Qualitative Modeling of the Haddock Supply Chain

Sarah Hope, MS

2025-04-22

Acknowledgements

Our Collaborators:

- Kate Masury
- Madeline Guyant
- Our Advisory Group
- Gavin Fay

Funding:

- NOAA SK

 

Why use qualitative models?

  • Our systems are complex, and our data are limited.
  • We are able to explore structures of our systems with only information about components and the relationships between them.
  • We can couple these models with quantitative models?

Challenges

  • Finding the Sweet Spot between robustness and functionality
  • Highly connected, or complex systems can grow dramatically by one addition, so simulation is necessary to evaluate

Benefits

  • Investigate feedbacks, links between key system components > Social Systems, Ecosystems, Ecosystem & Human dimensions [ supply chains ]] Markets into Models

Why study NE Seafood Supply Chains?

  • Reveal how the supply chains of different species are linked > Ex. If a change occurs that will make it difficult for fishers to harvest a species, is there another species that will be impacted?

  • Predict impacts of potential scenarios > Changes / What-if / Scenarios / Perturbations; Poor Recruitment Year, Quota Adjustment for Overfishing, a Pandemic, Tariffs, etc.

  • Produce Indicators of Resilience & Identify Vulnerabilities to Supply Chains > Analysis produces ? > Address uncertainty?

  • Impacts on Fishers, Fishing Communities & Food Systems > Understanding socioeconomic impacts on multiple scales > Including input, engagement, discussion with stakeholders about tradeoffs across systems and sectors > Knowledge and collaboration works towards expanding access to seafood.

Building a Model

The Community Matrix → Loop Analysis

# Define the base adjacency matrix for Model A

A <- matrix(c(
  -1, -1,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,
   1, -1,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,
   1,  0, -1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
   0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  1,  0,  0,
   1,  0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  0,  0,
   1,  0,  0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  0,
   0,  0,  0,  0,  1,  1, -1,  1,  0,  0,  0,  0,  0,
   0,  0,  0,  0,  0,  0,  0, -1,  0,  0,  1,  0,  0,
   0,  0,  0,  0,  0,  0,  1,  0, -1,  1,  0,  0,  0,
   0,  0,  0,  0,  0,  0,  1,  0,  0, -1,  0,  0,  0,
  -1,  0,  0,  0,  0,  0,  0, -1,  0,  0, -1,  0,  0,
   0,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0, -1,  0,
   0,  0,  0,  0,  0,  0,  0,  0,  0,  0, -1,  1, -1
), byrow = TRUE, nrow = 13)

B <- matrix(c(
  -1, -1,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,
   1, -1,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,
   1,  0, -1,  0,  0,  0,  0,  0,  0,  0,  0,  0,
   0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  1,  0,
   1,  0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  0,
   1,  0,  0,  0,  0, -1,  0,  0,  0,  0,  0,  0,
   0,  0,  0,  0,  1,  1, -1,  1,  0,  0,  0,  0,
   0,  0,  0,  0,  0,  0,  0, -1,  0,  0,  1,  0,
   0,  0,  0,  0,  0,  0,  1,  0, -1,  1,  0,  0,
   0,  0,  0,  0,  0,  0,  1,  0,  0, -1,  0,  0,
  -1,  0,  0,  0,  0,  0,  0, -1,  0,  0, -1,  0,
   0,  0,  0,  0,  0,  0,  0,  0,  1,  0, -1, -1
), nrow = 12, byrow = TRUE)

Loop Analysis

Model A

Model A Adjoint Matrix

Model A Impact Plot

Model B

Model B

Model Components

  • Each model component is a “node”
  • Connect the “nodes” with “edges” (each w/ a sign, weight)
  • Invert the matrix to reveal the
  • Raymond et al 2011
  • Model Stability

Uncertainty

  • Array of Model Alternatives
  • Well-defined and structurally unique models
  • “super model” approach

Adjacency Matrix

  • Response to change or perturbations
  • Management actions or any change can be considered a press perturbation
  • Loop Analysis
  • What does the matrix reveal about the model?
  • What implications might this have for the analysis?
  • What are the differences between alternatives?

Impact Analysis

  • Why these perturbations?
  • What story does the model tell under this scenario?

Results

  • What do these results mean?
  • Selection & Validation

sarah-hope.github.io