All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Similar Documents

Information Report

Category:
## Humor

Published:

Views: 0 | Pages: 4

Extension: PDF | Download: 0

Share

Description

Pre-harvest crop modelling of kharif rice using weather parameters in Valsad district of south Gujarat

Tags

Transcript

Pre-harvest crop modelling of
kharif
rice using weather parameters in Valsad district of south Gujarat
K. B. BANAKARA
1
,H. R. PANDYA
1
,Y. A. GARDE
2
andS.OJHA
1
1
Department of Agricultural Statistics, Navsari Agricultural University, Navsari, Gujarat
2
Department of Agricultural Statistics, Navsari Agricultural University, Waghai, Gujarat Email: kantheshbanakara@gmail.com
ABSTRACT
Rice is the most important staple food in India, which play crucial role in daily requisite of diet. In the Gujarat state, rice occupies about 7-8 per cent of the gross cropped area and accounts for about 14 per cent of the total food grain production. In the present study statistical forecasting models were employed to provide forecast before harvest of crop for taking timely decisions. In this paper Multiple Linear Regression (MLR) technique was utilized for estimating average rice production in Valsad district of South Gujarat. The weather indices were developed utilizing week number as weight by weekly weather parameters for the year 1975 to 2010 and for the cross-validation of the developed forecast models were tested by utilizing data from 2010 to 2014. It is observed that value of Adj. R
2
varied from 55.8 to 61.6 in different models. Based on the findings in the present study, itwas observed that model-5 found to be better than all other models for pre harvest forecasting of rice crop yield.
Keywords:
Weather indices, MLR techniques, Forecast Indian economy is mainly based on agriculture and rice is the most important staple food in India as well as Asia. More than 90 per cent of the
world’s rice is grown
and consumed in Asia, where 60
per cent of the world’s population lives. India ranks
second with 154.6 million tonnes of paddy (FAO, 2015). In the Gujarat state, rice occupies about 7 - 8 per cent of the gross cropped area of the state and accounts for around 14 per cent of the total food grain production. About 90 per cent of area under rice is confined to South and middle Gujarat. The productivity of the crop was highly influenced by weather parameters. Thus, development of forecasting models based on weather parameter for rice crop is very important role and developed pre-harvest forecast models were utilized in making policy decision regarding export and import, food procurement and distribution, price policies and exercising several administrative measures for storage and marketing of agricultural commodities. Thus, the use of statistical models in forecasting food production and prices for agriculture hold great significance. Although nostatistical model can help in forecasting the values exactly but by knowing even approximate values can help in formulating future plans.
MATERIALS AND METHODS
The investigation was carried out in Navsari Agricultural University, Navsari. The study utilized secondary yearly yield data (
Kharif
season) and weekly weather data for 39 years (1975-2014) which were collected from the Directorate of Economics and Statistics, Government of Gujarat, Gandhinagarand Indian Meteorological Department (IMD),Punerespectively. Five weather parameters were included in investigation
viz.
maximum temperature (
X
1
), minimum temperature (
X
2
), relative humidity (
X
3
), wind speed (
X
4
)and rain fall (
X
5
). However, weekly weather data related to
kharif
crop season starting from one month before sowing up to one month before harvest of crop (22
nd
to 37
th
Standard meteorological week (SMW)) was utilized for development of statistical model.The data of last one month of crop season was excluded keeping inview thatforecast crop yield at least one month before harvest.The association between yearly crop yield and different weekly weather parameters were studied by Karl-Person correlation coefficient approach.Multiple Linear Regressions (MLR) are widely suitable for short or intermediate term forecasting. In present study, MLR was used for developing forecasting models using predictors as appropriate un-weighted and weighted weather indices (Agrawal
et al
., 1980; Jain
et al.,
1980; Agrawal
et al.,
1986; Garde
et al.,
2012; Rajegowda
et al.,
2014; Singh
et al.,
2014; Dhekale
et al.,
2014;and Singh and Sharma, 2017). Weather indices were developed by using week number as weight.
