Thumbnail Image

Theses

Browse

2016 - 201922020 - 20211
Reset filters
Author: Kushwaha, Daniel Prakash × Start date: 2010 ×
Reset filters

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    ThesisItemOpen Access
    Effect of soil amendments on runoff, sediment yield, biomass, soil physico-chemical properties and loss of major nutrients from sloping land under natural and simulated rainfall conditions
    (G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand), 2021-06) Kushwaha, Daniel Prakash; Anil Kumar
    In this research, experimental study was made in three successive trials, viz. first and second trials were conducted under natural rainfall conditions in Mollisols soils of Pantnagar in the foothills of North Western Himalayas during monsoon season of 2018 and 2019, respectively; and the third trial was conducted in simulated rainfall conditions just after completion of natural experiments. During first trial under natural conditions, a pre-determined dose of soil amendments was applied, while in second trial an increased dose of amendments was applied to check their incremental effect. Three soil amendments viz. biochar, anionic PAM and gypsum were used to fix six treatments: control (C); gypsum (G); biochar (B); gypsum and polyacrylamide (G+PAM); biochar and polyacrylamide (B+PAM); and biochar and gypsum (B+G) and each treatment was replicated thrice on plots of size 3m × 3m using randomized block design on uniform land slope of 12%. Hypothesis was whether six treatments could reduce surface runoff, sediment yield, loss of major soil nutrients (N, P and K), and maintain and improve soil physico-chemical properties and above- and below-ground biomass or not? (B+PAM) treatment of second trial, in which biochar was applied @ 1500 g/m2 and PAM was applied @ 4 g/m2, was found to be the best treatment under natural conditions. Therefore, this particular treatment was considered again under simulated rainfall conditions, but this time variable land slopes of 0%, 6%, 12%, 18%, and 24% on plots of size 1m × 3m and variable application rates of amendments along with five simulated rainfall intensities viz. 7.06, 9.07, 11.05, 12.97 and 14.96 cm/h were applied. Trial under simulated rainfall conditions was completed in four consecutive stages, mainly control (no amendments), biochar mixed with soil @ 1000 g/m2, 1500 g/m2 and 2000 g/m2. In each stage, PAM was applied @ 3, 4 and 5 g/m2. Available N-P-K losses, biochar loss, sediment yield and runoff were taken into account in simulated rainfall conditions. In simulated experiment, low dose of anionic PAM as 3 g/m2 (0.03 t/ha) in all conditions and high dose of biochar as 2000 g/m2 (20 t/ha) was found to be acceptable for reduction of runoff, sediment yield and major nutrient losses. In the loss of biochar study, 1500 g/m2 (15 t/ha) was found acceptable instead of high dose 2000 g/m2 (20 t/ha) for reduction of biochar loss. In this study, three reasonable model scenarios of sediment yield, nutrient loss and biochar loss were also developed using the experimental data of both conditions. For these model scenarios, four modeling techniques viz. MLP-ANN, SVM-RKF, SVM-LKF and MLR were used. Under natural conditions, five variables of the best treatment viz. rainfall, runoff, sediment yield, biochar loss and nutrients (N, P and K) loss were selected, while under simulated conditions, seven variables viz. rainfall, runoff, sediment yield, biochar loss, nutrients (N, P and K) loss, application rate of amendments, and land slope were selected for the development of model scenarios. On the basis of modeling results under natural and simulated conditions, it was found that SVM-LKF model performed well in comparison to other models in simulating event based data.
  • Thumbnail Image
    ThesisItemOpen Access
    Suspended sediment yield modelling using artificial neural networks
    (G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand), 2016-06) Kushwaha, Daniel Prakash; Devendra Kumar
  • Thumbnail Image
    ThesisItemOpen Access
    Suspended sediment yield modelling using artificial neural networks
    (G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand), 2016-06) Kushwaha, Daniel Prakash; Devendra Kumar
    Eight artificial neural networks based models were developed to predict daily suspended sediment concentration for the Baitarani river at Anandpur and Champua stations using daily discharge and daily suspended sediment concentration. The 30 years data (June 1977 to September 2006) used in this study was divided into two sets viz. a training set (1977-1996) and a testing set (1997-2006). ANNs models were calibrated by using multilayer feedforward back propagation neural networks with sigmoid activation function and Levenberg-Marquardt (L-M) learning algorithm. The performance of the developed models was evaluated qualitatively and quantitatively. In qualitative evaluation of models, the observed and the computed suspended sediment concentration were compared using sediment hydrographs and scatter plots during testing period. Akaike’s information criterion (AIC), correlation coefficient (r), mean square error (MSE), root mean square error (RMSE), minimum description length (MDL), coefficient of efficiency (CE) and normalized mean square error (NMSE) indices were used for quantitative performance evaluation of the models. Results on the basis of qualitative and quantitative evaluation indicate that M-6 model with (7-5-5-1) network architecture is better than all models at Champua station and M-1 model with (2-7-7-1) network architecture is better than all models at Anandpur station.