Atish Sagar

Improving the forecasting accuracy of agricultural commodity prices is critical for many stakeholders namely, farmers, traders, exporters, governments, and all other partners in the price channel, to evade risks and enable appropriate policy interventions. However, the traditional mono-scale smoothing techniques often fail to capture the non-stationary and non-linear features due to their multifarious structure. This study has proposed a CEEMDAN (Complete Ensemble Empirical Mode Decomposition with Adaptive Noise)-TDNN (Time Delay Neural Network) model for forecasting non-linear, non-stationary agricultural price series. This study has evaluated its suitability in comparison with the other three major EMD (Empirical Mode Decomposition) variants (EMD, Ensemble EMD and Complementary Ensemble EMD) and the benchmark (Autoregressive Integrated Moving Average, Non-linear Support Vector Regression, Gradient Boosting Machine, Random Forest and TDNN) models using monthly wholesale prices of major oilseed crops in India. Outcomes from this investigation reflect that the CEEMDAN-TDNN hybrid models have outperformed all other forecasting models on the basis of evaluation metrics under consideration. For the proposed model, an average improvement of RMSE (Root Mean Square Error), Relative RMSE and MAPE (Mean Absolute Percentage Error) values has been observed to be 20.04%, 19.94% and 27.80%, respectively over the other EMD variant-based counterparts and 57.66%, 48.37% and 62.37%, respectively over the other benchmark stochastic and machine learning models. The CEEMD-TDNN and CEEMDAN-TDNN models have demonstrated superior performance in predicting the directional changes of monthly price series compared to other models. Additionally, the accuracy of forecasts generated by all models has been assessed using the Diebold-Mariano test, the Friedman test, and the Taylor diagram. The results confirm that the proposed hybrid model has outperformed the alternative models, providing a distinct advantage.

The management of water resources is a priority problem in agriculture, especially in areas with a limited water supply. The determination of crop water requirements and crop coefficient (K_c) of agricultural crops helps to create an appropriate irrigation schedule for the effective management of irrigation water. A portable smart weighing lysimeter (1000 × 1000 mm and 600 mm depth) was developed at CPCT, IARI, New Delhi for real-time measurement of Crop Coefficient (K_c) and water requirement of chrysanthemum crop and bulk data storage. The paper discusses the assembly, structural and operational design of the portable smart weighting lysimeter. The performance characteristics of the developed lysimeter were evaluated under different load conditions. The K_c values of the chrysanthemum crop obtained from the lysimeter installed inside the greenhouse were K_c ini. 0.43 and 0.38, K_c mid-1.27 and 1.25, and K_c end-0.67 and 0.59 for the years 2019–2020 and 2020–2021, respectively, which apprehensively corroborated with the FAO 56 paper for determination of crop coefficient. The K_c values decreased progressively at the late-season stage because of the maturity and aging of the leaves. The lysimeter’s edge temperature was somewhat higher, whereas the center temperature closely matched the field temperature. The temperature difference between the center and the edge increased as the ambient temperature rose. The developed smart lysimeter system has unique applications due to its real-time measurement, portable attribute, and ability to produce accurate results for determining crop water use and crop coefficient for greenhouse chrysanthemum crops.

Nowadays, great attention has been attributed to the study of runoff and its fluctuation over space and time. There is a crucial need for a good soil and water management system to overcome the challenges of water scarcity and other natural adverse events like floods and landslides, among others. Rainfall–runoff (R-R) modeling is an appropriate approach for runoff prediction, making it possible to take preventive measures to avoid damage caused by natural hazards such as floods. In the present study, several data-driven models, namely, multiple linear regression (MLR), multiple adaptive regression splines (MARS), support vector machine (SVM), and random forest (RF), were used for rainfall–runoff prediction of the Gola watershed, located in the south-eastern part of the Uttarakhand. The rainfall–runoff model analysis was conducted using daily rainfall and runoff data for 12 years (2009 to 2020) of the Gola watershed. The first 80% of the complete data was used to train the model, and the remaining 20% was used for the testing period. The performance of the models was evaluated based on the coefficient of determination (R²), root mean square error (RMSE), Nash–Sutcliffe efficiency (NSE), and percent bias (PBAIS) indices. In addition to the numerical comparison, the models were evaluated. Their performances were evaluated based on graphical plotting, i.e., time-series line diagram, scatter plot, violin plot, relative error plot, and Taylor diagram (TD). The comparison results revealed that the four heuristic methods gave higher accuracy than the MLR model. Among the machine learning models, the RF (RMSE (m³/s), R², NSE, and PBIAS (%) = 6.31, 0.96, 0.94, and −0.20 during the training period, respectively, and 5.53, 0.95, 0.92, and −0.20 during the testing period, respectively) surpassed the MARS, SVM, and the MLR models in forecasting daily runoff for all cases studied. The RF model outperformed in all four models’ training and testing periods. It can be summarized that the RF model is best-in-class and delivers a strong potential for the runoff prediction of the Gola watershed.

