EcoService Models Library (ESML)
Home
Search EMs
My
EMs 
Learn about
ESML
View ESML Data Maploading
Compare EMs
Which comparison is best for me?EM Variables by Variable Role
One quick way to compare ecological models (EMs) is by comparing their variables. Predictor variables show what kinds of influences a model is able to account for, and what kinds of data it requires. Response variables show what information a model is capable of estimating.
This first comparison shows the names (and units) of each EM’s variables, side-by-side, sorted by variable role. Variable roles in ESML are as follows:
- Predictor Variables
- Time- or Space-Varying Variables
- Constants and Parameters
- Intermediate (Computed) Variables
- Response Variables
- Computed Response Variables
- Measured Response Variables
EM Variables by Category
A second way to use variables to compare EMs is by focusing on the kind of information each variable represents. The top-level categories in the ESML Variable Classification Hierarchy are as follows:
- Policy Regarding Use or Management of Ecosystem Resources
- Land Surface (or Water Body Bed) Cover, Use or Substrate
- Human Demographic Data
- Human-Produced Stressor or Enhancer of Ecosystem Goods and Services Production
- Ecosystem Attributes and Potential Supply of Ecosystem Goods and Services
- Non-monetary Indicators of Human Demand, Use or Benefit of Ecosystem Goods and Services
- Monetary Values
Besides understanding model similarities, sorting the variables for each EM by these 7 categories makes it easier to see if the compared models can be linked using similar variables. For example, if one model estimates an ecosystem attribute (in Category 5), such as water clarity, as a response variable, and a second model uses a similar attribute (also in Category 5) as a predictor of recreational use, the two models can potentially be used in tandem. This comparison makes it easier to spot potential model linkages.
All EM Descriptors
This selection allows a more detailed comparison of EMs by model characteristics other than their variables. The 50-or-so EM descriptors for each model are presented, side-by-side, in the following categories:
- EM Identity and Description
- EM Modeling Approach
- EM Locations, Environments, Ecology
- EM Ecosystem Goods and Services (EGS) potentially modeled, by classification system
EM Descriptors by Modeling Concepts
This feature guides the user through the use of the following seven concepts for comparing and selecting EMs:
- Conceptual Model
- Modeling Objective
- Modeling Context
- Potential for Model Linkage
- Feasibility of Model Use
- Model Certainty
- Model Structural Information
Though presented separately, these concepts are interdependent, and information presented under one concept may have relevance to other concepts as well.
Compare Criteria:
Show Criteria
Hide CriteriaEM Identity and Description
EM Identification
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Short Name
em.detail.shortNameHelp
?
|
SAV occurrence, St. Louis River, MN/WI, USA | Chinook salmon value (household), Yaquina Bay, OR | ESII Tool method | Specific conductivity, USA |
|
EM Full Name
em.detail.fullNameHelp
?
|
Predicting submerged aquatic vegetation occurrence, St. Louis River Estuary, MN & WI, USA | Economic value of Chinook salmon per household method, Yaquina Bay, OR | ESII (Ecosystem Services Identification & Inventory) Tool method | Specific Conductivity, USA |
|
EM Source or Collection
em.detail.emSourceOrCollectionHelp
?
|
US EPA | US EPA | None | US EPA |
|
EM Source Document ID
|
330 | 324 |
391 ?Comment:Website for online user support |
460 |
|
Document Author
em.detail.documentAuthorHelp
?
|
Ted R. Angradi, Mark S. Pearson, David W. Bolgrien, Brent J. Bellinger, Matthew A. Starry, Carol Reschke | Stephen J. Jordan, Timothy O'Higgins and John A. Dittmar | EcoMetrix Solutions Group (ESG) | Olson, J.R., and S.M. Cormier |
|
Document Year
em.detail.documentYearHelp
?
|
2013 | 2012 | 2016 | 2019 |
|
Document Title
em.detail.sourceIdHelp
?
|
Predicting submerged aquatic vegetation cover and occurrence in a Lake Superior estuary | Ecosystem Services of Coastal Habitats and Fisheries: Multiscale Ecological and Economic Models in Support of Ecosystem-Based Management | ESII Tool | Modeling Spatial and Temporal Variation in Natural Background Specific Conductivity |
|
Document Status
em.detail.statusCategoryHelp
?
