[Link].

Sci (2025) 14(01): 316-320

International Journal of Current Microbiology and Applied Sciences

ISSN: 2319-7706 Volume 14 Number 1 (2025)
Journal homepage: [Link]

Original Research Article [Link]

Multiple Regression Model to Analyze the Effects of

Weather Variables on Crop Yield in the Kymore Plateau
and Satpura Hills
Navneet Raj Rathore*, S. S. Gautam and Suchita Jain

Department of Physical Science, MGCGVV, Chitrakoot, Madhya Pradesh-485334, India

*Corresponding author

ABSTRACT

Keywords This study investigates the influence of weather parameters on the yields of rice and wheat,
focusing on six key climate variables: maximum temperature, minimum temperature,
Multiple regression
model, yield, morning relative humidity, afternoon relative humidity, rainfall, wind direction, and wind
R square, RMSE, velocity in the Kymore Plateau and Satpura Hills. Using time series data from 2007 to
MAE, Percent 2023, the study employs multiple regression models to analyze the relationship between
deviation these climatic factors and crop yields. The results indicate that weather conditions
Article Info significantly impact rice and wheat production, highlighting the strong dependence of crop
productivity on meteorological variables. The findings further suggest that variations in
Received: climatic parameters play a crucial role in determining yield levels, alongside the influence
25 November 2024 of cultivated area on overall [Link] temperature and relative humidity in
Accepted:
30 December 2024
morning are negatively significantly influences the yield of rice and maximum temperature
Available Online: and wind velocity positively significantly influences and minimum temperature and wind
10 January 2025 direction are negatively significantly influences the yield of wheat.

Introduction The Kymore Plateau and Satpura Hills, located in the

state of Madhya Pradesh, represent ecologically diverse
Agricultural productivity is heavily influenced by agro-climatic zones that are particularly sensitive to
climatic and weather-related factors, especially in regions changes in temperature, rainfall, humidity, and other
where farming is largely dependent on natural rainfall climatic variables.
and seasonal variations.
In recent years, climate variability has posed significant
In India, where agriculture forms the backbone of the challenges to crop production in these regions, leading to
rural economy, understanding the impact of weather on fluctuating yields and increased vulnerability of farming
crop yield is essential for effective planning, risk communities. Accurate modeling and analysis of the
management, and sustainable agricultural development. relationship between weather parameters and crop yield

316
 [Link] (2025) 14(01): 316-320

can provide valuable insights for improving forecasting Source of data

models (Anik, et al., 2021), designing adaptive strategies,
and enhancing agricultural resilience. For the present research work secondary data of weather
variable and crop yield were collected from NASA Data
Among various crops, rice holds a crucial position as Access Viewer and Farmer welfare and Agriculture
India’s most important staple and the second most Development Department, Madhya Pradesh. Various
significant crop globally. Given the ongoing population meteorological parameters, including maximum
growth, there is a continuous need to enhance production temperature, minimum temperature, morning relative
to ensure sustainability, making climate change a key humidity, afternoon relative humidity, rainfall, wind
factor influencing rice yield variability. At specific direction, and wind velocity, were analyzed for the
growth stages, wheat is also influenced by climatic period from 2007 to 2023. This study focuses on two
factors, with some exerting restrictive effects. However, crops: rice and [Link] collected data have been
elements such as CO₂ levels, radiation, and temperature analysed with the help of suitable statistical techniques.
have been found to positively and significantly impact
wheat growth, with a direct correlation between these Multiple regression analysis
factors and wheat yield (Fuhrer, 2003). Climate change
refers to long-term shifts in temperature and weather Multiple Linear Regression (MLR) can be effectively
patterns, along with an increased frequency of extreme applied to time series data, provided key assumptions are
climatic events. The scientific and policy communities addressed. Techniques such as incorporating lagged
have acknowledged the vulnerability of crop agriculture variables, testing for autocorrelation, and ensuring
to these changes and have raised concerns about farmers' stationary help adapt MLR for temporal datasets. This
capacity to adapt, given the direct and substantial makes it suitable for analyzing relationships between
reliance of crop production on climatic conditions (Reid weather variables and crop yield over time.
et al., 2007; Roudier et al., 2011). In the current
demographic scenario, maintaining agricultural Multiple regression models
sustainability necessitates continuous increases in
production, with climate change playing a crucial role in
determining the yield variability of both rice and wheat.
Rice alone contributes to more than 60% of the total
value of crop agriculture (Asaduzzaman et al., 2010; Yu Where,
et al., 2010).
Y represents yield, while xi’s denotes weather
This research employs a multiple regression modeling parameters. b0 and bi’s are constants to be estimated, and
approach to quantify the effects of key weather e refers to the residual term.
variables—such as rainfall, temperature, and humidity—
on the yield of major crops grown in the Kymore Plateau Measures for the comparison and validation of
and Satpura Hills. By analyzing historical data and different models
identifying statistically significant relationships, this
study aims to contribute to more inform agricultural Coefficient of determination (R²)
decision-making, improved resource allocation, and
targeted interventions to mitigate the adverse effects of The coefficient of determination (R²) is a statistical
climatic fluctuations. measure that shows how well a model fits data and
predicts outcomes. It is a number between 0 and 1, and is
Materials and Methods often expressed as a percentage.

