Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
16 views13 pages

Environmental Models in Spatial Science

Uploaded by

kashyap jyoti gohain

There are different types of environmental models that can be categorized based on their underlying logic and processing methods. Deductive models draw conclusions from general premises and assumptions, while inductive models derive conclusions from collected facts and data. Deterministic models have fixed inputs and outputs, whereas stochastic models have random inputs and outputs. Common environmental model types include deductive models based on physical laws, inductive models developed through data analysis, deterministic empirical models, knowledge-driven models using expert systems, and process-driven conceptual models simulating underlying systems. Neural networks are an example of a stochastic model that uses a backpropagation algorithm to connect input variables to outputs through multiple network layers.

Copyright:

© All Rights Reserved
Download as PPTX, PDF, TXT or read online from Scribd
Download as pptx, pdf, or txt

Environmental Models in Spatial Science

Uploaded by

kashyap jyoti gohain
16 views13 pages
There are different types of environmental models that can be categorized based on their underlying logic and processing methods. Deductive models draw conclusions from general premises and assumptions, while inductive models derive conclusions from collected facts and data. Deterministic models have fixed inputs and outputs, whereas stochastic models have random inputs and outputs. Common environmental model types include deductive models based on physical laws, inductive models developed through data analysis, deterministic empirical models, knowledge-driven models using expert systems, and process-driven conceptual models simulating underlying systems. Neural networks are an example of a stochastic model that uses a backpropagation algorithm to connect input variables to outputs through multiple network layers.

Original Title

6

Copyright

© © All Rights Reserved

Available Formats

PPTX, PDF, TXT or read online from Scribd

Did you find this document useful?

Is this content inappropriate?

There are different types of environmental models that can be categorized based on their underlying logic and processing methods. Deductive models draw conclusions from general premises and assumptions, while inductive models derive conclusions from collected facts and data. Deterministic models have fixed inputs and outputs, whereas stochastic models have random inputs and outputs. Common environmental model types include deductive models based on physical laws, inductive models developed through data analysis, deterministic empirical models, knowledge-driven models using expert systems, and process-driven conceptual models simulating underlying systems. Neural networks are an example of a stochastic model that uses a backpropagation algorithm to connect input variables to outputs through multiple network layers.

Copyright:

© All Rights Reserved
Download as PPTX, PDF, TXT or read online from Scribd
Download as pptx, pdf, or txt
16 views13 pages

Environmental Models in Spatial Science

Uploaded by

kashyap jyoti gohain
There are different types of environmental models that can be categorized based on their underlying logic and processing methods. Deductive models draw conclusions from general premises and assumptions, while inductive models derive conclusions from collected facts and data. Deterministic models have fixed inputs and outputs, whereas stochastic models have random inputs and outputs. Common environmental model types include deductive models based on physical laws, inductive models developed through data analysis, deterministic empirical models, knowledge-driven models using expert systems, and process-driven conceptual models simulating underlying systems. Neural networks are an example of a stochastic model that uses a backpropagation algorithm to connect input variables to outputs through multiple network layers.

Copyright:

© All Rights Reserved
Download as PPTX, PDF, TXT or read online from Scribd
Download as pptx, pdf, or txt
You are on page 1/ 13

Environmental Models in Spatial

Science
Module 2

Reference text: Environmental Modeling with GIS and


Remote Sensing, Edt. By Andrew Skidmore

1
Environmental Models and their
types
 Environmental model
◦ Simulate the functioning of environmental
processes
◦ Better understanding of complexity
◦ Scale
 Local
 Regional
 Global
◦ GIS = allows change of scale, extrapolation,
etc.
2
Model types (Taxonomy of the
models)
Environmental
Model

Models based on
Models of logic
processing method

Inductive Deductive Deterministic Stochastic

A model has to be a combination of logic and process


3
Deductive models
 Draw conclusion from a set general
proposition (premises).
 One can not deny the conclusion if the
premises are assumed true. (Self-
contradiction)
 Example of deductive models
◦ Pythagoras’ theorem (assumes lines as true
identity)

4
◦ Three types of vegetation
 Grass
 Herb
 Shrub
◦ Rainfall and grazing intensity - control factors
◦ Using assumptions/literature
 Combination of grazing intensity and rainfall
amount will determine the change of one
vegetation category to other
 Deductive models are based on plausible
physical laws.
5
Inductive models
 Derive conclusion from facts
 A series of facts may be used to prove a
general statement
 May lead to understanding of patterns
 Most of the times, conclusion and the
factors contributing to it are known well.
The causal process may not be
understood.

6
 Steps in inductive models
◦ Define the problem and formulate a research
questions out of it
◦ Develop a hypothesis
◦ Collect data
◦ Exploratory analysis
◦ Confirmatory analysis - accept/reject
hypothesis

7
 In reality, a model can also be a mixed one
◦ Inductive and deductive
◦ Example, Global Climate Model
 Inductive
 Components of models depicting interaction between
atmosphere and biosphere
 Deductive
 Measurement of components such as Absorbed
Photosynthetically Active Radiation (APAR), Day length,
interpolated weather data, etc.

8
Deterministic models
 Empirical models
◦ A fixed input and fixed output
◦ Hence, empirical environmental models are
site-specific
 Example,
◦ Universal Soil Loss Equation
 Regression based
 Has multiple criteria to estimate soil loss
 Developed empirically using plot data

9
 Knowledge driven models
◦ Dependent and independent variables
◦ Establishment of rules to depict relationships
◦ Expert’s opinion used for formulating rules
 Combination of observations and expert’s
conclusion
◦ Expert system has:
 Knowledge base (rules)
 A method for processing those rules
 User interface to handle spatial data and interpret
results
 For example, land use land cover classification using
expert system

10
 Process Driven Models
◦ Also known as conceptual models
◦ White box (process underlying is understood)
◦ Process based models describe process based on
understanding and established concepts
 Many types
 Lumped
 Distributed, etc.
 Example, Hortonian Overland Flow model
 Q = (I-F)A
 Where, Q is overland flow, I rainfall intensity, F is infiltration
rate, and A is catchment area

11
Stochastic Models
 Input data parameters are variable/random
 Therefore, the output is also variable
 Uses back propagation method (BP)
 Multiple layers to connect input
data/variables to output
◦ BP works first in the forward direction to
calculate output
◦ Then in backward direction to compare the
calculated values to the target
◦ Errors are estimated to readjust the connectivity
◦ Example, Neural Network for digital classification
of satellite data

12
Structure of Neural Network

13

You might also like