Skip to content

Abrupt vs Gradual Shifts: Rate Factors

✏️

image

One sentence on impact: In three days we are mapping the rate factors that separate abrupt ecosystem flips from gradual transitions so ecosystem stewards can act before thresholds are crossed.

Sprint brief · Explore data · Persistent storage

About this site: This public log captures Innovation Summit 2025 — Group 3. Edit anything in your browser: open a file → pencil icon → Commit changes.

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.
  • Lead with visuals + short captions so this page reads like a scrolling slide deck.

Google Doc for team members

Day 1 — Define & Explore

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

Bridges to group 2

  • System-specific case study illustrating regime shift types, or counterfactuals of when shift was expected but didn't occur
  • Geospatial or time series datasets, adjusting predictor variables

Potential roles

  • ESIIL working group PI, co-PIs

Our product 📣

  • Conceptual figure containing our goals/hypotheses, leading to a perspective manuscript
  • List of terms and definitions, shared language with group #2
  • List of keyword terms for literature search (helpful in structuring meta-meta-analysis)
  • Team of teams of teams pizza3 party
  • A threshold decision chart highlighting rate factor triggers for abrupt vs gradual responses.
  • A concise narrative brief + two figures for Innovation Summit share-out.

Our question(s) 📣

  • What are the properties of stress and disturbance that lead to regime shifts?
  • What are the important predictors: rate (abrupt vs. gradual), intensity/severity, frequency/recurrence, duration (pulse vs. press), spatial extent, abiotic vs. biotic drivers?
  • What are the important responses: type of tipping (noise-induced, rate-induced, bifurcation-induced), resilience (resistance + recovery), community structure and function?
  • What is our definition of tipping point, regime shift, state change, critical transition, ecological transformation, etc.?
  • What is not within scope (e.g., succession, community assembly rules)?
  • Which climatic and ecological rate factors precede abrupt state changes compared with gradual drifts?
  • Can we flag leading indicators fast enough for managers to intervene within a single season?
  • How transferable are the signals between organisms with different life history traits?

Hypotheses / intentions

  • We think abrupt shifts are preceded by compound rate anomalies (e.g., concurrent moisture and temperature acceleration).
  • We intend to test whether gradual transitions exhibit lower derivative variance than abrupt flips in comparable time windows.
  • We will know we’re onto something if we can classify historical events with >75% accuracy using rate-derived features alone.

Why this matters (the “upshot”)

Rapid detection of tipping dynamics lets watershed groups and land managers deploy scarce mitigation resources before ecosystems cross points of no return. A clear rate-factor playbook can steer monitoring budgets and highlight where early warning dashboards add value.

Inspirations (papers, datasets, tools)

Field notes / visuals

Whiteboard sketch of rate-factor hypotheses Raw photo location: day1_whiteboard.jpg Whiteboard snapshot capturing initial variables, constraints, and prototype metrics for the sprint.

Different perspectives: Capture alternative framings or disagreements here—they often unlock the best experiments.

Day 2 — Data & Methods

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

Agenda

  • Create schedule
  • Skim Milkoreit et al. 2018
  • Define key terms
  • Identify critera for inclusion in the study

Data sources we’re exploring 📣

  • Source A
  • Source B — link and 1-line description

Methods / technologies we’re testing 📣

  • Approach 1 (e.g., time-series break detection)
  • Approach 2 (e.g., random forest on features)
  • Visualization (e.g., map tiles, small multiples)

Challenges identified

  • Data gaps / quality issues
  • Method limitations / compute constraints
  • Open questions we need to decide on

Data sources we’re exploring 📣

  • LandTrendr disturbance trajectories — Tracking vegetation change rates across western U.S. watersheds.

LandTrendr-derived slope changes across sample watersheds Raw photo location: explore_data_plot.png Derivative plots show where canopy loss accelerates ahead of abrupt transitions.

  • SNODAS + PRISM anomalies — Merging snow water equivalent trends with precipitation/temperature percent change summaries to capture gradual drifts.

Methods / technologies we’re testing 📣

  • Bayesian online change-point detection on rolling derivatives.
  • Gradient-boosted classification using rate-of-change features + lagged anomalies.
  • Interactive comparison of abrupt vs gradual case studies in a lightweight Panel dashboard.

Challenges identified

  • Aligning spatial resolution between remote sensing products and point-based climate grids.
  • Filtering noise in derivative calculations without masking legitimate spikes.
  • Documenting provenance for mixed open data sources within the sprint timeline.

Visuals

Static figure

Prototype comparison of abrupt vs gradual rate fingerprints Raw photo location: figure1.png Figure 1. One line on what this suggests.

Animated change (GIF)

Animation of rolling derivative anomalies through time Raw photo location: change.gif Figure 2. One line on what changes across time.

Interactive map (iframe)

Open full map

If an embed doesn’t load, drop the direct link underneath it.

Final Share Out — Insights & Sharing

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 📣

Rate-trigger matrix summarizing abrupt vs gradual signatures Raw photo location: fire_hull.png Visual 1. Swap in the primary graphic that clearly communicates your core takeaway.

Panel comparing contrasting watershed responses Raw photo location: hull_panels.png Visual 2. Use a complementary panel, collage, or set of snapshots that reinforces supporting evidence.

Rolling derivative dashboard mock-up 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
Sprint brief PDF
Open sprint brief 		Explore PRISM quicklook script
Explore data script 		Open Group 3 shared storage
Open storage

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 3. (2025). Abrupt vs Gradual Shifts: Rate Factors. https://github.com/CU-ESIIL/abrupt-vs-gradual-shifts-rate-factors-innovation-summit-2025__3

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