Validation of National Burn Severity Validation of National Burn Severity Validation of National Burn Severity Validation of National Burn Severity Mapping Project Techniques Within Mapping Project Techniques Within the Apalachicola National Forest the Apalachicola National Forest A A i i i i Joshua J. Picotte Dr. Kevin Robertson
Project Overview Project Overview � Background � Background Background Background � Project Questions Project Questions � Results � Results Results Results � Problems Problems � Conclusions � Conclusions Conclusions Conclusions
Background Background- -Burn Severity and Burn Severity and Joint Fires Science Program (JFSP) Joint Fires Science Program (JFSP) � Burn Severity Burn Severity- -ecosystem ecosystem change and landscape change and landscape h h d l d l d d change change � System developed in 1996 System developed in 1996 b C l K b C l K by Carl Key and Nate by Carl Key and Nate d N t d N t Benson Benson � Started in the Western Started in the Western Key and Benson 1995 U it d St t U it d St t United States United States � Initial Fires Assessed Initial Fires Assessed- -1994 1994 Glacier National Park Fires Glacier National Park Fires � Needs more work in East � Needs more work in East Needs more work in East Needs more work in East � JFSP was formed in 1998 to JFSP was formed in 1998 to provide support for fuel and provide support for fuel and fire management programs fire management programs g g p p g g Key and Benson 1995
Background- Background -Composite Burn Index Composite Burn Index (CBI) and Normalized Burn Ratio (CBI) and Normalized Burn Ratio (CBI) and Normalized Burn Ratio (CBI) and Normalized Burn Ratio (NBR) (NBR) � NBR NBR-Remote sensing of burn severity Remote sensing of burn severity i i f b f b i i � Landsat bands R Landsat bands R 4 and R and R 7 � NBR = (R NBR = (R 4 -R R 7 )/ (R )/ (R 4 +R +R 7 ) � dNBR=NBR prefire-NBR postfire � CBI CBI- -Ground measure of burn severity Ground measure of burn severity � 30 m plot 30 m plot p � CBI should validate NBR values CBI should validate NBR values
Background Background- -deltaNormalized Burn deltaNormalized Burn Ratio (dNBR) Ratio (dNBR) Ratio (dNBR) Ratio (dNBR) � dNBR: Change in reflectance between pre-fire (1 year pre-fire) 30m and post-fire NBR values 30m 30m � dNBR=NBR prefire -NBR postfire d � Weighted dNBR = Average of plot center and corners (N=5) t d (N 5) Severity dNBR Unburned Unburned -100 - 99 100 99 Low 99 - 269 Low-Moderate 269 - 439 Moderate High Moderate-High 439 439 - 659 659 High 659 - 1300
Background Background- -Composite Burn Composite Burn Index (CBI) Index (CBI) Index (CBI) Index (CBI) � Burn Index: 0 � Burn Index: 0 Burn Index: 0-3 Burn Index: 0 3 � � � 0-Unburned Unburned � 3-Severe Burn Severe Burn � Five Strata Five Strata � � 4-5 Ratings Factors 5 Ratings Factors � Averaged Averaged � CBI Score CBI Score � � Average of Five Strata Average of Five Strata
Background Background- -Apalachicola National Apalachicola National Forest (ANF) Forest (ANF) Forest (ANF) Forest (ANF) Gulf of Mexico 2006 Dormant dNBR Unburned Low Severity Low-Moderate Severity Moderate-High Severity High Severity g y ² ² 0 6,858 13,716 27,432 Meters
Background- Background -Habitats Within ANF Habitats Within ANF A. Sandhill Pineland Sandhill Pineland A. A. Turkey Oak Turkey Oak • Wiregrass Wiregrass • Bracken Fern Bracken Fern • Long Long- -Leaf Pine Leaf Pine • B. Flatwood Pineland Flatwood Pineland B. B. Palmetto Palmetto Palmetto Palmetto • Long Long- -Leaf/Slash Pine Leaf/Slash Pine • Depression Swamp Depression Swamp C. C. C Cypress Cypress • C. Titi Titi • Pond Pine Pond Pine •
Project Questions Project Questions Does this system adequately describe burn severity Does this system adequately describe burn severity y y q q y y y y 1. 1. with the South East ecosystems? with the South East ecosystems? How long after a fire can burn perimeters be How long after a fire can burn perimeters be 2. 2. remotely sensed? remotely sensed? l l d d Are there differences in the effectiveness of using Are there differences in the effectiveness of using 3. 3. CBI and dNBR to categorize burn severity in the CBI and dNBR to categorize burn severity in the CBI and dNBR to categorize burn severity in the CBI and dNBR to categorize burn severity in the three community types of the ANF? three community types of the ANF? W W What are the appropriate ranges dNBR for given What are the appropriate ranges dNBR for given e e e pp op e pp op e e ges d ges d o g ve o g ve 4. 4. burn severity categories? burn severity categories? What problems arise in assessing burn severity? What problems arise in assessing burn severity? 5. 5.
