CROP-YIELD WEATHER RELATIONSHIP OF PRAKASAM DISTRICT IN ANDHRA PRADESH
Loading...
Files
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Acharya N.G. Ranga Agricultural University
Abstract
Rainfall is one of the important climatic factor that influences crop
production in particular and agriculture in general. Here, the distribution of
rainfall can be studied by fitting suitable statistical distribution to the monthly
rainfall data recorded over 15 years (2000-01 to 2014-15) for Prakasam district in
Andhra Pradesh. It revealed that most of the months follow ‘Type I’ (or) Beta
distribution while the months of July and September assumed Pearsonian “Type
II” (or) “Type VII” i.e., Normal distribution and the month of November shown
“Pearsonian Type IV” distribution.
Statistically, estimation of a crop yield-weather relationship is fitting a
multiple regression equation with yield as the dependent variable and weather
parameters during the crop growth period as the independent variables. The
analysis has been carried out on the basis of crop yields and weather variables for
15 years of monthly time series data (2000-01 to 2014-15). Weather impact on the
crop yields was studied on the basis of rainfall, temperature (maximum and
minimum) and relative humidity (AM and PM).
In fitting the crop yield-weather relationships, the assumption of a
continuous time trend was found to be inappropriate when the impact of new
technology may exists in the form of quantum jumps over time which is termed as
discrete time effect. For this situation, concept of control charts was applied using
one sigma limits and sub-periods were identified. These sub-periods were formed
with the year of quantum jump as the cut-off point. Nature of time trend in the
sub-periods and overall yields was investigated by fitting time trend regression
equation respectively. All the crops revealed ‘differential’ trend effect in the
yields of the two sub-periods indicating that there could be a differential weather
response of the crop. Hence, it is appropriate to fit the crop yield-weather
relationships separately for each of the two sub-periods as well as for overall
period.
It was observed that an overall relationship may not be appropriate to
explain the yield variations as it consisted of certain irrelevant regressors.
Considering this behaviour, separate relationships were fitted for the different subperiods
existing in the crop yield data and the analysis revealed the existence of a
differential response of the yields to weather. The variables identified in these
relations suitably explain the weather response with respect to the crop growth
stages. Hence, it was concluded that yields under a given technology only could
be forecasted (based on weather variables) on the basis of the corresponding subperiod
relationship (equation).
Description
D5301
Keywords
null