Development of weather indices
J ournal of Agrometeorology 19 (Special Issue - AGMET 2016) :
1
96-199
(October 2017)
Where,
Z
ij
is the developed weather indices of
i
th
weather parameter for
j
th
weight.
Z
ikj
is the developed weather indices of product of
i
th
and
k
th
weather parameter for
j
th
weight.
Q
ij
is un-weighted (for
j
=0) and weighted(for
j
=1) weather indices for
i
th
weather parameter
Q
ikj
is the un-weighted (for
j
=0) and weighted (for
j
=1) weather indices for interaction between
i
th
and
k
th
weather parameters.
X
iw
is the value of the
i
th
weather parameter in
w
th
week.
m
is week of forecast. k=
i
= 1,2,...,
p
,
j
=0,1. and
w
=1,2,...,m.
Development of models 1.Model-1
This model was developed by using srcinalvariables and interaction between srcinal variables. The developed model was
01 1
p pi i ik i k i i k
Y A a X a X X cT e
Where,
Y
is the observed rice yield.
A
0
is the general mean.
X
i
and
X
k
are the weather parameters.
p
is number of weather parameters used.
T
is the trend parameter and
c
is the regression coefficients of trend parameter,
e
is the error term.
2.Model-2
This model was developed by using first andsecond developed weather indices, only weighted variables were used to develop the model. The developed model was
0 1 1 1 11 1
p pi i ik ik i i k
Y A a Z a Z cT e
Where,
Y
is the observed rice yield.
A
0
is the general mean.
Z
ij
and
Z
ikj
are the weather indices.
a
ij
and
a
ikj
are the regression coefficients of
Z
ij
and
Z
ikj
weather indices.
p
is number of weather parameters used.
T
is the trend parameterand
c
is the regression coefficients of trend parameter.
e
is the error term.
3.Model-3
This model was developed by using third andfourth developed weather indices, only weighted variables were used to develop the model. The developed model was
0 1 1 1 11 1
p pi i ik ik i i k
Y A a Q a Q cT e
Where,
Y
is the observed rice yield.
A
0
is the general mean.
Q
ij
and
Q
ikj
are the weather indices.
a
ij
and
a
ii'j
are the regression coefficients of
Q
ij
and
Q
ikj
weather indices.
p
is number of weather parameters used.
T
is the trend parameter and
c
is the regression coefficients of trend parameter.
e
is the error term.
4.Model-4
This model was developed by using firstand second developed weather indices, both un-weighted and weighted variables were used to develop the model. The developed model was
1 101 0 1 0
p pij ij ikj ikji j i k j
Y A a Z a Z cT e
Where,
Y
is the observed rice yield.
A
0
is the general mean.
Z
ij
and
Z
ikj
are the weather indices.
a
ij
and
a
ikj
are the regression coefficients of
Z
ij
and
Z
ikj
weather indices.
p
is number of weather parameters used.
T
is the trend parameter and
c
is the regression coefficients of trend parameter.
e
is the error term.
5.Model-5
This model was developed by using third and fourth developed weather indices, both un-weighted and weighted variables were used to develop the model. The developed model was
1 101 0 1 0
p pij ij ikj ikji j i k j
Y A a Q a Q cT e
Where,
Y
is the observed rice yield.
A
0
is the general mean.
Q
ij
and
Q
ikj
are the weather indices.
a
ij
and
a
ikj
are the regression coefficients of
Q
ij
and
Q
ikj
weather indices.
p
is number of weather parameters used.
T
is the trend parameter and
c
is the regression coefficients of trend parameter.
e
is the error term.
Comparison and Validation of Models
The models were compared on the basis of Forecast error:
ForecastError 100
i ii
O E O
Where,
O
i
the
E
i
are the observed and forecasted values of crop yield, respectively. The models were compared on the basis of adjusted coefficient of determination
R
2adj
as follows: Where,
ss
res
/(n-p)
is the residual mean square
ss
t
/(n-1)
is the total mean sum of square. From the fitted models, rice yield were forecasted for the years 2011-12 to 2014-15 and were compared on the basis of Root Mean Square Error (RMSE).
2
( )1( 1)
resadjt
SS n p RSS n
1221
1( )
ni ii
RMSE O E n
197Vol. 19, Special Issue (AGMET 2016)
Table 1
Yield Forecasting Models
Model NameWeek No.ModelAdj. R
2
Model-136
250 350 10
3644.0 19.64 0.042 0.009 5.24
Y T Z Z Z
55.8Model-232
451 21 121
2011.50 36.56 0.005 3.53 0.06
Y T Z Z Z
60.5Model-332
451 21 121
2011.45 36.57 0.317 233.23 3.60
Y T Q Q Q
60.5Model-432
451 20 11
975.05 35.22 0.004 11.02 1.46
Y T Z Z Z
61.6
Model-532
451 20 11
975.05 35.22 0.29 121.25 96.54
Y T Q Q Q
61.6
Table 2
Comparison between yield forecasting models
Model SMWYearForecast YieldActual YieldForecast error (%)RMSEAdj. R
2
