Field learning · local only

Field Lab

Turn a fishing day into evidence without pretending a small personal record is a forecast. Run one-variable experiments, keep a three-step field queue, replay the day, inspect the gaps in your own record, and separate source fact from application inference.

Trips, sessions, experiments, notes and debriefs stay in this browser. Habitat plates are physical archetypes, not locations. A hard stop still outranks every experiment.

On-water cognitive load

NOW / NEXT / IF NEEDED

The queue is derived from this session’s observations and the existing change tree. It does not create a second recommendation engine.

NOW

Start the on-water session

The queue only changes from observations made during this trip.

Open On the Water from a named trip and log what is actually happening.

Log the first useful signal

A rise form, follow, refusal, bottom contact, light change, water change, or temperature gives the desk something real to reason from.

Record the signal before changing the presentation.

IF NEEDED

Change one variable

Multiple simultaneous changes destroy the evidence the session is meant to preserve.

State the reason and the result that would count before making the change.

Experiment Engine

One variable. A declared test window.

State the observation, change one axis, name what stays still, and decide what result would count before the test starts. The engine records an outcome; it never labels a lure, fly or water “successful.”

Completed field experiments · 0

No completed experiments on this device yet.

Plan vs Reality

What the brief expected. What the day required.

Departing from the plan is not failure. The useful question is whether the same conditions keep making you leave the same starting assumption.

No debrief on this device yet contains both a planned presentation and what was actually fished.

Trip Replay

The day on a real minute axis

Changes divide the day into before and after. Responses sit where they actually happened. Quiet time stays quiet instead of being compressed into a tidy story.

No on-water session exists on this device yet. A replay is reconstructed from the session record, not invented from a debrief.

Personal Coverage Maps

The holes in your own record

Empty cells mean “not recorded,” not “bad.” One or two records stay thin. The board refuses to turn a handful of outings into a success percentage.

Coverage

Three records and this grid starts working

Coverage crosses two things at once — water colour against presentation, temperature against presentation — and shows you the squares you have never fished. On one or two records every square is empty, which is true and useless. Come back after a few debriefs.

Stillwater reasoning

Read the lake as layers and boundaries

Turnover / mixing transition

Cooling and wind are breaking down the summer layers. Temperature becomes more uniform and formerly stable depth relationships can disappear quickly.

Depth question: Has the lake actually mixed, or are remnants of the old layer still visible in temperature and forage observations?

Caution: Turnover timing varies by lake depth, elevation and weather. Treat calendar timing as a prompt to inspect, not a diagnosis.

Inspect · Windward bank

Look for accumulated plankton, insects, stained water, wave-created cover and bait pushed shallow.

Castability and boat control may be the limiting variables before presentation choice.

Inspect · First break

Look for a hard-bottom transition, scattered cover, bait concentration and fish that move between shallow and deep water.

A presentation that crosses the contour tells you more than one held entirely above or below it.

Inspect · Secondary break

Use when shallow water is physically less usable because of light, temperature, disturbance or season.

Depth control becomes the experiment; know where the presentation is in the column.

Inspect · Inlet

Look for plume boundaries, colder or warmer water, turbidity, oxygen and concentrated drifting food.

Read it as moving water nested inside stillwater; current angle matters again.

Habitat archetypes + plates

48 physical shapes of water

48 archetypes · no pins

moving water · habitat class

Current seam

A physical archetype, not a map or a hotspot. Use the drawing to identify geometry and decide what to inspect before choosing a presentation.

Field read

What creates it

Velocity difference around depth, bank shape, structure, or converging flows.

Inspect

Foam line, drifting food, fish rising on one side, and a repeatable lane rather than random surface texture.

Approach

Present along the boundary before crossing it.

Constraint

Drag is usually the first failure; standing position controls it.

Deeper species behavior

Behavior as field questions, not a bite forecast

Choose a species in Trip Prep or start an on-water session with a species to open the behavior layer.

Field Brief 2.0

The trip, the queue and the fallback on one artifact

Hook the Horizon · Field Ops Desk

Field Brief 2.0 — Named water not yet selected

Open target · local-only trip artifact

Start here

Water: Named water not yet selected
Target: not named
Primary: not locked
Holding water: not locked
Depth: not locked

Verify before fishing

Regulations still need review
Access still needs confirmation
Conditions have not been reviewed
Hydrology has not been reviewed

NOW / NEXT / IF NEEDED

NOW: Start the on-water session — Open On the Water from a named trip and log what is actually happening.
NEXT: Log the first useful signal — Record the signal before changing the presentation.
IF NEEDED: Change one variable — State the reason and the result that would count before making the change.

Fallback and stop

Fallback family: not written into the saved brief yet
If the water, access, regulation or handling condition becomes a hard stop, stop. Do not solve a stop with a presentation change.
If the first test fails, change one axis. Do not shotgun family, depth, speed and angle together.

Offline readiness

There is not yet a named trip to make field-ready.
Named water: Pick a named public water before calling this trip field-ready.
Species reference: No target species is stored in the working trip.
Starting presentation: The field queue needs a starting family before signal is lost.
Holding water + depth: Lock a starting holding class and depth band so the first test is explicit.
Regulations: Regulations are deliberately not trusted from an offline snapshot. Review the managing agency while connected.
Access: Public-access class is reference data; today’s closure, gate and launch status still need live verification.
Offline runtime: The browser has not yet been checked for service-worker control and cached field routes.

