ARIES                                        ARtificial Intelligence for Ecosystem ServicesARIES: Ecosystem Services asses...
ECOSYSTEM SERVICES in the MILLENNIUMASSESSMENTOptimal for communication, raising awareness   Supporting services          ...
A quantitatively based framework for ES
ARIES in a nutshell• A rapid assessment toolkit for ecosystem services and  their values; not a single model but an intell...
Project history Funded by the US National Science Foundation; follows  NSF-funded valuation database project NSF grant t...
AREAS of APPLICATION (changing rapidly)
Conceptual Ecosystem Service framework
ECOSYSTEM SERVICES in ARIESBenefit-oriented, optimal for quantification, modeling and spatial mapping    Ecosystem Service...
EXAMPLE ofARIES ONTOLOGY
EXAMPLE ofARIES ONTOLOGY
EXAMPLE ofARIES ONTOLOGY
EXAMPLE ofARIES ONTOLOGY
The three elements of ES modeling in ARIES                            1. provisionsheds1. Areas of provision ofES and biod...
“Conventional” ES source mapping                                                     Source mapping estimates             ...
STAKEHOLDER MAPPINGAnalysis can be performed for all relevant beneficiarygroupsMaps the location and levelof need of the p...
Modeling ES provision For entire model or model inputs:   • Use existing ecological models & their outputs if they exist ...
Modeling ES sinks Depending on the service,  sinks could provide a benefit:   • Absorption of flood water,     nutrients...
Modeling ES use   Similar process to modeling ES provision     •   How do we locate (potential) users of ES on the landsc...
Example: wave sink model (coastal protection MG)
Identifying carriers & flow paths                                                       Recreation, flood                 ...
Flow mappingAn agent-based approach   All flow districts start in    an empty state   Edges represent    transition prob...
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
4. ES flow propagationCarrier usedCarrier sunk
ES FLOW MAPPINGFlows connect sources and beneficiaries                                               Critical flow paths s...
EXAMPLES: Subsistence fisheries for coastal communities, Madagascar 1                         2                       31. ...
EXAMPLES: Veracruz , Mexico: water services                                                                      Selected ...
Novel results from flow analysis Analysis of flows wasn’t available before ARIES and  computes source, sink, use and flow...
TARGETING PRECISE AREAS THROUGH FLOWANALYSISapplicable to source, use and sink User draws source area of interest         ...
The “magic number”: new indicators     Using information about actual flows, new overall     indicators can be computed (w...
The public interface: a Web 2.0 application                                              37
STORYLINES are suggested by ARIES and created byusers to address each benefit deriving from ESServices are broken down int...
Ex-ante scenario definitionGlobal change scenarios can be merged with local land use changesPre-defined GLOBAL SCENARIOSe....
Routing linear features (roads,pipelines)  baseline                             reforested  Scenario 1: routing that minim...
Identification and ranking of areas for offsetting                                            ARIES can produce a full ES ...
Scenario based quantitative valuationQuantify impact of choices on specific stakeholder groups                          35...
VALUATION and ARIES: getting real?ARIES is agnostic about valuation and tries to counteract inaccuracy in the “state of  t...
Criticality thresholds and valuation                   Critical: Perfectly                               Important:       ...
VALUATION in ARIES can be INFORMED BY                    CRITICALITY                                                      ...
CONCORDANCE VALUE vs. MONETARYabstract valuation = concordance between benefit flows and stakeholderpriorities            ...
ARIES and economic valuation: asummary  Primary valuation:    • WTP surveys, hedonics, travel cost, consumer      expendi...
THANK YOUFor more information:    www.ariesonline.org    info@ariesonline.org
ARIES - ARtificial Intelligence for Ecosystem Services
Upcoming SlideShare
Loading in...5
×

ARIES - ARtificial Intelligence for Ecosystem Services

307

Published on

Published in: Technology, Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
307
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • 20 SLIDES MAX!!!