|
Peer reviewed and published | Peer reviewed and published | Other or unclear (explain in Comment) | Peer reviewed and published |
|
Comments on Status
em.detail.commentsOnStatusHelp
?
|
Published journal manuscript | Published journal manuscript | Website | Published journal manuscript |
Software and Access
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
| Not applicable | Not applicable | https://www.esiitool.com/ | (https://edg.epa.gov/ metadata/catalog/main/home.page) | |
|
Contact Name
em.detail.contactNameHelp
?
|
Ted R. Angradi | Stephen Jordan | Not reported | John Olson |
|
Contact Address
|
U.S. Environmental Protection Agency, Office of Research and Development, National Health and Environmental Effects Research Laboratory, Mid-Continent Ecology Division, 6201 Congdon Blvd., Duluth, MN 55804, USA | U.S. EPA, Gulf Ecology Div., 1 Sabine Island Dr., Gulf Breeze, FL 32561, USA | Not reported | California State Univ. Monterey Bay, 100 Campus Center, Seaside CA 93955 |
|
Contact Email
|
angradi.theodore@epa.gov | jordan.steve@epa.gov | Not reported | joolson@csumb.edu |
EM Description
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Summary Description
em.detail.summaryDescriptionHelp
?
|
ABSTRACT: “Submerged aquatic vegetation (SAV) provides the biophysical basis for multiple ecosystem services in Great Lakes estuaries. Understanding sources of variation in SAV is necessary for sustainable management of SAV habitat. From data collected using hydroacoustic survey methods, we created predictive models for SAV in the St. Louis River Estuary (SLRE) of western Lake Superior. The dominant SAV species in most areas of the estuary was American wild celery (Vallisneria americana Michx.)…” AUTHOR’S DESCRIPTION: “The SLRE is a Great Lakes “rivermouth” ecosystem as defined by Larson et al. (2013). The 5000-ha estuary forms a section of the state border between Duluth, Minnesota and Superior, Wisconsin…In the SLRE, SAV beds are often patchy, turbidity varies considerably among areas (DeVore, 1978) and over time, and the growing season is short. Given these conditions, hydroacoustic survey methods were the best option for generating the extensive, high resolution data needed for modeling. From late July through mid September in 2011, we surveyed SAV in Allouez Bay, part of Superior Bay, eastern half of St. Louis Bay, and Spirit Lake…We used the measured SAV percent cover at the location immediately previous to each useable record location along each transect as a lag variable to correct for possible serial autocorrelation of model error. SAV percent cover, substrate parameters, corrected depth, and exposure and bed slope data were combined in Arc-GIS...We created logistic regression models for each area of the SLRE to predict the probability of SAV being present at each report location. We created models for the training data set using the Logistic procedure in SAS v.9.1 with step wise elimination (?=0.05). Plots of cover by depth for selected predictor values (Supplementary Information Appendix C) suggested that interactions between depth and other predictors were likely to be significant, and so were included in regression models. We retained the main effect if their interaction terms were significant in the model. We examined the performance of the models using the area under the receiver operating characteristic (AUROC) curve. AUROC is the probability of concordance between random pairs of observations and ranges from 0.5 to 1 (Gönen, 2006). We cross-validated logistic occurrence models for their ability to classify correctly locations in the validation (holdout) dataset and in the Superior Bay dataset… Model performance, as indicated by the area under the receiver operating characteristic (AUROC) curve was >0.8 (Table 3). Assessed accuracy of models (the percent of records where the predicted probability of occurrence and actual SAV presence or absence agreed) for split datasets was 79% for Allouez Bay, 86% for St. Louis Bay, and 78% for Spirit Lake." | ABSTRACT:"Critical habitats for fish and wildlife are often small patches in landscapes, e.g., aquatic vegetation beds, reefs, isolated ponds and wetlands, remnant old-growth forests, etc., yet the same animal populations that depend on these patches for reproduction or survival can be extensive, ranging over large regions, even continents or major ocean basins. Whereas the ecological production functions that support these populations can be measured only at fine geographic scales and over brief periods of time, the ecosystem services (benefits that ecosystems convey to humans by supporting food production, water and air purification, recreational, esthetic, and cultural amenities, etc.) are delivered over extensive scales of space and time. These scale mismatches are particularly important