Normalized Difference Fraction
Index (NDFI): a new spectral
index for enhanced detection
of forest canopy damage
caused by selective logging
and forest fires
Carlos Souza Jr., Imazon
Dar Roberts, UCSB
Mark Cochrane, USD
Samia Nunes, Imazon
Deforestation vs Forest Degradation
Ikonos Image – Paragominas, PA
Creates a complex
environment:
{
{
{
{
Caused by:
{
{
{
Souza Jr. and Roberts (2005) - IJRS
Undisturbed forests
Canopy gaps
Exposed soils
Dead vegetation
Selective logging
Forest fires
Forest fragmentation
Very Challenge with Landsat!
2000
1998
Selective
logging and burning
Selective
logging
2001
1999
Old Selective
logging
Souza Jr. et al. (2003), Rev. Ciência Hoje
Old selective
logging and
burning
Objective
Develop a technique that combines spectral
and spatial information to enhance the
detection and mapping of canopy damage
caused by selective logging and forest fires
{
New spectral index: Normalized Difference
Spectral Index (NDFI)
{
Contextual classification algorithm (CCA)
Spectral: NDFI
Spatial: log landings
Degradation Intensity
Forest Transect Classes
Forest class
Field Description
Intact
(n=4)
Mature and undisturbed forest
Non-mechanized
logging
(n=5)
Logged forest without the use of vehicles such as
skidders and trucks, also known as traditional
logging. Log landings, roads and skid trails are not
built.
Managed logging
(n=5)
Planned selective logging where a tree inventory is
conducted, followed by road and log landing
planning to reduce harvesting impacts.
Conventional
Logging
(n=2)
Logged and burned
(n=3)
Conventional unplanned selective logging using
skidders and trucks. Log landings, roads and skid
trails are built.
Either non-mechanized or conventionally logged
forests that have subsequently been damaged by
intense surface fires.
Souza Jr. et al. (2005), Earth Interactions
Forest Transects
Forest inventories
{
Transects (10 m x 500 m = 0.5 ha)
Trees with DBH ≥ 10 cm
Sub-plots (10 m x 10 m)
{
All trees
{
Forest canopy cover
{
Vine density
{
% soil exposed
{
% of dead vegetation
Aboveground biomass estimation
{
Allometric equations (Gerwing, 2000)
Objective I
Forest Degradation Separability
SMA Fraction Images
GV
60
NPV
S hade
S o il
Fraction Value (%)
50
40
30
20
10
0
Intact
N on -me chan iz e d
L og ging
M anag e d
L og ging
Souza Jr. et al. (2005), Earth Interactions
C o nv e ntion al
L og gin g
L og ge d an d
B urn e d
Normalized Difference Fraction Index
GVShade − (NPV + Soil)
NDFI =
GVShade + NPV + Soil
GVShade
GV
=
100 − Shade
-1 ≤ NDFI ≤1
NDFI low to moderate
Low to moderate GV
Moderate to high NPV
and Soil
NDFI near 1
High GV
Low NPV and Soil
Souza Jr. et al. (2005), RSE
SMA Results
Soil
GVshade
Fraction Value (%)
Fraction Value (%)
NPV
NDFI
Fraction Value (%)
Souza Jr. et al. (2005), RSE
NDFI
NDFI vs. Fraction Images
Change from Intact Forest (%)
10
Percent Change from Intact Forest
5
GV
0
NPV
Soil
-5
Shade
-10
NDFI % Change
-15
-20
Nonmechanized
Logging
Souza Jr. et al. (2005), RSE
Managed
Logging
Logged
Logged and
Burned
Class Separability
Class
Nonmechanized
Logging
Intact
Mean
Stdev.
Mean
Stdev.
Managed
Logging
Mean
Stdev.
Conventional
Logging
Mean
Stdev.
Logged and
Burned
Mean
Stdev.
GV
40a
4
41a
5
41a
5
38b
9
25c
7
NPV
6a
2
5a
2
6a
2
10bc
4
11bd
3
Soil
2a
1
1a
1
3ab
1
4bc
3
7d
3
Shade
51a
3
53a
5
51ab
4
49bc
3
56d
3
NDFI
0.84a
0.08 0.87a 0.07 0.79b 0.07
Tukey test at P<0.01
Souza Jr. et al. (2005), RSE
0.58c
0.24 0.49d 0.22
Contextual Classification Algorithm - CCA
Step 1: Find log landings
Soil > 10%
Find regions and calculate area
1 ≤ Area ≤ 4 pixels
Soil
Fraction
Log
Landings
Step 2: Grow a canopy damage region around log
landings
Search for NDFI neighboring cell values
If NDFI > 0.75 then Intact Forest
if 0 ≤ NDFI ≤ 0.75 then Canopy Damage Canopy
Damage
Logged and
Burned
Conventional
Logging
CCA Results
NDFI
Canopy Damage
Accuracy Assessment
Classified
Pixels
Non-Forest
Reference Pixels
Non-forest
Forest
Canopy
Damage
Total
Users
Accuracy
(%)
454
0
17
471
96.4
Forest
6
625
117
748
83.6
Canopy Damage
40
0
616
656
93.9
Total
500
625
750
1875
Producers Accuracy (%)
91.0
100.0
82.0
Overall Accuracy = (1695/1875) = 90.4%
Kappa Coefficient = 0.85
Generic Image Endmembers
Objective IV
Standardized Fractions and NDFI
a) Paragominas,Pará State - 223/62
Objective IV
Soil
NPV
GV
NDFI
Standardized Fractions and NDFI
b) Santarém, Pará State - 227/62
Objective IV
Soil
NPV
GVShade
NDFI
Standardized Fractions and NDFI
c) Ji-paraná, Rondônia State - 231/67
Objective IV
Soil
NPV
GVshade
NDFI
Standardized
NDFI (2001)
Rondônia
Standardized
NDFI (2001)
N
Summary
NDFI enhances the detection and
mapping of degraded forests and
performs better than any individual
fraction.
CCA can unambiguously map canopy
damage due to selective logging and
forest fires.
These techniques can be applied over
the Amazon region.
© Copyright 2024 ExpyDoc