  • ARIES identifies areas of provision and use of ecosystem services and the flows that link the two. Provisionsheds are mapped as in this case where biodiversity and water supply are provided to nature-based tourists and rice farmers in Madagascar, which is one of our case study areas (along with Puget Sound and Mexico as we’ll see later)
  • The carrier may or may not actually used by a beneficiary, so may not be an actual (realized) ES
  • Mark Mulligan’s “flow traps” -> he’s clearly already looking at source/sink/flow/beneficiary dynamics
  • MENTION diversity of flow types associated with marine systems (incl. overland hydrologic flow, inland wave/storm movement, ocean currents, movement of biological organisms (fish & other marine species), human transportation/migration)
  • Stated & revealed preferenceFunction transfer: functions only exist for a small subset of ES & LULC types
  • Transcript of "ARIES - ARtificial Intelligence for Ecosystem Services"

    1. 1. ARIES ARtificial Intelligence for Ecosystem ServicesARIES: Ecosystem Services assessment for the 21st centuryFerdinando Villa, BC3 Ken Bagstad, Gary Johnson, Brian Voigt, Marta Ceroni, Josh Farley Rosimeiry Portela, Miroslav Honzák, Nalini Rao, Celia Harvey Dave Batker, Maya Kocian, Jen Harrison-Cox, Jim Pittman Silvia Silvestri, Kristian Teleki, Mónica Barcellos-Harris
    2. 2. ECOSYSTEM SERVICES in the MILLENNIUMASSESSMENTOptimal for communication, raising awareness Supporting services Regulating services -Nutrient cycling -Climate regulation -Net primary production -Disturbance regulation -Pollination & seed dispersal -Water regulation -Hydrologic cycle... -Nutrient regulation... Provisioning services Cultural services -Water supply -Recreation -Food -Aesthetic -Raw materials... -Spiritual & historic...Ecosystem Services are a multiple-scale problem where provision and use havedifferent modes and scales, and flow across the landscape in different manners.Quantitative, spatially explicit assessment and valuation require more systematicand less general definitions.
    3. 3. A quantitatively based framework for ES
    4. 4. ARIES in a nutshell• A rapid assessment toolkit for ecosystem services and their values; not a single model but an intelligent system that customizes models to user goals.• Demonstrate a mapping process for ecosystem service provision, use, and flow where most ES assessments only look at provision.• “Honest” probabilistic models inform decision-makers of likelihood/uncertainty of all possible outcomes; can operate in conditions of data scarcity.• Web based, customizable for specific user groups, geographic areas and policy goals; custom tools implement specific “bottom line”
    5. 5. Project history Funded by the US National Science Foundation; follows NSF-funded valuation database project NSF grant to UVM, CI, EE from ABI program ($927,000); additional funding from UNEP-WCMC, CI. Project lead moved to BC3, Bilbao in 9/2010. 1.0 release in Summer 2011 Interest from public sector (country, state, county governments) Interest from NGOs large and small, UNEP Interest from private sector (extractive industry)
    6. 6. AREAS of APPLICATION (changing rapidly)
    7. 7. Conceptual Ecosystem Service framework
    8. 8. ECOSYSTEM SERVICES in ARIESBenefit-oriented, optimal for quantification, modeling and spatial mapping Ecosystem Services: the effects on human well-being of the flow of benefits from an ecosystem endpoint to a human endpoint at given extents of space and time Flow
    9. 9. EXAMPLE ofARIES ONTOLOGY
    10. 10. EXAMPLE ofARIES ONTOLOGY
    11. 11. EXAMPLE ofARIES ONTOLOGY
    12. 12. EXAMPLE ofARIES ONTOLOGY
    13. 13. The three elements of ES modeling in ARIES 1. provisionsheds1. Areas of provision ofES and biodiversity3. Flow paths betweenareas of provision andareas of use2. Areas of use of ESand biodiversity wherebeneficiaries are located 14 2. benefitsheds
    14. 14. “Conventional” ES source mapping Source mapping estimates the potential value provided by each ES (tonnes of sequestered C in this image) ARIES builds the source models according to the geological, ecological and climate variables describing the areas. ARIES is the only approach that also estimates in conditions of Pink overlay is a visual cue to uncertainty data scarcity.
