A.K. Bhattacharya

Application of SALTMOD in Coastal Clay Soil in India

SALTMOD is a simulation model whichpredicts root zone soil salinity, drainagewater quality and water table depth inagricultural land under differentgeo-hydrological conditions and varyingwater management scenarios. The model wasapplied to the data from coastal AndhraPradesh of India where subsurface drainagesystem is laid out at several drainspacings at the experimental site. Fielddata for 1999, 2000 and 2001 were collectedfrom 35 and 55 m drain spacing plots forSALTMOD application. Modelling was doneconsidering two simulation approaches. Thefirst approach (Simulation-I) used the sameinitial values for the entire simulationperiod. In the second approach(Simulation-II), the computations wereperformed year-by-year, giving each yearthe current input values obtained from thesimulation results of the previous year.Results of these two approaches weredifferent from each other. Simulation-IIgave better predictions than that ofSimulation-I in terms of closeness to theobserved values. Simulation results ofsoil salinity in the root zone, drainagewater quality and quantity and depth towater table revealed that the salinity ofroot zone was predicted more accuratelythan that of drainage water quality anddepth to water table. Also throughsimulation, it was found that the salinityof drainage water was relativelyindependent of the root zone soil salinity. Model application study suggests thatSALTMOD can be used with confidence toevaluate various drain spacings of asubsurface drainage system and facilitatereasonable prediction of the reclamationperiod.

Water table draw down during drainage with evaporation/evapotranspiration

Abstract Groundwater table behavior in drained lands in addition to tile flow depends upon several other factors. Evaporation/evapotranspiration (E/ET) from the shallow water table is one of them. The Boussinesq equation for non-steady state groundwater flow with variable drainable porosity was suitably modified by incorporating a non-linear function simulating E/ET from the shallow water table. The modified equation was solved numerically for predicting the groundwater table behavior with respect to space and time. Column studies and soil tank model experiments were conducted to find out the functional relationship between the evaporation rate and water table height as well as for determination of other parameters required in the mathematical model. The water table fluctuation data were recorded in the soil tank model when both the process of evaporation and drainage were operative. The result of this study revealed that groundwater table behavior, predicted by modified Boussinesq equation was quite comparable with corresponding observed data. The quality of the predictions were much better compared to the conventional approach in which drainable porosity is assumed constant and the process of E/ET is neglected.

Salinity distribution in paddy root zone under subsurface drainage

Abstract One-dimensional convective dispersion solute transport model was solved by Crank–Nicholson finite difference scheme and was applied over the flow domain of subsurface drained paddy fields laid with tile drain at different spacings. The flow domain was divided into a number of stream tubes to predict salinities at different distances from drain centre, at different depths from the ground surface and at different times after the initiation of the operation of the subsurface drainage system. The sub division of the flow domain into a number of stream tubes was done for two purposes, viz. (i) to enable estimation of pore water flow velocity more appropriately with respect to the flow area within a stream tube and (ii) to enable comparison of predicted salinities with the observed values which were available within some of the stream tubes. The initial and the boundary conditions in solving one dimensional equation were based on the field investigated salinity values. In the solute transport model there are two essential input parameters, viz. the pore water velocity and the dispersion coefficient. The pore water velocity was calculated by dividing the Darcian velocity by the drainable porosity. The dispersion coefficient was fitted by trial and error till the average absolute deviation between the predicted and the observed salinities became minimum.

Chance-constrained optimal windmill irrigation system design

In this work, a methodology was developed to enable decision making with regard to the reservoir size and optimal crop plan for a windmill irrigation system. Wind speed and hence the windmill discharge were treated as stochastic variables, and the water pumped by a windmill was computer for different probabilities. Considering the probability of non-availability of a given quantity of pumped water as the risk, reservoir sizes to meet various daily water demands were computed. The available water supplies were then optimally allocated among different crop activities and the profits were calculated. The annual profits were converted to the total profit over the amortization period. The solution which gave the maximum profit was considered to be the best choice. It was found that the most profitable windmill irrigation system can effectively irrigate an area of 2.76 ha and yield an assured annual profit of nearly Rs. 0.089 million (US

Comparison of Artificial Neural Network and regression models for sediment loss prediction from Banha watershed in India

1.00≜12 Indian rupees approximately).