Tier 0 Hydrology Screening Metrics

Purpose

P114.4 adds fast watershed/year screening metrics from the disturbance ledger. They describe harvest and road exposure for review and UI use.

Tier 0 is not a hydrologic response model. It does not estimate peak-flow frequency, flood probability, or damage. Those begin in P114.5.

Metrics

For each watershed and year:

  • gross harvest, retention, and effective harvest area;

  • harvest_fraction;

  • an equivalent clear-cut area (ECA) diagnostic: eca_area_ha and eca_fraction;

  • active and connected road length;

  • road_density_km_per_km2 and connected_road_density_km_per_km2;

  • active road crossings and crossing_density_per_km2;

  • riparian_effective_harvest_ha, snow_transient_effective_harvest_ha, and high_elevation_effective_harvest_ha.

ECA Semantics

ECA is an explicitly diagnostic indicator:

  • it uses the most recent harvest per unit;

  • it can be undiscounted or discounted by a configurable linear recovery period;

  • it is not a frequency-based response;

  • it is not a substitute for the P114.5 stochastic analysis.

Without recovery_years, ECA is undiscounted and the report warns. With recovery_years, each unit’s last effective harvest area declines linearly to zero over that period.

Road Metrics

  • road_active_km is the sum of active road length in the watershed.

  • road_connected_km is the active length flagged as connected to streams.

  • road_density_km_per_km2 is active km divided by watershed area.

  • connected_road_density_km_per_km2 is connected km divided by watershed area.

  • road_connected_fraction is connected km divided by active km, or null when no road is active.

  • crossing_density_per_km2 is active crossings divided by watershed area.

Exposure Context

Riparian, snow-transient, and high-elevation exposure use reviewed per-unit fractions:

from femic.hydrology.screening import UnitExposureContext

contexts = (
    UnitExposureContext(
        "u1",
        riparian_fraction=0.25,
        snow_transient_fraction=0.5,
        high_elevation_fraction=0.1,
    ),
)

These are context-weighted fractions, not geometry intersections. The reviewed spatial analysis that produces them belongs to P114.2 or a later spatial overlay.

Regimes

A watershed regime can be recorded as rain, rain_on_snow, snowmelt, snow_transient, or unknown. Missing regimes default to unknown with a warning; this is a metadata flag, not a model.

Example

from femic.hydrology.screening import (
    WatershedRegime,
    compute_tier0_screening,
    write_tier0_screening_report,
)

report = compute_tier0_screening(
    ledger,
    domain,
    regimes=(WatershedRegime("w1", "snow_transient"),),
    unit_contexts=contexts,
    recovery_years=60,
)
write_tier0_screening_report(
    report,
    Path("metadata/hydrology/tier0_screening.json"),
)

Outputs

The report uses schema femic.hydrology.screening.v1 and records:

  • screening and ledger/domain identities;

  • recovery configuration;

  • one row per watershed/year;

  • caveats and warnings;

  • deterministic screening_id.

Limits

  • ECA is a screening diagnostic, not a response model.

  • Context fractions are supplied inputs, not computed geometry.

  • No instance-specific threshold or operational decision is applied.

  • Regime flags do not model climate or event frequency.

  • The report does not replace field or hydrologist review.

See Also