic surface reactions• Direct photochemical reactionsDirect photochemical reactions• Indirect photochemical reactionsIndirect photochemical reactions• Extracellular enzyme reactions on large moleculesExtracellular enzyme reactions on large molecules
– BacteriaBacteria– FungiFungi– AlgaeAlgae
• Microbial uptake by small moleculesMicrobial uptake by small molecules
Modeling (cont)Modeling (cont)
• Environmental parametersEnvironmental parameters– TemperatureTemperature– pHpH– Light intensityLight intensity– Metal concentrations (e.g., Al and Fe)Metal concentrations (e.g., Al and Fe)– Bacterial activityBacterial activity– Water flow rate/pressure gradientWater flow rate/pressure gradient
• Environment: 2D Grid, mineral surfaces, soil Environment: 2D Grid, mineral surfaces, soil porespores
• Simulation parameters: run time, data collectionSimulation parameters: run time, data collection
NOM 1.0NOM 1.0
• VisualizationVisualization– Simulation and Animation of MoleculesSimulation and Animation of Molecules
• Web-Based AccessWeb-Based Access– Standard Browser InterfaceStandard Browser Interface
• HTML Forms / JSPHTML Forms / JSP• Java ServletsJava Servlets• JDBC - Oracle DatabaseJDBC - Oracle Database• Oracle Forms and ReportsOracle Forms and Reports
– Shared Data and SimulationsShared Data and Simulations– Collaboration Support: Web-board, Chat, Collaboration Support: Web-board, Chat,
mail server, file upload/downloadmail server, file upload/download
Web Access to NOM Web Access to NOM SimulationSimulation
VisualizationVisualization Black - No AdsorptionGrays - Levels of AdsorptionRed - LigninsGreen - CelluloseBlue - ProteinsYellow - ReactedOrange - Adsorbed
QuickTime™ and aAnimation decompressorare needed to see this picture.
Visualization - NOM molecules in solution and adsorption Visualization - NOM molecules in solution and adsorption
Web Browser SetupWeb Browser Setup
VisualizationVisualizationColor coded moleculesColor coded molecules
- Solution- Solution- Adsorbed- Adsorbed- Mw- Mw
Web-based Web-based ReportsReports
Modeling and Simulation in Modeling and Simulation in support of Biocomplexity support of Biocomplexity ResearchResearch
Modeling
SimulationObservation
Theory
ComputerExperiments
Lab & FieldExperiments
SummarySummary• Challenges for biocomplexity researchChallenges for biocomplexity research• New computer science toolsNew computer science tools• Middle computational approachMiddle computational approach• Agent-Based Modeling approachAgent-Based Modeling approach• Stochastic (Monte Carlo based simulation)Stochastic (Monte Carlo based simulation)• NOM Molecules & Microbes as AgentsNOM Molecules & Microbes as Agents• Web-based Simulation, databases, data Web-based Simulation, databases, data
warehouse, visualization, database queries, warehouse, visualization, database queries, data miningdata mining
• Invitation to collaborate … Invitation to collaborate …
ACKNOWLEDGEMENTSACKNOWLEDGEMENTS
• Center for Environmental Science and Technology at the Center for Environmental Science and Technology at the University of Notre DameUniversity of Notre Dame
• National Science Foundation (National Science Foundation (Information Technology Information Technology Research - DEBResearch - DEB, Hydrologic Sciences, EMSI); Environmental , Hydrologic Sciences, EMSI); Environmental Protection Agency STAR; Department of Energy Protection Agency STAR; Department of Energy
• Notre Dame students & post-docs: L. Arthurs, Y-P Huang, X. Notre Dame students & post-docs: L. Arthurs, Y-P Huang, X. Xiang, and many moreXiang, and many more
• Steve Cabaniss, University of New MexicoSteve Cabaniss, University of New Mexico