Mapping Wildfire Disturbances in Southern

1
John Rogan jrogan_at_clarku.edu Geography
Department Clark University
Mapping Wildfire Disturbances in Southern
California Using Machine Learning Algorithms
2
Research Context

• Land cover/use change Mapping and Monitoring
• Growing interest in Rapid Response Information
  Systems
• Empirical information about wildfire cause and
  behavior can inform wildfire risk analysis
• Paucity of historic burn severity information

3
Wildfire Monitoring Programs
4
Monitoring Fire Effects

• The physical environment and its response to fire
  
• Factors affecting fire behavior
• Ecosystem/watershed damage assessment
• Evaluating success of a management ignited fire
• Appraising the potential for future treatments

5
Existing Monitoring Efforts (BAER) Watershed Scale
Source California Department of Forestry and
Fire Protection
6
Existing Monitoring Efforts Ecoregion Scale
Source California Department of Forestry and
Fire Protection
7
Existing Monitoring Efforts National Scale
A Detected from satelliteB Conventional
methods
Source Canadian Center for Remote Sensing
8
Existing Monitoring Efforts Global Scale
Source University of Maryland Global Fire
Product
9
Research Objectives

• Test a new methodology to map fire severity in
  San Diego County (1985-2000)
• Employ machine learning to map severity, while
  integrating environmental variables with spectral
  variables (categorical and continuous)
• Examine the contribution of ancillary variables
  to burn map accuracy

10
Fire/Burn Severity

• Severity - A descriptive term that integrates
  various phenomenological characteristics of a
  fire-altered landscape
• Physical and biological manifestation of
  combustion on vegetation and soil
• Direct Effects Influenced by
• Fuel consumption - Topography
• Crown scorch - Disturbance history
• Soil heating
• Bole Charring

11
Previous Research

• Focus on mapping burn scars (coarse resolution)
• Recent emphasis on burn severity/mortality/damage
  levels (fine-medium resolution) for impact
  assessment
• Retrospective burn area mapping at medium
  resolution (e.g., Hudak and Brockett 2004IJRS)
• BUT, challenges remain

12
Scene Model
13
Post-Fire IKONOS-2 Image
14
California Wildfire Threat
Source California Department of Forestry and
Fire Protection
15
San Diego County
16
Study Area Significance

• Impacts of natural disturbance processes are
  increasing in severity
• Public lands began burning more frequently than
  private lands in the mid-1970s. This trend is
  increasing
• Population increase and
• peri-urban spread into
• fire-prone areas (WUI)

17
Landsat TM and ETM Data
18
Environmental Variables
19
Ground Reference Data
Composite Burn Index Key and Benson (2000)

• SITE-Dominance
• Grassland (6)
• Chaparral (10)
• Conifer-Hardwood (5)
• Mixed (12)

20
Wildfire Effects (After Key and Benson)
21
Methods (Data Processing Flow)
22
Haze Correction (Pre-)
Band 1
Band 2
(a)
(b)
Band 3
Bands 4,3,2
23
Haze Correction (Post-)
(a)
(b)
24
Spectral Mixture Analysis

• Decomposition of mixed
• pixel spectral response
• Production of fractional
• representation of sub-
• pixel proportions
• Biophysically-meaningful
• estimates of land cover
• components

25
Classification Tree Analysis

• A type of MLA used to predict membership of cases
  of a categorical dependent variable from their
  measurements on one or more predictor variables
• Hierarchical, non-linear recursive partitioning
• Structurally explicit

26
Desired Map Accuracy
Source Rogan and Franklin (2001)
27
Case Study (Pre-Fire)
28
Case Study (Post-Fire)
29
Case Study Results - SMA
Shade
GV
BV
Soil
Vegetation Map
RMS
Fire perimeter
30
Case Study Results - Variable Selection
31
Case Study Results Burn Severity
ACCURACY (kappa)
CLASS
No burn Vegetation 90
No burn Water 100
Severe burn 87
82.5
Moderate burn 60
Light burn 74
No burn Bare Soil 84
32
County-Wide Results

• Variable Selection by Site(s)
• Grass Burn, Soil
• CHP Burn, GV, Soil, Veg, Slope
• CON/HDW Burn, GV, Soil, Veg, Slope, Shade
• Mixed Burn, GV, Slope, Veg, Shade, Slope,
  Aspect
• Mean Burn Map Accuracy by Site(s)
• Grass 87 (SD 11)
• CHP 81 (SD 10)
• CON/HDW 84 (SD 7)
• Mixed 70 (SD 16)

33
County-Wide Results

• Time since fire (TSF)
• Most problematic for grasslands, where TSF gt 3
  months
• Least problematic for CHP, CON/HDW
• Map accuracy
• Lowest for complex classes (e.g., mixed)
• Highest for simple classes (e.g., grassland)
• Variable Selection
• Many for complex classes (e.g., mixed)
• Few for simple classes (e.g., grassland)

34
Implications

• Map accuracy
• Range 70-80, depending on landscape type and
  TSF
• Subtle burn classes are least accurate
• Variable Selection
• Varied by landscape type (all used for complex
  areas)
• Implication for fire risk mapping?
• The larger the fire, the greater the potential
  for confusion caused by landscape heterogeneity

35
Wildland Fire Mapping Triangle
Burn Severity Map
Predictive Vegetation Modeling
Image Processing and Enhancement
Machine Learning
.search for standard methods for mapping fuels
and fire regimes at high (spatial) resolutions
over broad areas. Rollins et al. (2004, p.
86)
36
Acknowledgements

• NASA Land Cover Land Use Change Program
• US Forest Service and CDF
• SDSU Janet Franklin and Doug Stow
• UCSB Dar Roberts and Alexandria Digital Library
• U of Arizona Steve Yool