Stressors: Order, Duration, Frequency & Intensity — Innovation Summit 2025 (Group 2)

Hayman Fire, June 8, 2002 - land cover change

One sentence on impact: In 3 days we will determine a workflow to evaluate how the order, duration, frequency, and intensity of disturbances affect the likelihood of regime shifts from forests to nonforests.

View shared code · Data & access**

About this site: This public log captures our Innovation Summit sprint. Update it directly in GitHub (open a file → ✏️ → Commit changes) so the homepage always reflects the latest thinking.

How to use this page (for the team)

Edit this file: docs/index.md → ✎ → change text → Commit changes.
Add images: upload to docs/assets/ and reference like assets/your_file.png.
Keep text short and visuals first. Think “slide captions,” not essays.

Day 1 — Define & Explore

Focus: questions, hypotheses, context; add at least one visual (photo of whiteboard/notes).

Our product(s) 📣

Short term:
Identify our response!
Two graphs illustrating the effects of disturbance
Code for processing datasets to 1) identify tipping points from the datasets, and 2) calculate disturbacne severity, order, frequency, duration (SOFD)from datasets.
Graph of Landfire filtered vegetaiton or region of choice
Long term:
Paper! Providing a case study of tipping point drivers in one sample system (forests?)
Grant proposal!

Our question(s) 📣

How does the FODS of fire, drought, and development stressors influence ecosystem recovery windows?
Where do short, intense stress clusters lead to the greatest community or ecological vulnerability?
LT Who needs this information first (agency partners, community groups, funders) to take action?

Hypotheses / intentions [Optional: probably not relevant if you are creating an educational tool]

We think that the spacing between stressors is as important as their intensity for predicting recovery needs.
We intend to test whether compact clusters of high-intensity events correlate with regime shifts? .

Why this matters (the “upshot”) 📣

Colorado resource managers need fast, visual explanations of how multiple stressors overlap. By translating the order, duration, frequency, and intensity of those events into a simple story, we can point to interventions that reduce risk for people and ecosystems.

Field notes / visuals

Whiteboard brainstorm Raw photo location: day1_whiteboard.jpg Caption: Day 1 brainstorm.

Norms Raw photo location: day1_norms.jpg Caption: Caption: Day 1 norms.

Different perspectives: Briefly capture disagreements or alternate framings. These can unlock innovation. [emergency management triggers] - We will know we’re onto something if we can visualize at least two contrasting stressor sequences with clear decision cues.

Day 2 — Data & Methods

Focus: what we’re testing and building; show a first visual (plot/map/screenshot/GIF).

Study Area: Southern Rockies Ecoregion
Spatial extent: a small part of the ecoregion
Spatial resolution: 30x30m
Temporal extent: 1999-2024
Temporal resolution: Annual
Projection: Albers Equal Area Conic projection (EPSG 9822)

Response Variables 1. Was there a regime shift? (Change from forest to nonforest that did not recover within 10 years) 2. If there WAS NOT a regime shift, how long did forests take to recover? 3. If there WAS a regime shift, what system did the forest transition to? (e.g. grassland, shrubs)

Questions: - How does the FODS of fire, drought, insects, and their interaction influence forest shifts to non-forest between 1999 and 2024 in the Southern Rockies ecoregion? - How does the FODS of fire, drought, insects, and their interaction influence forest recovery between 1999 and 2024 in the Southern Rockies ecoregion?

Data sources we’re exploring 📣

Snapshot showing initial data patterns.

Predictors: Frequency, order, duration, severity of
A) Fire, insect outbreaks from LANDFIRE
B) Drought measured as SPEI from PRISM
C) Fire severity from MTBS

Methods / technologies we’re testing 📣

Approach 1 - generalized linear (mixed) models with a binomial response
Approach 2 - generalized linear (mixed) model with a continuous response
Visualization (e.g., map tiles, small multiples)

Challenges identified

We speak different languages - coding, disciplines, chosen technologies
Gaps in the data
Computational constraints - working with big data is hard!
Ensuring models accurately represent our understanding of the system (DAGS)
Scaling up
Identifying final predictors (How do we calculate FODS for each disturbance)
Defining regime shift (Is 10 years long enough?) Need theoretical frameworks!

Visuals

Study Area Subset

Hayman Fire, June 8, 2002 - land cover change Figure 1. Area containing the Hayman Fire showing changes in land cover from 1999 to 2024.

Pixel State Change for Hayman Fire Area (Using NLCD 2019)

Figure 2. Plot of land cover change for a pixel in chosen area.

Interactive map

Interactive exploratory map of Hayman Fire ESIIL 2025 Earth Engine App

Landfire disturbance history

Figure 3. Landfire disturbance.

SPI

Figure 3. Initial SPI from PRISM.

DAG

Figure 4. Initial DAG.

Thought process day 2

Whiteboard Figure 5. Datasets, response, workflow.

Next steps:

Short term
Continue processing the data to arrive at final responses and predictors
Refine the DAG
Run models
Scale to Southern Rockies/USA Moderate term
Submit a WG proposal with Group 3
Include more disturbances Write up our results and submit as a manuscript
Long term
Submit a grant proposal with Group 3

Focus: synthesis; highlight 2–3 visuals that tell the story; keep text crisp. Practice a 2-minute walkthrough of the homepage 📣: Why → Questions → Data/Methods → Findings → Next.

Team photo at start of Day 3 Raw photo location: team_photo.jpg

Findings at a glance 📣

Headline 1 — what, where, how much
Headline 2 — change/trend/contrast
Headline 3 — implication for practice or policy

Visuals that tell the story 📣

Lead conclusion visual placeholder Raw photo location: fire_hull.png Visual 1. Swap in the primary graphic that clearly communicates your core takeaway.

Supporting panels for key insights Raw photo location: hull_panels.png Visual 2. Use a complementary panel, collage, or set of snapshots that reinforces supporting evidence.

Complementary result figure placeholder Raw photo location: main_result.png Visual 3. Highlight an additional visual that captures a secondary insight or next step.

What’s next? 📣

Immediate follow-ups
What we would do with one more week/month
Who should see this next

Featured links (image buttons)

Read the brief

View code

Explore data

Team

Name	Role	Contact	GitHub
Jane Doe	Lead	jane.doe@example.org	@janedoe
John Smith	Analyst	john.smith@example.org	@jsmith

Storage

Code Keep shared scripts, notebooks, and utilities in the code/ directory. Document how to run them in a README or within the files so teammates and visitors can reproduce your workflow.

Documentation Use the docs/ folder to publish project updates on this site. Longer internal notes can live in documentation/; summarize key takeaways here so the public story stays current.

Cite & reuse

If you use these materials, please cite:

Innovation Summit 2025 Group 2 Team. (2025). Stressors: Order, Duration, Frequency & Intensity — Innovation Summit 2025 (Group 2). https://github.com/CU-ESIIL/stressors-order-duration-frequency-intensity-innovation-summit-2025__2

License: CC-BY-4.0 unless noted. See dataset licenses on the Data page.

Challenges identified/NA

Aligning spatial footprints between hydrologic gauges, fire perimeters, and community boundaries.
Limited overlap in temporal resolution between hazard products (daily vs. sub-daily events).
Deciding which stressor combinations best illustrate contrasting management decisions.

Methods / technologies we’re testing 📣 /NA

Sequence analysis of multi-hazard timelines (fire → drought → flood).
Change point detection on 7-day rolling anomalies to surface stress clusters.
Interactive story map prototypes that layer time, intensity, and affected communities.