Trip-specific offline readiness

Can this phone carry this trip without signal?

not ready

There is not yet a named trip to make field-ready.

0 ready · 5 attention · 2 live-only

attention · Named water

Pick a named public water before calling this trip field-ready.

attention · Species reference

No target species is stored in the working trip.

attention · Starting presentation

The field queue needs a starting family before signal is lost.

attention · Holding water + depth

Lock a starting holding class and depth band so the first test is explicit.

live-only · Regulations

Regulations are deliberately not trusted from an offline snapshot. Review the managing agency while connected.

live-only · Access

Public-access class is reference data; today’s closure, gate and launch status still need live verification.

attention · Offline runtime

The browser has not yet been checked for service-worker control and cached field routes.

Source / provenance viewer

What is fact, what is observed, what is derived

Provenance is part of the product contract. A public observation, an agency source, a durable editorial fact, an application inference and your own record are not displayed as though they have the same authority.

editorialNamed-water catalogueInspect

Authority: Curated Field Ops reference rows with managing-agency identity and links.
Transport: Supabase public-reference SELECT → same-origin /api/catalogue → row validation → compiled offline floor.
Freshness: Editorial corrections can update the server catalogue without an app deploy. Agency links still require live verification before a trip.
Offline: Names, states, water type and the compiled operational floor remain available without signal.
Personal data: No personal trip record is sent to this catalogue path.
Inference boundary: A named water is an identity, not a claim that access, season or conditions are currently open or safe.

editorialSpecies referenceInspect

Authority: Curated species record: accepted name, thermal ranges, spawn context, forage classes and field note.
Transport: Supabase public-reference SELECT → same-origin /api/catalogue with a compiled offline fallback.
Freshness: General biological ranges are durable reference facts; local populations and regulations can differ.
Offline: The compiled species catalogue and behavior layer are available offline.
Personal data: No personal observation is required to read species reference data.
Inference boundary: Thermal and behavior layers are decision context, not a prediction that a fish is present or feeding.

public observationPublic weather conditionsInspect

Authority: Public weather observation/forecast source identified in the conditions read.
Transport: Fetched through the application conditions boundary and explicitly carried into a trip when the angler chooses to keep it.
Freshness: Short-lived by design. The readiness model expires carried weather context rather than serving it as current indefinitely.
Offline: A carried read can remain in the local trip record, but the app does not claim it is still current offline.
Personal data: The request concerns a public named water, not an exact private angler coordinate.
Inference boundary: Weather describes the environment. It is not converted into a bite score.

agencyOfficial gauge / hydrologyInspect

Authority: USGS or the official gauge identity mapped to the named public water.
Transport: Official observation → server/runtime fetch → displayed with site id and observation time.
Freshness: Observation time is shown. Stale or unavailable flow remains stale or unavailable; it is never converted to normal.
Offline: A prior observation can be shown as a historical carried value only when labelled with its timestamp.
Personal data: No user location is sent to the gauge service.
Inference boundary: Flow and temperature are physical observations. The desk interprets constraints but does not claim they guarantee fishing quality.

derived modelChange / experiment reasoningInspect

Authority: Field Ops decision rules plus the observations recorded in the active on-water session.
Transport: Runs locally in the browser from session state; no AI or remote inference service is required.
Freshness: As current as the most recent observation the angler logged.
Offline: Fully offline.
Personal data: Session observations remain on this device.
Inference boundary: Every suggestion names the observation that produced it, the one axis to test, what to hold constant and how to read the result.

derived modelHabitat archetype libraryInspect

Authority: Physical habitat classes expressed as geometry, current/depth/light mechanisms and field inspection questions.
Transport: Shipped with the application as reference content.
Freshness: Durable physical descriptions; editorial review is tracked by the freshness workflow.
Offline: Fully offline.
Personal data: No personal location is attached to an archetype.
Inference boundary: An archetype describes what a shape of water does. It never claims a particular named spot contains fish.

personal recordTrips, sessions, experiments and debriefsInspect

Authority: The angler’s own choices and observations.
Transport: Browser local storage only; export/import is the explicit way records move between devices.
Freshness: Timestamped when written; historical records remain historical rather than being silently refreshed.
Offline: The primary record works offline.
Personal data: No account, login, server sync or analytics upload path. Exact coordinates are outside the record model.
Inference boundary: Coverage, plan-vs-reality and replay report what was recorded. Small samples are labelled thin rather than turned into confidence scores.

editorialExternal procedural learning resourcesInspect

Authority: Curated external publisher/video links selected for procedural instruction.
Transport: External link opens outside the offline application.
Freshness: The editorial freshness automation re-checks reachability on a schedule.
Offline: The explanation and plates remain offline; the external video does not.
Personal data: Opening an external resource leaves Field Ops and is subject to that provider’s policies.
Inference boundary: External videos are never evidence for regulations, safety stops or conservation thresholds.