monitoring agriculture using remote sensing at the sebele dar farm

Report
MONITORING AGRICULTURE USING
REMOTE SENSING AT THE SEBELE
DAR FARM
PRESENTATION
BY
MMOLOKI G. MOLWANTWA
Introduction
• The issue of monitoring crop performance and
productivity is one of the major challenges facing
the agricultural sector in Botswana. Often
difficulties arise in monitoring the agricultural
sector in pastoral or arable commercial farms
especially where there is limited in situ data.
Remote sensing has been identified to be used
widely for monitoring agriculture in areas where
there is limited in situ data and also offers
farmers the opportunity to evaluate the crop
productivity level of large areas more efficiently
and cost effectively (Flynn, 2008).
Need of the study
• Monitoring is an important aspect in
agriculture because with it we can predict
seasonal outlook of crops.
• In Botswana monitoring crop performance
and productivity using remote sensing is not
done at a satisfactory level, therefore a need
then arises to monitor crop performance and
productivity using remote sensing.
Aim
• The aim of this study was to use remote
sensing products to monitor crop
performance in the Department of Agriculture
Research Content Farm in Sebele.
Methodology
• Study Area (Sebele DAR Farm, Gaborone Botswana)
Methodology cont`d
Shape file
Methodology cont`d
• Data requirements
Rainfall products
The rainfall products were derived from FEWSNET
Rainfall Estimation (RFE) dekadal imagery.
Yield data
The yield data was obtained from the DAR archive. This
was record of yield from the 2008-09 growing season
up to the 2011-2012 cropping season.
Methodology cont`d
Vegetation products
The vegetation products were derived from the SPOTVegetation NDVI at 1km spatial resolution provided by
DevCoCast (www.devcocast.eu). The data is obtained
from the BCA remote sensing station which is under
African Monitoring of the Environment for Sustainable
Development (AMSED) Organisation.
Methodology cont`d
Shape file and NDVI Overlay
Methodology cont`d
• Data analysis
Regression analysis model for long term average
NDVI against long term average rainfall for the
years 2008-2010 were made using Microsoft
excel whereby the rainfall was an independent
variable and NDVI a dependent variable.
Results
Long term average VS current average NDVI
0.700
0.600
NDVI values
0.500
0.400
current average(2011)
0.300
long term avg(2008-2010)
0.200
0.100
0.000
1
6
11
16
21
Dekadals
26
31
36
Results cont`d
Image a
November 7 2011
Image b
November 13 2011
Results cont`d
Image c
November 18 2011
Image d
December 10 2011
Results cont`d
Image e
December 20 2011
Image f
December 15 2011
Results cont`d
NDVI long term average and Rainfall time series for DAR Farm(2008-2010)
0.60
70
60
0.50
50
long term average
rainfall(2008-2010)
40
0.30
30
0.20
rainfall (mm)
NDVI values
0.40
long term avg(2008-2010)NDVI
20
0.10
10
0.00
0
1
6
11
16
Dekadals
21
26
31
36
Results cont`d
0.70
70
0.60
60
0.50
50
0.40
40
0.30
30
0.20
20
0.10
10
0.00
0
1
4
7
10
13
16
19
Dekadals
22
25
28
31
34
Rainfall (mm)
NDVI values
Current Rainfall and NDVI for 2011
Current Average
Rainfall( 2011)
Current Average
NDVI(2011)
Results cont`d
Yield and NDVI time series
25
0.368
0.366
20
0.364
0.362
NDVI values
Yield (tonnes)
15
10
0.360
0.358
0.356
Long term avg NDVI
0.354
5
0.352
0.350
0
0.348
2008/2009
2009/2010
2010/2011
Cropping seaon
2011/2012
Yield (tonnes)
Results cont`d
RFEs and Recorded Longterm Average Rainfall for DAR farm(2008-2010)
70
60
50
rainfall (mm)
40
30
20
10
01
4
7
10
13
Long term average RFES
16
19
recorded long term average rainfall
22
25
28
31
34
Results cont`d
long term average rainfall(2008-2010) Line Fit Plot
1.00
0.90
long term avg(2008-2010)NDVI
0.80
0.70
0.60
long term avg(2008-2010)NDVI
0.50
0.40
0.30
Linear (Predicted long term
avg(2008-2010)NDVI)
0.20
0.10
0.00
0.0
10.0
20.0
30.0
40.0
long term average rainfall (mm)
50.0
60.0
70.0
Conclusion
• The aim of this project was to monitor
agriculture in DAR farm using remote sensing
products. From the results it shows that
monitoring has been achieved as rainfall
estimates and NDVI can be used to make
inferences about crop performance.
Recommendations
• The commercial farms particularly those that
produce large crop outputs should adopt the
use of remote sensing products to monitor
crop performance in Botswana.

similar documents