TH scores update: WILDFIRE

[matamadio]

Hazard Processing Method

• Data Source: 3 return periods (RP5, RP25, RP50) - FWI rasters
• Unit: FWI (Fire Weather Index)
• Default Value Threshold: 50
• Default Area Threshold: 20%

Workflow: Same as Earthquake/Cyclone

1. Extract all pixel values for each RP raster within each admin unit
2. Calculate mean (for reporting) and area percentage (% of pixels > value threshold: 50 FWI)
3. Check if area threshold % met per RP
4. Score = count of RPs meeting threshold (0-3)
5. Special case: If NO pixels > 0 (FWI = 0) in ALL RPs → Score -1 (not affected)

Scoring Logic:

• No FWI values > 0 in any RP → Score -1 (not affected - no wildfire risk)
• 3 RPs meet threshold → Score 3
• 2 RPs meet threshold → Score 2
• 1 RP meets threshold → Score 1
• 0 RPs meet threshold (but some FWI > 0) → Score 0

---

Comparison

New score distribution: ADM2

New score distribution: URBAN

[matamadio]

Original (2019) vs. 2025 Update

Data Overview

• Original Dataset: 38,229 administrative regions (ADM2 level)
• 2025 Dataset: 39,370 administrative regions (ADM2 level)
• High-Quality Matches: 33,207 regions (90%+ matching confidence)
• Unmatched Original: 5,022 regions
• Unmatched 2025: 6,268 regions

Technical Notes

• Hazard ranks: -1 = N/A, 0 = Very Low, 1 = Low, 2 = Medium, 3 = High
• Original data used text codes (VLO, LOW, MED, HIG), converted to numeric for easier comparison (delta)
• Matching threshold: 90% similarity required for inclusion in analysis
• Medium-quality matches (70-90%) available in separate sheet for manual review

Matching Method

Fuzzy string matching using hierarchical location names (Country | Admin1 | Admin2).
Only high-quality matches (≥90% similarity) were used for hazard comparison.

---

• Regions Compared: 31,742
• Agreement: 51.3% (16,269 unchanged)
• Net Trend: Strong Downgrades (15,071 downgrades vs 402 upgrades)
• Average Change: -1.64 ranks
• Notable Shifts:
  • High → Very Low: 2,502 regions
  • High risk regions decreased from 22,744 to 14,978

Comment

The new data and classification approch combined lead to a general downgrade of risk compared to original scores, especially high values: yet it stills puts 19.5% of ADM and 54.3% of urban areas in high risk. Overall this seems more balanced and realistic than original scores, but I might be wrong.