monitoring the dynamics of vegetation components in...
TRANSCRIPT
David Helman and Itamar M. Lensky Bar-Ilan University
I thank Prof. S. Brenner and Prof. Y. Carmel for their helpful comments and pleased
by their willingness to assist and thus contribute to my PhD work in the future. The
changes I made in the research proposal following the comments are delineated
hereafter.
1. I corrected all phrases and spelling mistakes.
2. I added detailed information regard the SEBE with the relevant equations
therein (Chapter 1.2.1 – “The SEB method”, pages 9-10).
3. I added an explanation on the Structural Equation Modeling (SEM) in
Materials and Methods (Chapter 3.3 – “Software”, page 23).
4. I added information on the total size of Yatir forest area for comparison with
satellite resolution in Materials and Methods (Chapter 3.1 – “Remote sensing
data”, page 19).
Sincerely,
David Helman
Department of Geography and Environment
Bar Ilan University
From the 40th Annual Conference of the Israel Society of Ecology and Environmental Science (ISEES), Tel Aviv, October 2012
Monitoring the dynamics of vegetation components in Mediterranean forests:
Time-series analysis of satellite-derived NDVI
Implications for management
Pre-fire mapping of fire-risk Mapping the optimal grazing regime
Conservation and afforestation planning
Chlorophyll a
Chlorophyll b
Carotenoids
Photosynthesis
Rate
Absorption
100% 0%
100% 0%
NIR + Red NIR - Red
NDVI = -1 < NDVI < +1
The Normalized Difference Vegetation Index (NDVI)
1 0.8 0.6 0.4 0.2 0
Time [days]
Growing Senescence
NDVI TIME SERIES
ND
VI
NDVI time series describes well the vegetation's phenology
Measuring plant phenology with NDVI
250 m
Satellite
200 100 0
Precipitation
0.8 0.6 0.4 0.2 0
Time [days]
ND
VI
summer winter
Trees Trees
grasses
Soil Soil
SUMMER WINTER
summer winter
Safriel et al. (2010)
Pixel’s complexity – different contributions to NDVI
(fW)
�
grasses
Soil
Trees 1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
Decomposing NDVI time series into trees and grasses
NDVI = Trees + grasses + Soil (1)
NDVI = Trees + grasses + Soil Soil ≈ constant (1) (2)
(fw)
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
�
grasses
Soil
Trees
Decomposing NDVI time series into trees and grasses
(fw)
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
�
grasses
Soil
Trees
NDVI = Trees + grasses + Soil Soil ≈ constant (1) (2)
Decomposing NDVI time series into trees and grasses
(fw)
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
NDVI’
�
grasses
Soil
Trees
NDVI’ = Trees + grasses (3)
Decomposing NDVI time series into trees and grasses
(fw)
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
NDVI’
Plotting the minimum NDVI’ during the summer for the contribution of trees to NDVI (no grasses are present)
�
grasses
Soil
Trees
NDVI’ = Trees + grasses (3)
Decomposing NDVI time series into trees and grasses
(fw)
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
NDVI’
�
grasses
Soil
Trees
NDVI’ = Trees + grasses (3)
Trees = NDVI’min (in summer)
(4)
Trees
Decomposing NDVI time series into trees and grasses
grasses = NDVI’ – Trees (5)
(fw)
grasses
1 0.8 0.6 0.4 0.2 0
Time
ND
VI
2005 2006 2007 2008
NDVI’
�
grasses
Soil
Trees
Trees = NDVI’min (in summer)
(4)
NDVI’ = Trees + grasses (3)
Trees
Decomposing NDVI time series into trees and grasses
Murray – Darling Basin Queensland
Y = 150 X – 22.5
Lu et al. (2003)
FPC − Foliage Projection Cover (%)
