Foundational Orientation Case · 11 min
What is a Scientific Model?
A purposeful representation of a system: useful because it is selective, testable, and revisable.
- Position
- Foundation
- Evidence
- 3 public sources
- Boundary
- Introductory orientation
- Status
- Canonical
Human Intention
“I want to understand what a scientific model can show—and why it is not the thing itself.”
This asks how models support scientific understanding. It does not teach one modeling technique or evaluate one disputed model.
Starting Orientation
A model is a simplified version of something scientists want to understand or predict.
Simplification is not merely a flaw: assumptions and purpose determine what a model can usefully represent.
Subject
A scientific model is an explicit representation used to describe, explain, investigate, or predict aspects of a system.
Models can be physical, conceptual, mathematical, computational, or graphical. Each keeps some relationships visible while holding others constant, approximating them, or leaving them outside scope.
A model is judged through use: whether its structure is clear, its assumptions are declared, and its outputs remain open to comparison with observations and other models.
- 01PhenomenonThe system under study
- 02QuestionWhat the model is for
- 03RepresentationVariables, relations, assumptions
- 04OutputExplanation, simulation, prediction
- 05ComparisonObservation tests the fit
Context
- ConceptualMake relationships explicit
A diagram or analogy can organize current understanding and expose questions.
- MathematicalExpress constraints
Equations make selected quantities and relationships precise enough to analyze.
- ComputationalExplore behavior
Simulation can examine consequences when direct observation or calculation is limited.
Relationships
Evidence
Models become trustworthy through explicit assumptions and continued comparison—not resemblance alone.
Each source supports a bounded part of this orientation. Its limitation remains attached.
National Academies of Sciences
Developing and Using Models
- Bears on
- Models represent current understanding, support questions and explanations, and communicate ideas.
- Limit
- An education framework; specialist fields use additional model classes and validation practices.
NOAA National Centers for Environmental Information
Numerical Weather Prediction
- Bears on
- A concrete example of observations entering a model framework to produce forecasts.
- Limit
- One computational domain; it does not represent conceptual, physical, or non-predictive scientific models generally.
NOAA Physical Sciences Laboratory
Modeling and Data Assimilation
- Bears on
- Shows models being calibrated, diagnosed, improved, and combined with observations across time scales.
- Limit
- Operational weather and climate practice, not a universal validation procedure for all sciences.
Assumptions
- 01
The case treats scientific models as a family of representations rather than only equations or simulations.
- 02
Usefulness is evaluated relative to a declared question, domain, scale, and validation practice.
- 03
Agreement with observations can support a model without proving it is uniquely correct.
- 04
Orientation remains distinct from prediction: a model may do one well without doing the other.
Uncertainty
Uncertainty has a location.
Parameter and input uncertainty
Outputs can change when measured inputs, initial conditions, or parameter estimates change.
Structural uncertainty
Another defensible model may organize the same system through different relationships or approximations.
Domain of validity
Performance in one range, place, or time period may not transfer beyond it.
Boundaries
Unknowns
The open edge is part of the map.
Alternative models
Could a different representation fit the same observations while implying another explanation?
Untested range
Where has the model not yet been compared with appropriate observations?
Decision use
Which additional values, risks, or local evidence are required before a model output informs action?
Continuations
Choose a path—or stop here.
Inspect a working model
See observations, a numerical framework, outputs, and known forecast limits together.
Open NOAA weather models (opens in a new tab)Return to evidence
Examine what comparison can support and why uncertainty remains visible.
Continue to evidenceChange the scale
Ask how extent and resolution alter variables, patterns, and domains of validity.
Continue to scaleReflection
What changed?
A scientific model has changed from “a simplified copy” into a purposeful, inspectable representation built around a question.
Its assumptions, outputs, comparisons, and limits now show where it can help—and where orientation must continue beyond it.