Overview |
Crop Production |
Crop Pattern |
Crop Yield |
Irrigation |
Soil Management |
Precision Farming |
Relevant Products |
Relevant Links
Crop Area Estimation (Olive tree) Using Satellite Images
A Case Study in Tarom Region-Zanjan Province Iran
Mohsen Ahadnejad Reveshty
University of Zanjan, Dept. of Geography
Email: ahadnejad@mail.znu.ac.ir
Abstract
Olive is a low cost, valuable nutrient agriculture product of
high economic value. Olive grows between latitudes 30 to 45 degrees,
in very restricted climatic conditions (Mediterranean), in places
with mild winter (no frost condition), in poor soils. Food products
as shortening and canned olive may be used as a nutrient food.
Due to climatic condition of Tarom area in Zanjan Province, olive
plantation has recently been promoted in this area. The present
research aims to estimate olive farming area in 2002 using satellite
data. The processing techniques employed are as NDVI, supervised
classification and Tasseled Cape Index. Results of study disclosed
that olive-farming area covers about 3843 ha in this area.
Introduction
Land use and land cover is found as significant information in
an area that may assist managers and decision-makers to take drastic
measures. One of the most important land uses is agriculture lands
as well as orchards, which may play an important role in providing
man’s food. Crop yield forecast in an area usually requires
crop area estimation, which is mainly concerned by some relating
organizations.
Satellite data along with remote sensing technique may be employed
as a useful and effective tool to estimate crop area. Recent developments
of image processing techniques as well as availability of high
resolution satellite imageries avail to use this technique as
a quick and low cost method in compare to conventional methods
for crop area estimation while it is obvious that due to limited
period of growing season in case of some crops it is a very tedious
and difficult task to estimate crop area using time consuming
conventional methods.
Many studies have been conducted concerning the use of satellite
data for different crop type area estimation throughout the world.
Olive area estimation has been carried in a few olive-growing
countries as Turkey, Spain and Portugal. Ediz Unal (2004) used
Landsat 7 and IRS data to survey olive, pistachio and Vineyard
in Gazian-Tab area Turkey .He used image supervised classification
technique to estimate olive plantation area. Teresa Barata (2004)
has compared land sat images with aerial photo to map olive gardens
in Portugal.
Present study focuses on identification and mapping of olive in
Tarom-Zanjan region based on LANDSAT 7 ETM images using PCA analysis,
NDVI and supervised classification technique. Such study is being
conducted for the first time in our country in case of olive yards.
Study area
Study area is located between northern latitudes of 36 39 36 -37
01 48 and longitudes of 48 43 12- 49 18. The average elevation
of the study area is 700 meters above the sea level, covering
an area around 93210 ha. Administrative boundary of the study
area includes Tarom Township along northern portion of Zanjan
Province. From the physiographic point of view gently sloping
foothills consisting upper red formation as well as flat lying
alluvial of Qezel Uzan River cover the area. Mediterranean climatic
condition governing the area allows different crop type to be
grown in this area, so this area is namely called as Iranian India.
Furthermore unique climatic and physiographic condition governing
the area allows conventional olive farming along Qezel Uzan River
and its tributaries by conventional methods. During recent year’s
government policy as well as increasing demand is promoting more
farming of this crop and development of olive processing factories
in this region. Study area is shown in fig 1.
Fig 1: Case Study Area
Methods and Materials
Data:
LANDSAT 7 ETM 166-35 of 7 August 2002 was used to map olive farming
area. Topographic maps of 1: 25000 prepared by National Cartographic
Center are used for geo-referencing of land sat images. GPS was
also employed for training area selection in the field. The following
flow chart illustrates different research steps.
Fig 2: Research flowchart
Methodology:
Different image processing techniques are usually available to
highlight a certain land use. In present research four major techniques
were employed to highlight olive trees which are going to be described
in following.
Principal Component Analysis (PCA)
PCA analysis is one of the most commonly used techniques for remote
sensing data analysis. This technique helps for better discrimination
of different classes. False color composites were made of PCA1,
PCA2, and PCA3 using ENVI 4.0 software. Majorities of Olive trees
were highlighted as deep green color in this image. Olive framing
in the area is shown in Fig 3 and 4.
Fig 3 & 4: True false color composite of PCA1, PCA2, PCA3
Supervised Classification
Training area selection were accomplished based on field control
as well as topographic base maps of 1:25000 scale. Consequently
MLC classification was applied for final discrimination of olive
farming from other land uses. Finally classification results converted
from raster format to vector (Arc/view) for further area calculations
and map preparation. Using this method olive farming area was
calculated around 3843 ha. Fig 5 illustrates olive framing in
the study area.
Fig 5: spatial distribution of olive framing in year 2002, (MLC
classification)
NDVI Method
Normalized Vegetation Index was calculated using two spectral
bands of b4 and b3 in which vegetation cover has the highest reflectance.
Due to higher density of olive trees in this area as well as its
higher chlorophyll content in compare to other vegetation cover,
when threshold is applied over NDVI image, olive farming area
are better highlighted in compare to other crops. However considering
the fact that data used for this study belong to summer season,
therefore many other crops may reflect similar signature causing
interference, so the accuracy of this method would considerably
be lower comparing the classification method. Since olive tree
is an evergreen tree so satellite data of wintertime would give
more accurate results. Olive tree distribution map prepared by
NDVI method is shown in fig 6.
Fig 6: Olive tree distribution map prepared by NDVI method in
year 2002
Tasseled Cap Index
One of the other indices used for olive tree discrimination is
Tasseled Cap. Using these method parameters like brightness, greenness,
soil humidity and so on could be calculated. Greenness factor
induced by olive trees is considered in this method to estimate
olive tree distribution. Tasseled Cap index map applying a threshold
discriminated olive trees from other agriculture lands. Olive
tree distribution map prepared by Tasseled Cap method is shown
in fig 7.
Fig 7: olive tree distribution map using Tasseled Cap method
in year 2002
Conclusion and suggestions
Results of study revealed that PCA analysis is the most effective
method to increase discrimination factor among different classes.
Color composites of PCA1 PCA2, PCA3, consisting the majority of
information were used for training area selection. MLC classification
was employed to highlight olive farming area. Consequently olive
area estimated around 3843 ha using the two mentioned methods.
Also between the two indices Tasseled Cap indices reveals better
result in compare to NDVI method.
In order to address problems related to food security planners
using multidisciplinary decision support systems require among
other information on where and how much crops are accurately grown
in order to monitor agriculture production over vast area so high
resolution images as Ikonos and Quick Bird are suggested for more
accurate crop area estimation. Also satellite data acquired in
wintertime could give better results in case of olive farm mapping.