Model-13620102058263421.87644.1055.820111965257523.7020121915236418.9720132003243117.6020141896229217.26Model-23220101987263424.56659.1060.520111949257524.2820121836236422.3220132032243116.4220142055229210.34Model-33220102255263414.39352.2760.520112216257513.9620122100236411.162013225824317.10201422972292-0.23Model-43220102227263415.44394.80
61.6
20112175257515.5320122058236412.932013221324318.962014225722921.52Model-53220102239263415.00
381.4561.6
20112185257515.1420122063236412.722013223524318.042014228122920.49
Where,
O
i
and the
E
i
are the observed and forecasted values of crop yield, respectively and n is the number of years for which forecasting will be done. Selection of model was made based on highest Adjusted R
2
value and lowest RMSE and forecast error value among the method.
RESULT AND DISCUSSION
The best model was identified on the basis of adjusted coefficient of determination, (Adj. R
2
). The forecasting models were developed at different time periods
i.e.
32
th
week onwards at weekly interval which shown in Table 1. The values of adj. R
2
were varied from 55.8 per cent in model-1 to 61.6 per cent for model-4 and model-5 which indicates model-4 and model-5 are best among all other models. This 61.6 per cent variation accounted by weather indices
T, Z
451
, Z
20
and
Z
11
for model-4 and
T, Q
451
, Q
20
and
Q
11
for model-5. Thus, the model using un-weighted and weighted indices was found to be appropriate in 32
th
SMW (ten weeks before harvest of crop). Comparison of models was made by using values of root mean squared error (RMSE) and forecast error (Fig. 5). The validation of the models was done only for those showed meaningful results (for all models). The comparison of results given in Table 2 showed that for forecasting rice yield, model-
198October 2017
3was better with lower RMSE value of 352.27 ascompared to all other models. But Adj. R
2
was less compared model-4 and model-5 with RMSE value of 394.80 and 381.45 which slight difference with mode-3. Both model-4 and model-5 yields same Adj. R
2
but model-5 has lower RMSE and forecast error (0.49 to 15.14) value compared to model-4 hence model-5 was suitable for Valsad district of South Gujarat.
Fig1:
Comparison of expected and actual rice yield using model-5
CONCLUSION
Using the forecast model, pre-harvest estimates of rice crop yield for Valsad district could be computed successfully before ten weeks of actual harvest
i.e.
during panicle initiation stage of the crop period. The weather variables involved in models were interaction between wind speed and rainfall, minimum temperature and maximum temperature.
REFERENCES
Agrawal, R.;Jain, R. C.;Jha, M. P. and Singh, D. (1980). Forecasting of rice yield using climatic variables.
Indian Journal of Agricultural Science
.50(9): 680-684. Agrawal, R.;Jain, R. C. and Jha, M. P. (1986). Modes for studying rice-weather relationship.
MAUSAM
.37(1): 67-70. Dhekale, B. S.; Mahdi, S. and Sawant, P. K. (2014). Forecast models for groundnut using meteorological variables in Kolhapur,Maharashtra.
Journal of Agrometeorology
, 16(2): 238-239.Food and Agriculture Organization (2015). Rice market monitor, report 18(2): 2-6.Garde, Y. A.; Shukla, A. K. and Singh, S. (2012).Pre-harvest forecasting of rice yield using weather indices in Pantnagar.
International Journal of Agricultural Statistical Science.
8(1): 233-241. Jain, R. C.; Agrawal, R. and Jha, M. P. (1980). Effect of climatic variables on rice yield and its forecast,
MAUSAM
,31(3): 591-596. Rajegowda, M. B.; Soumy, D. V.; Padmashri, H. S.; Janardhanagowda, N. A. and Nagesha, L. (2014). Ragi and groundnut yield forecasting in Karnataka
–
statistical model.
Journal of Agrometeorology
16 (2): 203-206. Singh, M. and Sharma, S. (2017). Forecasting the maize yield in Himachal Pradesh using climatic variables.
Journal of Agrometeorology
19 (2): 167-169. Singh, R. S.; Patel, C.; Yadav, M. K. and Singh, K. K. (2014).Yield forecasting of rice and wheat crops for eastern Uttar Pradesh.
Journal of Agrometeorology
16(2): 199-202.
199Vol. 19, Special Issue (AGMET 2016)

Recommended

Related Search

Role of Pre-harvest Factors In Post-harvest LModelling of Solar Sell With Using of Matlab Pre-Hispanic Complex Cultures of the AndesModelling of Concrete Transport ProcessesModelling of Preference Data (particularly FoConstitutive modelling of unsaturated soilsAnalysis and modelling of nonlinear systemsMathematical Modelling of Biological SystemsGIS-based weights-of-evidence modelling of raMathematical modelling of processes

We Need Your Support

Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...Sign Now!

We are very appreciated for your Prompt Action!

x