Miguel A. González-Botello Stephen H. Bullock and J. Mario Salazar Ceseña Comparison of Satellite and Ground Vegetation Indexes to Estimate Erosion in.
Post on 27-Mar-2015
216 Views
Preview:
Transcript
Miguel A. González-BotelloStephen H. Bullock and J. Mario Salazar Ceseña
Comparison of Satellite and Ground Vegetation Indexes to
Estimate Erosion in a Mediterranean-climate Watershed
Terrestrial Ecology LabConservation Biology Department
Centro de Investigaciones Científicas y de Estudios Superiores de Ensenada
No 2
Introductionw
Soil erosion is a natural process that reworks the distribution of organic matter, nutrients, and sediments.
Accelerated erosion is part of the desertification process: loss of productivity, stored carbon and biodiversity
No 3
Soil erosion across a 5,000 km2 watershed in NW Baja California has been estimated* using the Revised Universal Soil Loss Equation (RUSLE), in GIS.
Introduction
*Smith, S. V. et al. in press. Soil Erosion and its Potential Significance for Carbon Fluxes in a Mountainous Mediterranean-Climate Watershed. Ecological Applications
No 4
In a small area of southern France, DeJong* found a weak, linear relation of C to the Normalized Difference Vegetation Index (NDVI).
Introduction
*De Jong, S.M. 1994. Derivation of vegetative variables from a Landsat TM image for modelling soil erosion. Earth Surface Processes and Landforms, Vol. 19, 165-178.
C = 0.45 - 0.805 (NDVI)
(NIR – R)(NIR + R)NDVI=
No 5
Our project will evaluate the relation of C (measured in the field with precise methods and large samples distributed over
~3000 km2) to NDVI and to terrain variables (exposure, slope and elevation) and vegetation variables.
Our objective
No 6
Mediterranean-climate part of Mexico.Rainfall of 265 mmBetween November and April.
Three Terrestrial Prioritized Regions.Chaparral and coastal scrub.
Study area
No 7
GIS has been an important tool in designing the field work and will be essential to integration of ground and satellite data.
The location of the field sites (c. 67) involved the following seven steps:
Use of GIS in Site Selection
No 8
1. Update land use maps
No 9
2. Exclude non-shrub areas
Settled, agricultural and woodland or grassland/meadow areas
No 1
0
3.Digitize paved and dirt roads
No 1
1
4.Create Map of Accessible Areas
40 and 250 mFrom the roads
No 1
2
5.Compare frequency distributions
Watershed Accessible area
ASPECT ASPECT
NDVI* NDVI*
* May 2005
No 1
3
Watershed Accessible area
ELEVATION ELEVATION
SLOPE SLOPE
No 1
4
6.Select potential field sites
Ca. 120 sites. substantially uniform over more than 1 hectare regarding slope, exposure and vegetation (visual or NDVI).
No 1
5
Site selection involved High resolution images of Digital Globe (Google Earth), INEGI’s Ortophotos, NDVI, and Slope, Aspect & Height.
Google Earth ImagesDigital Globe
INEGI’s Digital Orthophoto Web Service (WMS)http://antares.inegi.gob.mx/cgi-bin/map4/mapserv_orto?)On ArcGIS 9
NDVI (May 2005)On ArcGIS 9
Aspect derived from Inegi’s DEMOn ArcGIS 9
Slope derived from Inegi’s DEMOn ArcGIS 9
No 1
6
7.Select among potential sites
We selected the sites to best represent the frequencies from the previous slides.
No 1
7
0
5
10
15
20
25
30
5 15 25 35 45 55
Category
Sites
Area
0
2
4
6
8
10
12
14
16
18
20
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Category
Sites
Area
Aspect (North)
Slope
South North
No 1
8
0
2
4
6
8
10
12
50 200 350 500 650 800 950 1100 1250 1400 1550 1700 1850
Category
Sites
Area
Height
0
5
10
15
20
25
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Category
Sites
AreaNDVI
No 1
9
With KML’er 1.2, extensive interaction between Google Earth 3 and ArcGIS 9 was possible.
Versatility
No 2
0
Field measurements were based on 30 m line transects.
Drip height, soil surface cover, and types of cover were recorded at 20 random points along the line.
Field Sampling Method
No 2
1
To assess aerial cover, we recorded the interception of each plant >20 cm diameter along the entire transect.
We also recorded plant species, height, and perpendicular diameter.
Field Sampling Method
No 2
2
Also, to better calibrate the erosion model, we collected samples of soil and litter.
Terrain variables were recorded to compare with estimates from the digital elevation model.
Field Sampling Method
No 2
3
Landsat data from late April-early May 2007 will be processed for NDVI (and EVI) and variance among years of contrasting rainfall will also be analyzed (2001, 2003, 2005).
Satellite ImageProcessing
No 2
4
Preliminary Results
0%
20%
40%
60%
80%
100%
0.1000 0.2000 0.3000 0.4000 0.5000 0.6000 0.7000 0.8000 0.9000 1.0000
Subfactor categories
% o
f si
tes
G Factor (Soil)
H Factor (Drip-height)
P Factor (Vegetation)
The major part of sites shows a high soil cover (G Subfactor, 0 – 0.1). This reduces substantially the field calculated C Factor.
No 2
5
0
10
20
30
40
50
60
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Subfactor categories
% o
f si
tes
C deJ
C in field
Preliminary Results
De Jong Model tend to overestimate the C factor.
No 2
6
y = 0.0266Ln(x) + 0.4303
R2 = 0.2566
0.0
0.1
0.2
0.3
0.4
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035
C[m]
C[J
on
g]
Our preliminary results suggests that the deJong Model is not suitable to assess erosion in Baja California Chaparral and Coastal Scrub.
No 2
7
Next stepsMost analyses are pending as field work is recently finished.
The acquisition of 2007 Landsat images are in process, the NDVI is not yet available.
No 2
8
Thanks to:
Conservation
Program
2007 Global Scholarship Program
Centro de Investigaciones Científicas y de Estudios Superiores de Ensenada
top related