Evaluating the Quality of a Wildfire Defensible Space with Airborne LiDAR and GIS Jason Harshman GEOG 596A Capstone Proposal August 2015 1.

Evaluating the Quality of a Wildfire Defensible Space with Airborne LiDAR and

GIS

Jason HarshmanGEOG 596A Capstone Proposal

August 2015

1

Presentation Outline

• Background

• Project Goals

• Study Area

• Methodology

• Initial Results

• Future Plans

2

CAL FIRE

Background• Prolonged drought in California has greatly increased wildfire risk• Year-round wildfire season for much of California• 11,000 structures lost to wildfires between 2003-2013 in

California• Over $30 million in damages in 2013• CAL FIRE annual budget over $200 million

Wildfire near a San Marcos, CA home, May 2014 (Anthony Carrasco/San Diego Union-Tribune)

Wildfire near a San Marcos, CA home, May 2014 (Sean M Haffey/San Diego Union-Tribune) 3

What is a defensible space?• Prevent structure ignition• Requires vegetation maintenance surrounding a structure• 100 feet from a structure’s edge• May extend beyond 100 feet• Requirements can vary between different jurisdictions• Split into 2 zones

Illustration of a defensible space (CAL FIRE). 4

Defensible Space0 to 30 feet (Zone A)

•“Free and Clear”

•Remove tree canopy overlapping structure

•Remove highly flammable vegetation

30 to 100 feet (Zone B)

•Reduce vegetation density

•Reduce ladder fuels

•Requirements vary depending on slopeIllustration of Zone B requirements (CAL FIRE).

5

Project Goals

• Analyze the quality of a structure’s defensible space

• Quality equates to how well a structure’s defensible space follows

defensible space regulations

• Use LiDAR data to map vegetation and terrain surrounding structures

• Develop a general methodology that can be applied to other locations

6

Project Location• Sonoma County, CA• 77% of county wildfire responsibility is CAL FIRE• 20% is rated “Very High” for wildfire risk

Regional location of Sonoma County, CAWildfire responsibility areas in Sonoma County, California

7

Locate a Study Area• Identify a test location for analysis of a small group of structures

o CAL FIRE responsibility area

o 1-mile of the last 10 years of fires

o “Very High” wildfire hazard areas

o “Maximum” wildfire threat areas

o Remove conservation/sensitive land and public land

o Majority conifer and shrub land cover

o Area must have structures

8

Study Area

Proposed Study Area Sample selection of structures

9

LiDAR Data

• Airborne LiDAR and Orthoimagery collected in late 2013

• Currently largest coverage of public domain LiDAR data within

California

• Classified point cloud of LiDAR (http://sonomavegmap.org)

• Structure footprints polygons derived from the LiDAR also

available

Source Sonoma County Vegetation Mapping and LiDAR Program (http://sonomavegmap.org/)

Collected by Watershed Sciences, Inc. (WSI)

Altitude 900 AGL

Laser Pulse Rate 105,000 Hz

Field of View 30°

Average Point Density 2.88 p/m2

Horizontal Accuracy Not provided

Vertical Accuracy ~6 cm

LAS ClassificationsUnused (0), Unclassified (1), Ground (2), High Vegetation (5), Buildings (6), Noise (7), Water (9), Excluded (11)