Nowadays, great attention has been attributed to the study of runoff and its fluctuation over space and time. There is a crucial need for a good soil and water management system to overcome the challenges of water scarcity and other natural adverse events like floods and landslides, among others. Rainfall-runoff modeling is an appropriate approach for runoff prediction, making it possible to take preventive measures to avoid damage caused by natural hazards such as floods. In the present study, several data driven models, namely: Multiple linear regression (MLR), Multiple adaptive regression splines (MARS), Support vector machine (SVM), and Random Forest (RF), were used for rainfall-runoff prediction of the Gola watershed, located in the south-eastern part of the Uttarakhand. The performance of the models was evaluated based on the coefficient of determination (R2), root mean square error (RMSE), Nash-Sutcliffe efficiency (NSE), and percent bias (PBAIS) indices. In addition to the numerical comparison, the models were evaluated and their performances were evaluated base on graphical plotting, i.e., line diagram, scatter plot, Violin plot, relative error plot and Taylor diagram (TD). The comparison results revealed that the four heuristic methods gave higher accuracy than the MLR model. Among the machine learning models, the RF (RMSE (m3/s), R2, NSE, and PBIAS (%) = 6.31, 0.96, 0.94, and -0.20 during the training period, respectively, and 5.53, 0.95, 0.92, and -0.20 during the testing period, respectively) surpassed the MARS, SVM, and the MLR models in forecasting daily runoff for all cases studies. Among all four models, the RF model outperformed in the training and testing periods. It can be summarized that the RF model is best-in-class and delivers a strong potential for runoff prediction of the Gola watershed.

Mahanadi is one of the major inter-state east flowing perennial rivers in peninsular India. Hamp watershed of Seonath Sub-basin of upper Mahanadi basin was considered for the study to estimate the sediment yield and nutrient loss-based identification of critical agricultural sub-watershed and its critical Hydrological Response Unit (HRU) using Soil and Water Assessment Tool (SWAT) in-terfaced with GIS i.e., ArcSWAT. The study area was divided into 14 sub-watersheds considering topographical parameters derived from DEM and drainage network. The land cover, soil layers, and DEM were used to generate 207 HRUs for analysis of annual runoff, sediment yield and nu-trient loss for 2004-2008 (calibration period) and 2010-2013 (validation period). The sediment yield, runoff estimation and nutrient loss matched consistently well with the monthly and seasonal measured values. On the basis of average annual sediment yield (18.18 t/ha), runoff (245.97 mm) and nutrient loss NO3-N (1.62 kg/ha), respectively, sub-watershed WS4 was categorized under high priority for critical are identification. The sub watershed WS4 comprises of 15 HRUs (No. 36 - 50) with four kharif crops viz rice, soybean, maize and sugarcane. Results showed that the crops soy-bean, maize and sugarcane reduced the average annual runoff by 18.1, 31.4 and 18.0 per cent, respectively whereas the sediment yield was increased drastically by 104.5, 37.5 and 5.7 per cent, respectively as compared to rice. Soybean and maize crops HRU generate significant amount of soil and nutrient loss and were found to be as the critical HRUs for the upper Mahanadi River basin

Rainfall prediction is clearly of great importance for any country. One would like to make long term prediction, i.e. predict total monsoon rainfall a few weeks or months and in advance short term prediction, i.e. predict rainfall over different locations a few days in advance [1]. Predicted by using its correlation with observed parameter. Several regression and neural network based models are currently available. While Artificial Neural Network provide a great deal of promise, they also embody much uncertainty [2,3]. In this paper, different artificial neural network models have been created for the rainfall prediction of Uttarakhand region in India. These ANN models were created using training algorithms namely, feed-forward back propagation algorithm [4,5]. The number of neurons for all the models was kept at 10. The mean squared error was measured for each model and the best accuracy was obtained by the feed-forward back propagation algorithm with MSE value as low as 0.00547823.