    15. 15. STAKEHOLDER MAPPINGAnalysis can be performed for all relevant beneficiarygroupsMaps the location and levelof need of the potentialbeneficiaries of eachservice.Beneficiary maps can bealso made for actual andpotential beneficiariesthrough flow analysis.Potential beneficiaries canbe the object of planningenhanced flows for positiveimpact assessment. Blue overlay represents local farming communities and their dependence on soil deposition/erosion
    16. 16. Modeling ES provision For entire model or model inputs: • Use existing ecological models & their outputs if they exist • If no good models exist, build ad hoc models based on expert ecological knowledge How much of a given benefit is produced for each landscape district?
    17. 17. Modeling ES sinks Depending on the service, sinks could provide a benefit: • Absorption of flood water, nutrients Or a be a detriment: • Visual blight reducing the quality of views • Dampening out of values over distance
    18. 18. Modeling ES use Similar process to modeling ES provision • How do we locate (potential) users of ES on the landscape and quantify their need?
    19. 19. Example: wave sink model (coastal protection MG)
    20. 20. Identifying carriers & flow paths Recreation, flood regulation, manyHydrologic services Aesthetic viewsheds ecosystem goods Carbon sequestration, some cultural values Recreation, aesthetic proximity, some cultural services
    21. 21. Flow mappingAn agent-based approach All flow districts start in an empty state Edges represent transition probabilities Each location contains:  Source value  Sink and use rates & capacities  Sink cache  Use cache  Carrier cache
    22. 22. 4. ES flow propagationCarrier usedCarrier sunk
    23. 23. 4. ES flow propagationCarrier usedCarrier sunk
    24. 24. 4. ES flow propagationCarrier usedCarrier sunk
    25. 25. 4. ES flow propagationCarrier usedCarrier sunk
    26. 26. 4. ES flow propagationCarrier usedCarrier sunk
    27. 27. 4. ES flow propagationCarrier usedCarrier sunk
    28. 28. 4. ES flow propagationCarrier usedCarrier sunk
    29. 29. 4. ES flow propagationCarrier usedCarrier sunk
    30. 30. ES FLOW MAPPINGFlows connect sources and beneficiaries Critical flow paths show areas most critical to ensure ES flow to the More suited to development intended beneficiaries. Regions of high flow density should be protected or enhanced for positive impact Regions of lower flow Critical ES flow regions density can be developed without impacting ES provision.
    31. 31. EXAMPLES: Subsistence fisheries for coastal communities, Madagascar 1 2 31. total demand for subsistence fisheries2. met demand fraction (from flow analysis)3. unmet demand fraction.The model uses poverty, population density,pollution, habitat suitability and harvest data.Problem areas are immediately visible. 32
    32. 32. EXAMPLES: Veracruz , Mexico: water services Selected results of flow modelingtotal demand from: Stream network, elevation, porosity… Possible (usable) source Agriculture Aquaculture Actual surface water use Industrial Actual flow to beneficiaries Residential used to compute…. Inaccessible water source 33
    33. 33. Novel results from flow analysis Analysis of flows wasn’t available before ARIES and computes source, sink, use and flow profiles. Flow analysis yields crucial maps to assist decision, such as critical flow contours, unmet service demand or unused service production. Quantification is based on flow strength, actual use and provision. Policy scenarios can be analyzed by comparing such contextual information, resulting in more accurate, beneficiary-dependent, science-based estimates of values. Uncertainty is preserved in flow computation and can be visualized.
    34. 34. TARGETING PRECISE AREAS THROUGH FLOWANALYSISapplicable to source, use and sink User draws source area of interest Total value of identified area can be precisely computed; Analysis can also Flows trajectories identify sources identified that supply given users or sinks in their way. Specific use areas computed 35
    35. 35. The “magic number”: new indicators Using information about actual flows, new overall indicators can be computed (with associated uncertainties) for:  EFFICIENCY of provision (actual vs. potential) 0 -1  EFFICIENCY of use (need met or unmet vs. total) 0 -1  EQUITY of distribution (winners and losers) 0 -1  TOTAL: actual use, actual production, unused potential, unmet need Such indicators can be used as good objective functions in scenario analysis.