for quantifying the economic values of ecosystem services. Examples can be seen in fish, shellfish, game, and bird populations. Moreover, there can be wide-scale mismatches in management regimes, e.g., coastal fisheries management versus habitat management in the coastal zone. We present concepts and case studies linking the production functions (contributions to recruitment) of critical habitats to commercial and recreational fishery values by combining site specific research data with spatial analysis and population models. We present examples illustrating various spatial scales of analysis, with indicators of economic value, for recreational Chinook (Oncorhynchus tshawytscha) salmon fisheries in the U.S. Pacific Northwest (Washington and Oregon) and commercial blue crab (Callinectes sapidus) and penaeid shrimp fisheries in the Gulf of Mexico. | AUTHORS DESCRIPTION: "The Nature Conservancy (TNC) and The Dow Chemical Company (Dow) initiated a collaborative effort to develop models that would help Dow and the wider business community identify and incorporate the value of nature into business decision making…the ESII Tool models and outputs were constructed and tested with an engineering and design perspective to facilitate actionable land use and management decisions. The ESII Tool helps non-ecologists make relative comparisons of the expected levels of ecosystem service performance across a given site, under a variety of conditions. As a planning-level tool, it can inform business decisions while enhancing the user’s relationship with nature. However, other uses that require ecological models of a higher degree of accuracy and/or precision, expert data collection, extensive sampling, and analysis of ecological relationships are beyond the intended scope of this tool." "The ESII App is your remote interface to the ESII Tool. It enables you to collect spatially-explicit ecological data, make maps, collect survey data, take photos, and record notes about your observations. With a Wi-Fi connection, the ESII App can upload and download information stored on the ESII Project Workspace, where final analyses and reports are generated. Because sites may be large and may include several different types of habitats, each Site to be assessed using the ESII Tool is divided into smaller areas called map units, and field data is collected on a map unit basis." "Once a map unit has been selected from the list of map units, the first survey question will always be “Map Unit Habitat Type” (Figure 12). The survey will progress through four categories of questions: habitat, vegetation, surface characteristics, and noise and visual screening. The questions are designed to enable you to select the most appropriate response easily and quickly." "Ecosystem Functions and Services scores are shown in units of percent performance, while each Units of Measure score will be shown in the engineering units appropriate to each attribute. At a map unit level, percent performance predicts how well a map unit would perform a given function or service as a proportion of the maximum potential you would expect from ideal attribute conditions. At a Site or Scenario level, percent performance is calculated as the area weighted average of the individual map unit’s percent performance values; it provides a normalized comparative metric between Sites or Scenarios. At both the map unit and the Site or Scenario levels, the units of measure represent absolute values (such as gallons of runoff or BTU reduction through shading) and can be either summed to show absolute performance of a Scenario, or normalized by area to show area-based rates of performance." | We developed a random forest model that predicts natural background specific conductivity (SC), a measure of total dissolved ions, for all stream segments in the contiguous United States at monthly time steps between the years 2001 to 2015. Models were trained using 11 796 observations made at 1785 minimally impaired stream segments and validated with observations from an additional 92 segments. Static predictors of SC included geology, soils, and vegetation parameters. Temporal predictors were related to climate and enabled the model to make predictions for different dates. The model explained 95% of the variation in SC among validation observations (mean absolute error = 29 μS/cm, Nash-Sutcliffe efficiency = 0.85). The model performed well across the period of interest but exhibited bias in Coastal Plain and Xeric regions (26 and 30%, respectively). National model predictions showed large spatial variation with the greatest SC predicted to occur in the desert southwest and plains. Model predictions also reflected changes at individual streams during drought. |
|
Specific Policy or Decision Context Cited
em.detail.policyDecisionContextHelp
?
