Synchrony Roadmap

This roadmap records the design intent behind the CubeDynamics synchrony framework. The goal is to support synchrony as coordinated dynamics, not simply as correlation.

Scientific Foundations

The synchrony grammar is intended to bridge several research traditions:

• Ecological synchrony, including the Moran effect, rescue effects, portfolio effects, and community synchrony.
• Climate coherence, teleconnections, spatial covariance, and climate networks.
• Event synchronization and event coincidence analysis.
• Conditional dependence, quantile dependence, and tail statistics.
• Compound, sequential, and state-dependent environmental events.

The manuscript framing should position CubeDynamics as an implementation of an event-based synchrony grammar: raw fields become states, states become event histories, and synchrony operators ask mechanism-specific questions about those histories.

Canonical Data Model

• Raw cubes: continuous environmental or biological values over time and space.
• State cubes: boolean or weighted conditions with state, magnitude, and threshold variables.
• Event catalogs: one row per event with timing, duration, magnitude, recurrence, and links to previous events where available.
• Synchrony outputs: maps, edge tables, block comparisons, matrices, and regional summaries.
• Legacy and response cubes: future support for time-since-event, cumulative exposure, response kernels, and event graphs.

Primitive Operators

The first implementation provides four primitive synchrony verbs:

• Occurrence synchrony: shared event or state occurrence.
• Severity synchrony: shared magnitude during joint occurrence.
• Timing synchrony: similarity in onset, peak, end, or threshold crossing.
• Duration synchrony: similarity in event persistence.

Future primitives may include legacy synchrony, recurrence synchrony, event-sequence synchrony, and graph-based mechanism workflows.

Spatial Modes

The shared spatial comparison engine should continue to support:

• Reference pixel or reference series comparisons.
• Neighborhood synchrony maps.
• All-pairs edge outputs for small problems.
• Regional summaries and synchrony-distance curves.
• Block comparisons and climate-network style analyses.

Relational results should remain relational. All-pairs and network outputs should not be forced into fake (time, y, x) cubes.

QA and Validation Plan

Synchrony operators need diagnostics beyond a single score:

• Event prevalence maps.
• Synchrony-versus-distance curves.
• Local synchrony maps.
• Threshold sensitivity surfaces.
• Joint sample-size diagnostics.
• Bootstrap confidence intervals, preferably with block methods.
• Null models preserving autocorrelation and seasonality.
• Metric comparison panels.
• Matched-event timelines.
• Cross-process lag curves.

Synthetic truth cases should demonstrate:

• Independent fields.
• Shared events.
• Lagged events.
• Shared occurrence with independent severity.
• Shared timing with different duration.
• Distance-decay synchrony.
• Spatial modules.
• Frost followed by biological boom.
• Repeated fire with recurrence effects.
• Compound drought-plus-heat events.

Each primitive should recover only the feature it is designed to measure.

Development Phases

1. State cubes and occurrence synchrony.
2. Severity synchrony and compatibility with the rolling median-split recipe.
3. Event catalogs, timing synchrony, and duration synchrony.
4. Biological cubes and sync_with.
5. Event graphs, legacy effects, sequence analysis, null models, and climate networks.

The current implementation covers the first reviewable pass across phases 1-4. Phase 5 remains intentionally deferred.

Manuscript Path

Possible paper sequence:

1. What is synchrony?
2. Climate synchrony.
3. Primitive synchrony operators.
4. Mechanism-specific synchrony.
5. Complex ecological grammars.

The long-term target is a formal grammar for state-dependent environmental synchrony, implemented as composable CubeDynamics verbs.