Kymore Plateau and Satpura Hills is nestled in the

Madhya Pradesh of India. The Agroclimatic zones lies
between latitude and longitude are 230 58' to 250 12' N
and 800 20' to 810 23' E respectively. It is located at an Where and is the value of the ith observation,
altitude ranging from approximately 300 to 340 meters estimated value of the ith observation and mean value of
above mean sea level. Y.

317
 [Link] (2025) 14(01): 316-320

Root Mean Square Error (RMSE) The performance of the model in predicting rice yield
was evaluated using multiple statistical indicators,
Root Mean Square Error (RMSE) is a statistical metric including the coefficient of determination (R²), Root
that measures how far predictions are from actual values. Mean Square Error (RMSE), Mean Absolute Error
(MAE), and percent deviation. The model achieved an R²
The RMSE is defined as value of 0.77, indicating that approximately 77% of the
variability in rice yield can be explained by the selected
meteorological variables. This suggests a strong
relationship between the input variables and the yield,
confirming the model’s effectiveness in capturing the
underlying patterns in the data.
Where is the value of the ith observation of y,
estimated value of the yi, n is the number of The RMSE value of 17.56 further supports the model’s
robustness, indicating a relatively low standard deviation
observations.
of the residuals (prediction errors). This implies that the
predicted values are generally close to the observed
(i) Mean Absolute Error (MAE) values, and the model fits the data well.
The MAE is defined as
Additionally, the MAE value of 4.39 indicates that, on
average, the absolute difference between the predicted
and actual rice yield values is 4.39 units, highlighting the
model’s accuracy and its capability to generate reliable
forecasts.
Where is the value of the ith observation of y,
estimated value of the yi, n is the number of The percent deviation was calculated at 62.44%,
observations. suggesting a notable variation between predicted and
actual values. Although higher than ideal, this may
Percent Deviation (PD) reflect the inherent variability in rice yield due to
unaccounted-for factors or noise in the dataset.
Nevertheless, the other performance metrics collectively
affirm the predictive reliability of the model.

Regarding the influence of climatic variables on rice

yield, the analysis revealed that maximum temperature,
Results and Discussion afternoon relative humidity, rainfall, and wind velocity
positively influenced rice yield. In contrast, minimum
Statistical models were developed by using sixteen-year temperature and morning relative humidity were found to
dataset (2007-2023) considering all the Weather have a significant negative influence at the 5%
variables, including maximum temperature, minimum significance level, indicating that lower nighttime
temperature, morning relative humidity, afternoon temperatures and higher morning humidity may
relative humidity, rainfall, wind direction, and wind adversely affect rice production. Furthermore, wind
speed. Here we consider yield as dependent variable and direction was also found to negatively influence yield,
maximum temperature, minimum temperature, relative albeit less significantly.
humidity in morning, relative humidity in afternoon,
rainfall, wind direction, wind velocity as independent These findings align closely with those of Kingra et al.,
variables for the above said two different crops. The (2019), who also reported similar climatic impacts on
multiple regression analysis was performed to determine rice yield. The consistency with previous literature
the effects of variables related to weather on crop yield. strengthens the validity of the current model and
The regression equations along with R2, RMSE, MAE underscores the critical role of agro-meteorological
and percent deviation values are represented in Table 3.1. parameters in rice yield prediction.

318
 [Link] (2025) 14(01): 316-320

Table.1 Multiple regression models on rice and wheat yield of Kymore Plateau and Satpura Hills

Model Model R2 RMSE MAE %

Deviation
Rice Y=18.54+2.76X1-5.42*X2-5.31*X3+4.64X4+0.01X5- 0.77 17.56 4.39 62.44
0.26X6+12.25X7
Wheat Y=32.49+2.17*X1-2.61*X2-2.25X3+1.41X4+0.01X5- 0.89 2.12 2.60 33.71
0.29**X6+7.99*X7
Note ** significance at 1 % and * significance at 5 %

The regression analysis for wheat yield prediction cooler nighttime temperatures may adversely affect
exhibited a strong model performance, as indicated by an wheat development. Wind direction exhibited a
R² value of 0.89, implying that 89% of the variability in significant negative influence at the 1% level, potentially
wheat yield can be explained by the selected agro- due to directional patterns associated with unfavorable
meteorological variables. This high coefficient of weather conditions or microclimate disturbances.
determination confirms that the model captures the Additionally, morning relative humidity was negatively
majority of the influential factors governing wheat associated with wheat yield, suggesting possible links to
production. increased disease pressure or suboptimal transpiration
dynamics during early hours.
The model’s predictive accuracy is further substantiated
by a Root Mean Square Error (RMSE) of 2.12, which This study underscores the significant influence of
demonstrates a low standard deviation of prediction weather parameters on the yields of rice and wheat in the
errors and suggests that the model's estimations closely Kymore Plateau and Satpura Hills region. By employing
align with the observed yield values. Additionally, the multiple regression models on time series data from 2007
Mean Absolute Error (MAE) was 2.60, reflecting a small to 2023, it is evident that climatic variables such as
average absolute deviation between predicted and actual temperature, humidity, rainfall, wind direction, and wind
yields. These low error metrics validate the model's velocity play a crucial role in determining crop
precision and suitability for reliable yield forecasting. productivity.

