Mapping Urban Forest Productivity
and Growth Rate by Remote Sensing
Imagery
Huan Gu and Phil Townsend
[email protected], [email protected]
1
Urban Forest
They provide “ecosystem services”
water runoff mitigation
shade
absorb CO2
increase property values
They’re pretty
http://www.sustainablecities.net
provide recreation areas
people get attached to them
http://www.thedailypage.com/isthmus/archive.php
http://www.deseretnews.com/article/1,5143,705286478,00.html
LIDAR
Forest Structure : “Biomass”
Forest Biochemistry : “Health”
Hyper-spectral Imagery
Study Area
Study Data
LIDAR data in 2005 and 2009
-acquired by Dane County and municipalities
AVIRIS imagery in 2009, 2010 and 2011
-hyperspectral images from NASA
MASTER imagery in 2010
-thermal images from NASA
NAIP imagery in 2008
-color infrared air photos from USDA
Street tree inventory
-from municipalities ~100,000 trees
LIDAR
last return
1st return
LIDAR
sensor
1st and last
return
1st return – last return =
vegetation height
http://forsys.cfr.washington.edu/JFSP06/lidar_technology.htm
LIDAR
Grid with
specified size
Lidar points
Lidar points are each a few feet apart – so we calculate statistics on the points within a
given GRID CELL, for example, a 10 ft X 10 ft cell might have 16 Lidar measurements.
LIDAR Statistics of Forest Structure
10% quantile
70% quantile
80% quantile
90% quantile
95% quantile
Maximum height
Mean height
Standard deviation
Coefficient of Variance
Skewness
Kurtosis
Lidar Statistical Indices
Highest tree
Mean tree height
Crown base height
Crown length
Basal area
Mean stem diameter
Stand stem density
Aboveground biomass
Branch biomass
Foliage biomass
Tree Structure Indices
Aboveground
Biomass (mg/ha)
Water
No Forest
0 - 50
50 - 100
100 - 150
150 - 200
> 200
Mean Stem
Diameter (in)
Water
No Forest
0-5
5 - 10
10 - 15
15 - 20
20 - 25
Mean Tree
Height (ft)
Water
No Forest
0-5
5 - 15
15 - 50
50 - 75
> 75
Aboveground
Biomass (mg/ha)
Water
No Forest
0 - 50
50 - 100
100 - 150
150 - 200
> 200
Mean Stem
Diameter (in)
Water
No Forest
0-5
5 - 10
10 - 15
15 - 20
20 - 25
Mean Tree
Height (ft)
Water
No Forest
0-5
5 - 15
15 - 50
50 - 75
> 75
Aboveground
Biomass (mg/ha)
Water
No Forest
< -20
-20 - 0
0 - 20
> 20
Mean Stem
Diameter (in)
Water
No Forest
< -6
-6 - 0
0-6
>6
Mean Tree
Height (ft)
Water
No Forest
< -6
-6 - 0
0-6
>6
Aboveground
Biomass (mg/ha)
Water
No Forest
0 - 50
50 - 100
100 - 150
150 - 200
> 200
Mean Stem
Diameter (in)
Water
No Forest
0-5
5 - 10
10 - 15
15 - 20
20 - 25
Mean Tree
Height (ft)
Water
No Forest
0-5
5 - 15
15 - 50
50 - 75
> 75
Aboveground
Biomass (mg/ha)
Water
No Forest
0 - 50
50 - 100
100 - 150
150 - 200
> 200
Mean Stem
Diameter (in)
Water
No Forest
0-5
5 - 10
10 - 15
15 - 20
20 - 25
Mean Tree
Height (ft)
Water
No Forest
0-5
5 - 15
15 - 50
50 - 75
> 75
Aboveground
Biomass (mg/ha)
Water
No Forest
< -20
-20 - 0
0 - 20
> 20
Mean Stem
Diameter (in)
Water
No Forest
< -6
-6 - 0
0-6
>6
Mean Tree
Height (ft)
Water
No Forest
< -6
-6 - 0
0-6
>6
AVIRIS
http://masterweb.jpl.nasa.gov/sensor/sensor.htm
AVIRIS (VIS-NIR-SWIR)
• 224 channels
• 380-2500 nm
450
1450
Wavelength (nm)
2450
Foliar N concentrations, Madison, WI
AVIRIS Image, Madison 2009
False Color Composite Imagery
Madison 2009
Foliar N concentration (%)
7.9%
0%
Fully Characterize Urban Ecosystem At A Point In Time
Acknowledgement:
Aditya Singh
Shawn Serbin
Clayton Kingdon
Bernard Isaacson
Peter Wolter
John Couture
Benjamin Spaier
Wesley Fox
Field Assistants
Marla Eddy
Dave Davis
Kirk Contrucci