MTBS Classification

Description

Classifies fire sectors using the Monitoring Trends in Burn Severity (MTBS) protocol. Provides standardized severity class assignments (1-4) with MTBS data source citations. This skill focuses on the classification methodology and sector-level assignments, complementing the Soil Burn Severity skill's detailed assessment.

Triggers

When should the agent invoke this skill?

• •User asks for MTBS classification of a fire
• •Query requests severity class numbers (1-4 scale)
• •Questions about MTBS methodology or protocol
• •Request for severity map classes
• •Questions about severity thresholding methodology
• •Comparison of sectors by MTBS class

Instructions

Step-by-step reasoning for the agent:

1. •Load Fire Data: Retrieve sector data for the specified fire
   • •Accept fire_id parameter to identify the fire
   • •Load from Cedar Creek fixtures or provided data
   • •Validate that sector data includes dNBR values
2. •Apply MTBS Protocol: Classify using MTBS thresholds
   • •Class 1: Unburned/Unchanged (dNBR < 0.1)
   • •Class 2: Low Severity (0.1 ≤ dNBR < 0.27)
   • •Class 3: Moderate Severity (0.27 ≤ dNBR < 0.66)
   • •Class 4: High Severity (dNBR ≥ 0.66)
3. •Assign Class Labels: Apply standard MTBS nomenclature
   • •Use numeric class (1-4) for mapping systems
   • •Use text label for human readability
4. •Calculate Class Distribution: Aggregate by severity class
   • •Sector count per class
   • •Acreage per class
   • •Percentage of total area
5. •Generate Classification Report: Document methodology
   • •Include MTBS source citation
   • •Note imagery date and sensor
   • •Provide class definitions

Inputs

Outputs

Reasoning Chain

Step-by-step reasoning for the agent:

1. •First, identify the fire and load sector-level dNBR data
2. •Then, apply MTBS classification thresholds to each sector
3. •Next, assign numeric class (1-4) and text label
4. •Then, aggregate class distribution across all sectors
5. •Finally, identify the dominant severity class and generate report

Resources

• •resources/mtbs-thresholds.json - Official MTBS classification thresholds

Scripts

• •scripts/classify_mtbs.py - Python implementation of MTBS classification
  • •Function: execute(inputs: dict) -> dict
  • •Inputs: {"fire_id": "cedar-creek-2022"}
  • •Returns: Classification report with sector assignments

Examples

Example 1: Basic MTBS Classification

Input:

{
  "fire_id": "cedar-creek-2022"
}

Output:

{
  "fire_id": "cedar-creek-2022",
  "fire_name": "Cedar Creek Fire",
  "total_acres": 127341,
  "classification_summary": {
    "class_4": {
      "label": "High Severity",
      "acres": 81041,
      "percentage": 63.6,
      "sector_count": 4
    },
    "class_3": {
      "label": "Moderate Severity",
      "acres": 37900,
      "percentage": 29.8,
      "sector_count": 3
    },
    "class_2": {
      "label": "Low Severity",
      "acres": 8400,
      "percentage": 6.6,
      "sector_count": 1
    }
  },
  "sector_classifications": [
    {
      "id": "CORE-1",
      "name": "Central Fire Origin",
      "mtbs_class": 4,
      "mtbs_label": "High Severity",
      "dnbr_mean": 0.81,
      "acres": 27001
    }
  ],
  "dominant_class": {
    "class": 4,
    "label": "High Severity",
    "percentage": 63.6
  },
  "mtbs_metadata": {
    "source": "MTBS",
    "imagery_date": "2022-09-15",
    "thresholds": "Key & Benson (2006)",
    "sensor": "Landsat 8 OLI"
  },
  "reasoning_chain": [
    "Loaded 8 sectors for Cedar Creek Fire",
    "Applied MTBS classification to each sector",
    "Class 4 (High Severity) is dominant at 63.6% of area"
  ]
}

References

• •MTBS Classification: https://www.mtbs.gov/
• •Key & Benson (2006) Landscape Assessment
• •Landsat dNBR methodology: https://www.usgs.gov/landsat-missions