The percent deviation, calculated at 33.71%, denotes a Specifically, minimum temperature and morning relative
moderate level of deviation between predicted and actual humidity were found to have a significant negative
values. While this figure is higher than ideal, it is still impact on rice yield, while for wheat, maximum
within an acceptable range considering the variability temperature and wind velocity exhibited a significant
typically associated with biological and environmental positive influence. Conversely, minimum temperature
data. and wind direction were negatively associated with
wheat yield. These findings highlight the vulnerability of
In terms of variable influence, the results revealed that agricultural output to climatic variability and emphasize
maximum temperature and wind velocity have a the need for adaptive strategies in crop management and
significant positive influence on wheat yield at the 5% planning. Understanding these relationships can aid
level of significance. This suggests that warmer daytime policymakers, agronomists, and farmers in making
conditions and increased air movement during the informed decisions to enhance resilience and ensure
growing season may support improved physiological sustainable crop production in the face of changing
activity and yield outcomes. Furthermore, afternoon climatic conditions.
relative humidity and rainfall also exert a positive
influence on wheat production, likely by enhancing Author Contributions
moisture availability and reducing crop water stress
during critical growth stages. Navneet Raj Rathore: Investigation, formal analysis,
writing—original draft. S. S. Gautam: Validation,
Conversely, minimum temperature was found to have a methodology, writing—reviewing. Suchita Jain:—
significant negative effect at the 5% level, indicating that Formal analysis, writing—review and editing.

319
 [Link] (2025) 14(01): 316-320

Data Availability 2003;97(1-3):1-20

[Link]
The datasets generated during and/or analyzed during the Islam, F., Alam, G.M., Begum, R., Islam, M.N., &
current study are available from the corresponding author Bhandari, H. (2022). Farm level adaptation to
on reasonable request. climate change: insight from rice farmers in the
coastal region of Bangladesh. Local
Declarations Environment, 27(6), 671-681.
Kingra, P.K., Kaur, R., & Kaur, S. (2019). Climate
Ethical Approval Not applicable. change impacts on rice (Oryza sativa)
productivity and strategies for its sustainable
Consent to Participate Not applicable. management. The Indian Journal of Agricultural
Sciences, 89(2), 329-336.
Consent to Publish Not applicable. Kumar, K.S. Kavi, Parikh, J., 1998. Climate change
impacts on Indian agriculture: The Ricardian
Conflict of Interest The authors declare no competing approach. In: Dinar [Link]., Measuring the Impacts
interests. of Climate Change on Indian Agriculture. World
Bank Technical Paper No. 402, Washington D.C.
References Reid, S., Smit, B., Caldwell, W., Belliveau, S., 2007.
Vulnerability and adaptation to climate risk in
Anik, A.R., Rahman, S., Sarker, J.R., Al Hasan, M., & Ontario agriculture. Mitig. Adapt. Strat. Glob.
Bhandari, H. (2021). Farmers’ adaptation Change 12, 609–637.
strategies to combat climate change in drought [Link]
prone areas in Bangladesh. International Journal Roudier, P., Sultan, B., Quirion, P., & Berg, A. (2011).
of Disaster Risk Reduction, 65, 102562. The impact of future climate change on West
[Link] African crop yields: What does the recent
Asaduzzaman, M., Ringler, C., Thurlow, J., Alam, S., literature say? Global Environmental Change,
2010. Investing in Crop Agriculture in 21(3), 1073–1083.
Bangladesh for Higher Growth and Productivity, [Link]
and Adaptation to Climate Change. Bangladesh Yu, W., Alam, M., Hassan, A., Khan, A. S., Ruane, A.
Food Security Investment Forum, Dhaka. C., Rosenzweig, C., Major, D. C., & Thurlow, J.
Fuhrer J. Agroecosystem responses to combinations of (2010). Climate change risks and food security in
elevated CO2, ozone, and global climate change. Bangladesh. The World Bank.
Agriculture, Ecosystems & Environment. [Link]

How to cite this article:

Navneet Raj Rathore, S. S. Gautam and Suchita Jain. 2025. Multiple Regression Model to Analyze the Effects of
Weather Variables on Crop Yield in the Kymore Plateau and Satpura Hills. [Link]. 14(01):
316-320. doi: [Link]

320