|
None identified | None identified | None identified | N/A |
|
Biophysical Context
|
submerged aquatic vegetation | Yaquina Bay estuary | Not applicable | Stream segment taken from StreamCat database |
|
EM Scenario Drivers
em.detail.scenarioDriverHelp
?
|
No scenarios presented | No scenarios presented | No scenarios presented | N/A |
EM Relationship to Other EMs or Applications
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Method Only, Application of Method or Model Run
em.detail.methodOrAppHelp
?
|
Method + Application | Method + Application | Method Only | Method + Application |
|
New or Pre-existing EM?
em.detail.newOrExistHelp
?
|
New or revised model | New or revised model | New or revised model | New or revised model |
Related EMs (for example, other versions or derivations of this EM) described in ESML
em.detail.relatedEmHelp
?
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Document ID for related EM
em.detail.relatedEmDocumentIdHelp
?
|
None | Doc-324 | None | None |
|
EM ID for related EM
em.detail.relatedEmEmIdHelp
?
|
None | EM-603 | EM-397 | EM-713 | None |
EM Modeling Approach
EM Relationship to Time
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Temporal Extent
em.detail.tempExtentHelp
?
|
2010 - 2012 | 2003-2008 | Not applicable | 2001-2015 |
|
EM Time Dependence
em.detail.timeDependencyHelp
?
|
time-stationary | time-stationary | time-stationary | time-dependent |
|
EM Time Reference (Future/Past)
em.detail.futurePastHelp
?
|
Not applicable | Not applicable | Not applicable | past time |
|
EM Time Continuity
em.detail.continueDiscreteHelp
?
|
Not applicable | Not applicable | Not applicable | discrete |
|
EM Temporal Grain Size Value
em.detail.tempGrainSizeHelp
?
|
Not applicable | Not applicable | Not applicable | 3 |
|
EM Temporal Grain Size Unit
em.detail.tempGrainSizeUnitHelp
?
|
Not applicable | Not applicable | Not applicable | Month |
EM Spatial Extent
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Bounding Type
em.detail.boundingTypeHelp
?
|
Physiographic or ecological | Geopolitical | Not applicable | Geopolitical |
|
Spatial Extent Name
em.detail.extentNameHelp
?
|
St. Louis River Estuary | Pacific Northwest | Not applicable | Contiguous United States |
|
Spatial Extent Area (Magnitude)
em.detail.extentAreaHelp
?
|
10-100 km^2 | >1,000,000 km^2 | Not applicable | >1,000,000 km^2 |
Spatial Distribution of Computations
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Spatial Distribution
em.detail.distributeLumpHelp
?
|
spatially distributed (in at least some cases) ?Comment:BH: Each individual transect?s data was parceled into location reports, and that each report?s ?quadrat? area was dependent upon the angle of the hydroacoustic sampling beam. The spatial grain is 0.07 m^2, 0.20 m^2 and 0.70 m^2 for depths of 1 meter, 2 meters and 3 meters, respectively. |
spatially lumped (in all cases) |
spatially distributed (in at least some cases) ?Comment:map units delineated by user based on project. |
spatially distributed (in at least some cases) |
|
Spatial Grain Type
em.detail.spGrainTypeHelp
?
|
area, for pixel or radial feature | Not applicable | other (specify), for irregular (e.g., stream reach, lake basin) | area, for pixel or radial feature |
|
Spatial Grain Size
em.detail.spGrainSizeHelp
?
|
0.07 m^2 to 0.70 m^2 | Not applicable | map units | 3.1 km2 |
EM Structure and Computation Approach
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Computational Approach
em.detail.emComputationalApproachHelp
?
|
Analytic | Analytic | Analytic | Analytic |
|
EM Determinism
em.detail.deterStochHelp
?
|
deterministic | deterministic | deterministic | deterministic |
|
Statistical Estimation of EM
em.detail.statisticalEstimationHelp
?
|
|
|
|
|
Model Checking Procedures Used
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Model Calibration Reported?
em.detail.calibrationHelp
?
|
Yes | No | Not applicable | Yes |
|
Model Goodness of Fit Reported?
em.detail.goodnessFitHelp
?