    36. 36. The public interface: a Web 2.0 application 37
    37. 37. STORYLINES are suggested by ARIES and created byusers to address each benefit deriving from ESServices are broken down into storylinesthat address specific benefits for specificusers.Computed storylines look likepresentations, with explanatory text andsequence, and can be exported to PDFand web pages.The ARIES knowledge base looks at thedata to suggest storylines that match thegeographical and policy context.ARIES case studies develop newstorylines that augment the knowledge A scenario is a user-defined storylinebase in more areas. based on an existing one. 38
    38. 38. Ex-ante scenario definitionGlobal change scenarios can be merged with local land use changesPre-defined GLOBAL SCENARIOSe.g. IPCC climate change SPATIAL CHANGE EDITOR Hand-draw or upload planned intervention, e.g. land conversion forest MODEL PARAMETERS Completed scenarios are saved and compared and THRESHOLDS of RELEVANCE (options, law or governance indications)
    39. 39. Routing linear features (roads,pipelines) baseline reforested Scenario 1: routing that minimizes Scenario 2: routing that impact to flows of ES under minimizes impact on flows of ES business as usual scenario. A long with reforested corridors. Shorter feature is required to avoid feature offsets reforestation impacting water provision. costs.
    40. 40. Identification and ranking of areas for offsetting ARIES can produce a full ES profile for a set of areas under consideration for offsetting, under baseline or ex- C ante intervention scenarios. A Such profiles help selection of areas and documentation of ES offsets. B D CO2 sequestration Agricultural Water Provision Production A B CMultiple Criteria analysis allows D Soil Retention Flood Protectioncustomizing the ES profiles to pre-existing priorities or legal constraints. Aesthetic views
    41. 41. Scenario based quantitative valuationQuantify impact of choices on specific stakeholder groups 350 300 250 200 Residents/Narrow Residents/Wide 150 Farmers/Narrow 100 Farmers/Wide 50 0 Water Flood Aesthetic Soil Agricultural Provision Protection views Retention ProductionTwo alternative options …against the(different buffer zone different needs ofwidths) evaluated forimpact on ecosystem two differentservices… farmers residents stakeholder groups.
    42. 42. VALUATION and ARIES: getting real?ARIES is agnostic about valuation and tries to counteract inaccuracy in the “state of the art” by incorporating: • explicit uncertainty • flexible definition of value • flexibility and innovation in methods • validation opportunities.VALUE can be BASED ON: ACTUAL or POTENTIAL physical flows or source values Concordance value with stakeholder priorities • MCA (Electre3, Prometheus, Evamix) • AHP Economic valuation • Bayesian and Econometric modeling can be easily integrated • Intelligent benefit transfer methods are in development 43
    43. 43. Criticality thresholds and valuation Critical: Perfectly Important: inelastic demand Valuable: elastic demand. inelastic demand Elastic demand: Values change Marginal valuation no Values highly slowly with supply, and change is longer relevant sensitive to supply reversible Marginal ValueMarginal value Sustainability Threshold Josh’s slide with curve, get from Ferd Supply Ecological determines restoration price, e.g. Cap & Price determines supply (green taxes…) auction Natural Capital stocks Natural Capital Stocks
    44. 44. VALUATION in ARIES can be INFORMED BY CRITICALITY Users can set thresholds Marginal Value based on scenarios Natural Capital Stocks Beneficiaries are classified based on criticality of actual Marginal Value provision Provision areas contributing Natural Capital Stocks to different beneficiaryMarginal Value classes can be computed Natural Capital Stocks and valued independently 45
    45. 45. CONCORDANCE VALUE vs. MONETARYabstract valuation = concordance between benefit flows and stakeholderpriorities C F W R Value Climate 1 9 9 9 377 Flood 1 9 3 455 Water 1 2 51 Raw mat 1 51 Relative importance values for benefits are input by users Overall value map is recalculated to reflect stated priorities in each scenario of management Concordance values are the equivalent of value to stakeholders, and vary between 0 (no value) to 1 (complete concordance)
    46. 46. ARIES and economic valuation: asummary  Primary valuation: • WTP surveys, hedonics, travel cost, consumer expenditures, avoided/replacement cost  Secondary valuation: • Value transfer: apply primary values from a database to site of interest  Land use driven  Traditional multiple regression  Bayesian multiple regression  Artificial Intelligence mediated (choice of source studies and transfer function learning)  Flow-based  Ecological-economic modeling • may become part of ARIES as new model content
    47. 47. THANK YOUFor more information: www.ariesonline.org info@ariesonline.org

    ×