Results Results- -Sandhill Burn Severity Sandhill Burn Severity ² 25 0 335 671 1341 Meters 20 Severity Acres Area Unburned 146.7 15 Percent Fire Low 617.625 Low-Moderate 318.6 10 Moderate-High 9.675 High 0.225 5 Total 1092.825 Total Burned 946.125 0 dNBR
Results Results- -Flatwood Burn Severity Flatwood Burn Severity Wet Flatwoods, Initial Assessment ² 20 0 8,001 16,002 32,004 Meters 18 18 16 14 e Area Severity Acres 12 Percent Fire Unburned 9850.5 10 Low 5673.825 8 Low-Moderate 3442.725 6 Moderate-High 1312.2 4 4 High 168.975 2 Total 20448.23 Total Burned 10597.73 0 dNBR
Results- Results -Correlation Between CBI Correlation Between CBI and dNBR and dNBR and dNBR and dNBR Sandhill Sandhill Overall 2.5 2.5 2 2 1.5 1.5 CBI CBI 1 1 0.5 0.5 0 0 -100 0 100 200 300 400 500 600 700 0 50 100 150 200 250 300 350 400 450 500 dNBR Weighted dNBR Weighted Flatwoods 2.5 2 2 R 2 P>Fr 1.5 Overall 0.125 0.005 CBI Sandhills 0.000 0.913 1 Flatwoods Flatwoods 0 767 0.767 <0 001 <0.001 0.5 0 -100 0 100 200 300 400 500 600 700 dNBR Weighted
Results Results- -Correlations between CBI Correlations between CBI and dNBR Within Substrates and dNBR Within Substrates and dNBR Within Substrates and dNBR Within Substrates Overall R 2 P>Fr Substrates Substrates 0.002 0.002 0.733 0.733 Vegetation < 1m 0.212 <0.001 1m < Vegetation < 5m 0.187 <0.001 Subcanopy Trees 0.040 0.174 Upper Canopy Trees Upper Canopy Trees 0.032 0.032 0.079 0.079 Sandhills Flatwoods R 2 2 P>Fr R 2 2 P>Fr Substrates 0.033 0.312 Substrates 0.304 0.002 Vegetation < 1m 0.011 0.556 Vegetation < 1m 0.372 <.001 1m < Vegetation < 5m 0.038 0.278 1m < Vegetation < 5m 0.499 <.001 Subcanopy Trees S 0.005 0 00 0.739 0 39 S Subcanopy Trees 0.509 0 09 <.001 001 Upper Canopy Trees 0.032 0.328 Upper Canopy Trees 0.516 <.001
Burn Severity Change Detection Burn Severity Change Detection- - Water Water Water Water ² ² ² 0 50 100 201 Meters 0 411 823 1646 Meters 2006 Dormant dNBR Unburned Low Severity Low-Moderate Severity Moderate-High Severity Moderate High Severity High Severity Lakes and Ponds Actual Pond Boundary
Problematic detection of Low Burn Problematic detection of Low Burn Severity Severity Severity Severity A. 2006 Dormant dNBR B. Unburned Low Severity Low-Moderate Severity Moderate-High Severity High Severity ! 30m Plots C. C. Plot CBI dNBR A. 1.23 1 B. 1.23 59.2 C. 1.27 89.6 ² ² 0 305 610 1,020 Meters
Conclusions Conclusions Conclusions Conclusions � Successful in high fire frequency communities � Successful in high fire frequency communities Successful in high fire frequency communities Successful in high fire frequency communities � Extrapolation to other communities within the East Extrapolation to other communities within the East � Useful system for burn monitoring Useful system for burn monitoring U U f l f l t t f f b b it it i i � Sense perimeter months following the burn Sense perimeter months following the burn � Season of remote Season of remote- -sensing (dNBR) may be sensing (dNBR) may be limited limited � Useful in assessing the “success” of burn Useful in assessing the “success” of burn management plans management plans g g p p
Acknowledgements Acknowledgements � Nate Benson Nate Benson � Jason Drake Jason Drake � Leigh Lentile Leigh Lentile � Caroline Noble Caroline Noble � Joe Noble Joe Noble � Don Ohlen Don Ohlen � Kathy Marois Kathy Marois � David Printiss � David Printiss David Printiss David Printiss � Greg Titus Greg Titus � Eugene Watkins � Eugene Watkins Eugene Watkins Eugene Watkins
Recommend
More recommend