|
Yes | No | Not applicable | Yes |
|
Goodness of Fit (metric| value | unit)
em.detail.goodnessFitValuesHelp
?
|
|
None | None |
|
|
Model Operational Validation Reported?
em.detail.validationHelp
?
|
Yes | Yes | Not applicable | Yes |
|
Model Uncertainty Analysis Reported?
em.detail.uncertaintyAnalysisHelp
?
|
No | No | Not applicable | No |
|
Model Sensitivity Analysis Reported?
em.detail.sensAnalysisHelp
?
|
No | No | Not applicable | Yes |
|
Model Sensitivity Analysis Include Interactions?
em.detail.interactionConsiderHelp
?
|
Not applicable | Not applicable | Not applicable | Yes |
EM Locations, Environments, Ecology
Location of EM Application
Terrestrial location (Classification hierarchy: Continent > Country > U.S. State [United States only])
em.detail.relationToSpaceTerrestrialHelp
?
| EM-414 | EM-604 | EM-712 | EM-982 |
|
|
None |
|
Marine location (Classification hierarchy: Realm > Region > Province > Ecoregion)
em.detail.relationToSpaceMarineHelp
?
| EM-414 | EM-604 | EM-712 | EM-982 |
| None |
|
None | None |
Centroid Lat/Long (Decimal Degree)
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
Centroid Latitude
em.detail.ddLatHelp
?
|
46.72 | 44.62 | Not applicable | 39.83 |
|
Centroid Longitude
em.detail.ddLongHelp
?
|
-96.13 | -124.02 | Not applicable | 98.58 |
|
Centroid Datum
em.detail.datumHelp
?
|
WGS84 | WGS84 | Not applicable | WGS84 |
|
Centroid Coordinates Status
em.detail.coordinateStatusHelp
?
|
Estimated | Estimated | Not applicable | Estimated |
Environments and Scales Modeled
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Environmental Sub-Class
em.detail.emEnvironmentalSubclassHelp
?
|
Aquatic Environment (sub-classes not fully specified) | Rivers and Streams | Inland Wetlands | Lakes and Ponds | Near Coastal Marine and Estuarine | Terrestrial Environment (sub-classes not fully specified) | Rivers and Streams | Inland Wetlands | Lakes and Ponds | Terrestrial Environment (sub-classes not fully specified) | Rivers and Streams |
|
Specific Environment Type
em.detail.specificEnvTypeHelp
?
|
Freshwater estuarine system | Yaquina Bay estuary and ocean | Not applicable | Stream segment |
|
EM Ecological Scale
em.detail.ecoScaleHelp
?
|
Ecological scale corresponds to the Environmental Sub-class | Ecological scale is finer than that of the Environmental Sub-class | Not applicable | Ecological scale is finer than that of the Environmental Sub-class |
Scale and taxa of organisms modeled
Scale of differentiation of organisms modeled
em.detail.nameOfOrgsOrGroupsHelp
?
|
EM ID
em.detail.idHelp
?
|
EM-414 | EM-604 | EM-712 | EM-982 |
|
EM Organismal Scale
em.detail.orgScaleHelp
?
|
Not applicable | Other (multiple scales) | Not applicable | Not applicable |
Taxonomic level and name of organisms or groups identified
taxonomyHelp
?
| EM-414 | EM-604 | EM-712 | EM-982 |
| None Available |
|
None Available | None Available |
EnviroAtlas URL
EM Ecosystem Goods and Services (EGS) potentially modeled, by classification system
CICES v 4.3 - Common International Classification of Ecosystem Services (Section > Division > Group > Class)
em.detail.cicesHelp
?
| EM-414 | EM-604 | EM-712 | EM-982 |
|
|
|
|
<a target="_blank" rel="noopener noreferrer" href="https://www.epa.gov/eco-research/national-ecosystem-services-classification-system-nescs-plus">National Ecosystem Services Classification System (NESCS) Plus</a>
fegs1Help
?
(Environmental Subclass > Ecological End-Product (EEP) > EEP Subclass > EEP Modifier)
fegs2Help
?
| EM-414 | EM-604 | EM-712 | EM-982 |
| None |
|
|
|