LAS Format 1.2

10

Methodology Overview

• Create several surface datasets from the LiDAR

• Separate analysis for Zone A and Zone B

• Average results from Zone A and B into final score

11

Surface Datasets• Create several surface datasets from the LiDAR

Dataset Acronym Resolution FunctionData Type

Interpolation Method Source

Digital Elevation Model DEM 2 Feet

Ground elevation surface Grid

Natural Neighbors Triangulation

Ground classified points, all returns

Digital Surface Model DSM 2 Feet

Terrain elevation surface Grid

Natural Neighbors Triangulation First return points

Canopy Height Model CHM 2 Feet

Relative height from the ground Grid N/A DSM-DEM

Slope N/A 2 FeetSlope of ground surface Grid N/A DEM

Tree Canopy Surface Model CSM 2 Feet

Canopy elevation surface Grid Simple

High vegetation classified points, first return

Building Surface Model BSM 2 Feet

Building elevation surface Grid Simple

Building classified points, all returns

12

Zone A• Calculate amount of vegetation coverage within Zone A

Zone A30 feet

Get count within

structure footprints

Calculate % of structure with overlap

Zone A Value

Calculate canopy

overlap of structures

BSM - CSMReclassify positive

values to 1

Get CSM count within

Zone A

Calculate % of CSM in Zone A

Calculate canopy

coverage within Zone

A

Average values

13

• Determine if tree canopy overlaps structure footprint

Zone B• Determine if distance between tree canopy is sufficient

Zone B30-100 feet

Reclassify slope to 1, 2,

3

Create individual tree data

with FUSION

Plot locations in

GIS

Buffer based max

crown widths

Euclidean Distance of

crowns

Reclass distances of 10, 20, 30 ft to 1,

2, 3

Distance Reclass minus Slope

Reclass

14

Zone B

15

Slope % Reclass Tree Distance Reclass

1 (10 feet) 2 (20 Feet) 3 (30 Feet)

1 (0-20%)0 1 2

2 (20-40%)-1 0 1

3 (40%+)-2 -1 0

• Slope Reclass - Distance Reclass matrix

Zone A + Zone B

• Average Scores together

• Scores closer to 0 suggest a better quality defensible space

16

0.0 1.0

High Quality

Poor Quality

Zone A Initial Results• Values range from 0.026 to 0.528

• Lower value = higher quality of defensible space

17

Zone A Initial Results

Lowest Zone A score Lowest Zone B score

18

Zone B

19

Zone A Zone B

Poor quality Zone B

• Identify vegetation coverage that exceed minimum requirements Zone A Zone

B

High quality Zone B

Discussion

20

• Validity of overall methodology

• Zone A initial results promising

• Need quality LiDAR for quality surface datasets

• Surface datasets key to analysis

Justin Sullivan / Getty Images

Future Plans

• Refine methodology

• Scale-up analysis

• Diversify landscape types

• Analysis of other location(s)

• Build a tool or set of tools?21

CAL FIRE

Timeline

22

August

Refine Methodology

ID Other Locations & Gather Data

Submit Abstract

AAG Annual Meeting

Perform

Analysis

October

January

March2016

2015

May

596B

Submit Write-

up

References• CAL FIRE. (2013). Wildfire activity statistics. Sacramento, CA: California Department of Forestry and Fire Protection.

Retrieved from http://www.fire.ca.gov/downloads/redbooks/2013Redbook/2013_Redbook_Final.pdf• CAL FIRE. (2015a). Defensible space and hardening your home Retrieved from

http://www.readyforwildfire.org/docs/files/CALFIRE_Ready_Brochure.pdf • CAL FIRE. (2015b). Emergency fund fire suppression expenditures. Sacramento, California: California Department of

Forestry and Fire Protection. Retrieved from http://www.fire.ca.gov/fire_protection/downloads/SuppressionCostsOnepage.pdf

• Amy Hubbard. (2014, September). California burns through $209-million wildfire budget, taps $70 million more. Los Angeles TImes Retrieved from http://www.latimes.com/local/lanow/la-me-ln-california-exhausts-budget-wildfires-20140930-story.html

• San Diego Union-Tribune. (2014). Day 2: San diego county fires. Retrieved from http://www.sandiegouniontribune.com/photos/galleries/2014/may/14/bernardo-fire-14-contained/

• Alexandra D. Syphard, Teresa J. Brennan, & Jon E. Keeley. (2014). The role of defensible space for residential structure protection during wildfires. International Journal of Wildland Fire, , 1165-1175. doi:10.1071/WF13158

• Philip E. Dennison, Simon C. Brewer, James D. Arnold, & Max A. Moritz. (2014). Large wildfire trends in the western unitedstates, 1984–2011. Geophysical Research Letters, 41, 2928-2933. doi:10.1002/2014GL059576

• Susan M. Stein, Sara J. Comas, James P. Menakis, Mary A. Carr, Susan I. Stewart, Helene Cleveland, Lincoln Bramwell and Volker C. Radeloff. (2013). Wildfire, wildlands, and people: Understanding and preparing for wildfire in the wildland-urban interface. (). Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station.

• Robert J. McGaughey. (2014). FUSION/LDV: Software for LIDAR data analysis and visualization Retrieved from http://forsys.cfr.washington.edu/fusion/FUSION_manual.pdf

• US Forest Service.Smokey bear. Retrieved from http://www.smokeybear.com/design-resources.asp23

Acknowledgements

Doug Miller

24

Questions?

25