FPC = a y NDVI + b
AVHRR 8 Km data
NDVImin = Soil (0% vegetation)
NDVImax = pixel of full vegetation (~100%)
Linear correlation between field measured data and NDVI from satellite
NDVImax − NDVImin FPC
NDVI Trees =
FPC can be estimated relatively:
Roderick et al. (1999)
NDVI – The biophysical meaning
NDVI
מ"ק 0 1 2 3 4
כרמל-דלית א
עוספיה
אוניברסיטת חיפה
עשבונים )חרוב מצוי, אלון מצוי(עד רחבי עלים -עצים ירוקי )אורן ירושלים(עד מחטניים -עצים ירוקי חקלאות קרקע
)סירה קוצנית(בני שיחים )אלת המסטיק, רתמה קוצנית, קידה שעירה(שיחים
Carmel et al. (2009)
בית אורן
טירת הכרמל טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
ניות
כרמל-דלית א
ל "כרם מהר
ניות
(Before December 2010)
> +25 %
0
> -25 %
Carmel Forest – patterns of tree cover (2001-2010)
Decadal change in tree canopy cover [%]
טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
כרמל-דלית א
ל "כרם מהר
ניות
> +25 %
0
> -25 %
Woody trend
Herbs
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.0
0.8
0.6
0.4
0.2
0
2006 Fire
(Before December 2010)
Carmel Forest – patterns of tree cover (2001-2010)
Decadal change in tree canopy cover [%]
טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
כרמל-דלית א
ל "כרם מהר
ניות
> +25 %
0
> -25 %
Woody trend
Herbs
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.0
0.8
0.6
0.4
0.2
0
2006 Fire 2010 Fire
Carmel Forest – patterns of tree cover (2001-2010)
(Before December 2010)
Decadal change in tree canopy cover [%]
טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
כרמל-דלית א
ל "כרם מהר
ניות
> +25 %
0
> -25 %
Woody trend
Herbs
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.0
0.8
0.6
0.4
0.2
0
2005 Fire
Carmel Forest – patterns of tree cover (2001-2010)
(Before December 2010)
Decadal change in tree canopy cover [%]
טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
כרמל-דלית א
ל "כרם מהר
ניות
> +25 %
0
> -25 %
Decadal change in tree canopy cover [%]
Woody trend
Herbs
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.0
0.8
0.6
0.4
0.2
0
2010 Fire
Carmel Forest – patterns of tree cover (2001-2010)
(Before December 2010)
בית אורן
טירת הכרמל טירת הכרמל אוניברסיטת חיפה
עוספיה
בית אורן
ניות
כרמל-דלית א
ל "כרם מהר
ניות
> +25 %
0
> -25 %
יער נטוע
Carmel Forest – patterns of tree cover (2001-2010)
(Before December 2010)
Decadal change in tree canopy cover [%]
ניר עציון
חיפה אוניברסיטת
ניות
טירת הכרמל חיפה אוניברסיטת
בית אורןGrasses
Decadal change in annual grasses
> +35 %
0
> -35 %
Grazing areas map
ניר עציון
בית אורן
]2003[אשכנזי . ש )ל"קק –ממשק הצומח המעוצה של הכרמל (
Carmel Forest – patterns of herbs cover (2001-2010)
(Before December 2010)
Paz S. et al. (2011)
בית אורן
אוניברסיטת חיפה טירת הכרמל
ניות
חיפה אוניברסיטת טירת הכרמל טירת הכרמל
חיפה אוניברסיטת
בית אורן בית אורן
Trees Grasses
> +25 %
0
> -25 %
Decadal change in tree canopy cover
Decadal change in annual grasses
> +35 %
0
> -35 %
December 2010 wildfire on Mt. Carmel and Fire-risk map
ניר עציון
ניר עציון ניר עציון
Carmel Forest – patterns of trees and grasses cover
Local Scale
1. Pre-fire mapping of fire-risk
2. Grazing management
3. Monitoring severe disturbances in forests (wildfires, tree
mortality etc.)
4. Monitoring biodiversity hotspots (vascular vegetation)
5. Climate change effect on forest structure
6. Patterns of change in vegetation in Mediterranean forests
7. Mapping climate-sensitive zones within the forests
Regional to global scale
Implications for management and research