Landslide Susceptibility Assessment

Purpose

Landslide Susceptibility Assessment quantifies landscape susceptibility to shallow and deep-seated landslides using terrain, geology, and hydrologic data. Outputs probability-based susceptibility maps for hazard zoning, planning, and risk reduction.

Typical Questions This Tool Helps Answer

• Which areas of the study region have the highest combined slope, curvature, and elevation susceptibility score, and what fraction exceeds the high-risk threshold?
• How does susceptibility change when elevated rainfall intensity is added, and which areas shift from moderate to high or extreme risk class?
• Which high-susceptibility slopes intersect infrastructure corridors or settlement footprints and require priority mitigation attention?

Background

Environmental risk workflows are built around source-pathway-receptor logic: identify stressors, model transport or exposure pathways, and estimate consequence to assets or ecosystems. Inputs and model assumptions define the validity envelope, so uncertainty documentation is as important as the final risk score.

In practice, these tools are most defensible when used comparatively across scenarios, with explicit thresholds and confidence narratives. Interpretation should focus on rank ordering and mitigation prioritization, not absolute certainty.

Susceptibility mapping combines terrain predisposition, hydrologic forcing proxies, and material/land-cover context to rank likely failure zones. Outputs are probabilistic or relative-risk indicators, not deterministic failure predictions.

Methodological Considerations

• Calibrate thresholds and class boundaries against local context whenever possible; transferred defaults should be treated as provisional.
• Evaluate scenario sensitivity explicitly so mitigation priorities are robust to uncertainty in assumptions.
• Keep lineage between assumptions, intermediate metrics, and summary outputs for audit-ready interpretation.

Practical Interpretation Pitfalls

A frequent mistake is interpreting risk classes as deterministic outcomes. These products are comparative planning aids and should be paired with field validation and expert review.

Inputs

Parameters

Profile setting adjusts relative factor influence:

• fast: stronger slope weighting for rapid screening.
• balanced: mixed slope/curvature/elevation/rain weighting.
• conservative: higher curvature emphasis for stricter terrain interpretation.

Factor contribution mode controls summary explainability detail:

• none: hides factor-weight details.
• basic: includes core weights and dominant factor.
• detailed: includes ranked factor percentages.

Outputs

Output artifact keys below are runtime outputs, not input parameters.

Output filenames:

• <output_prefix>_susceptibility.tif
• <output_prefix>_trigger_pressure.tif
• <output_prefix>_confidence.tif
• <output_prefix>_risk_zones.gpkg
• <output_prefix>_summary.json
• <output_prefix>_report.html

Risk-zone attributes:

• ZONE_ID
• SUSC
• CONF
• CLASS

Summary-contract additions:

• output_semantics: machine-readable output intent and certification scope for each output key.
• confidence_contract: confidence metric declaration including threshold and low-confidence fraction.
• interpretation_warnings: plain-language warnings describing proxy limits and run-to-run comparison constraints.

Validation

• Confirm inputs and parameter ranges are valid.
• Verify threshold and profile values in summary JSON.
• Verify zone count is bounded by max_zone_features.
• Review explainability notes before external delivery.
• Confirm summary includes output_semantics, confidence_contract, and interpretation_warnings.

Example

import whitebox_workflows as wbw

env = wbw.WbEnvironment()
env.landslide_susceptibility_assessment(
        dem="dem_10m_region.tif",
        rainfall_intensity="rainfall_intensity_norm.tif",
        profile="balanced",
        factor_contribution_mode="basic",
        susceptibility_threshold=0.65,
        max_zone_features=5000,
        output_prefix="outputs/landslide/assessment"
)

References

• Guzzetti, F., Carrara, A., Cardinali, M., & Reichenbach, P. (1999). "Landslide Hazard Evaluation." Eng. Geol. 58(3–4), 353–372.

Parameter Interaction Notes

Results are most sensitive to profile selection, rainfall availability, and threshold choice.

• Higher thresholds produce fewer high-risk zones.
• Lower thresholds broaden warning areas and zone counts.
• If rainfall is absent, the workflow uses a neutral trigger proxy and this should be documented.

QA and Acceptance Criteria

Use a staged acceptance approach for Landslide Susceptibility Assessment:

1. Confirm DEM and optional rainfall inputs are valid.
2. Confirm expected rasters, vector zones, and summary outputs are generated.
3. Validate susceptibility threshold and resulting high-risk fractions.
4. Confirm factor contribution detail level matches selected mode.

Recommended acceptance checks:

• Summary workflow ID is correct.
• High-risk fraction aligns with mapped hotspots.
• Zone output size is within operational limits.

Advanced Operational Guidance

For production deployment of Landslide Susceptibility Assessment:

• Use basic mode for internal operations and detailed mode for formal reporting.
• Tune threshold + zone cap together to control delivery volume.
• Archive summary JSON with project records.

Implementation Patterns

Common implementation patterns with Landslide Susceptibility Assessment:

• Baseline screening run (balanced, basic).
• Explainability run (detailed) for stakeholder package.
• Threshold sensitivity run to bracket operational risk.

Related Tools

Use Landslide Susceptibility Assessment together with upstream conditioning and downstream validation tools in the same bundle to ensure end-to-end consistency and stronger decision confidence.

When To Use This Workflow

Landslide Susceptibility Assessment is best used when you need defensible outputs for permitting, compliance reporting, or external audit review.

What this workflow helps you achieve:

• Reduces interpretation ambiguity by standardizing method and output structure.
• Produces documentation-ready outputs that can be shared with regulators and stakeholders.
• Shortens time from analysis to decision package.

Results Delivery Checklist

Before handing results to a customer or regulator:

1. Confirm AOI, CRS, and temporal scope match project statement of work.
2. Verify assumptions and threshold values are documented in the run metadata.
3. Export both primary outputs and summary tables for non-technical stakeholders.
4. Include at least one validation artifact (field check, independent layer comparison, or sensitivity run).
5. Archive parameter profile used for reproducibility.

Common Questions

Q: Can this output be used directly in compliance submissions?
A: Yes, when accompanied by documented inputs, parameter profile, and validation evidence.

Q: How do we explain uncertainty to a non-technical reviewer?
A: Use confidence/quality layers and summarize uncertainty in plain-language ranges (high/moderate/low confidence).

Q: What is the most common failure mode in delivery?
A: Scope mismatch (extent, date range, or baseline assumptions) rather than algorithmic failure.

Q: What does factor_contribution_mode=none do?
A: It intentionally suppresses factor-weight details in summary outputs.

Q: Why are there many risk-zone polygons?
A: Each qualifying high-risk area can generate many features; adjust threshold and max_zone_features to control output volume.

Q: What should teams inspect first?
A: Review susceptibility and confidence together, then check summary high-risk fraction and zone counts.

Q: Is rainfall required?
A: No. It is optional, but including it improves trigger realism.