No abstract

Reliable forecast of groundwater level is necessary for its sustainable use and for planning land and water management strategies. This paper deals with an application of artificial neural network (ANN) approach to the weekly forecasting of groundwater levels in multiple wells located over a river basin. Gradient descent with momentum and adaptive learning rate backpropagation (GDX) algorithm was employed to predict groundwater levels 1 week ahead at 18 sites over the study area. Based on the domain knowledge and pertinent statistical analysis, appropriate set of inputs for the ANN model was selected. This consisted of weekly rainfall, pan evaporation, river stage, water level in the surface drain, pumping rates of 18 sites and groundwater levels of 18 sites in the previous week, which led to 40 input nodes and 18 output nodes. During training of the ANN model, the optimum number of hidden neurons was found to be 40 and the model performance was found satisfactory (RMSE= 0.2397 m, r=0.9861, and NSE=0.9722). During testing of the model, the values of statistical indicators RMSE, r and NSE were 0.4118 m, 0.9715 and 0.9288, respectively. Using the same inputs, the developed ANN model was further used for forecasting groundwater levels 2, 3 and 4 weeks ahead in 18 tubewells. The model performance was better while forecasting groundwater levels at shorter lead times (up to 2 weeks) than that for larger lead times.

No abstract

The present study deals with the estimation of soil loss from the Upper Subarnarekha catchment in Odisha (India) using Soil and Water Assessment Tool (SWAT). Sequential uncertainty fitting (SUFI-2) algorithm of the SWAT calibration uncertainty programs (SWAT-CUP) was used for model simulation. The model was calibrated with the observed data for the period from 1996 to 2008 with first 3 years (1996)(1997)(1998) as warm-up period. Further, validation of the model was done using 5-year data from 2009 to 2013. Reliable evaluation of the model performance during calibration has been substantiated by the coefficient of determination (R 2 ), Nash-Sutcliffe efficiency (NSE), and percentage bias (PBIAS) as 0.81, 0.81, and −0.15, respectively, and the respective values for the validated model were found to be 0.79, 0.78, and −0.19. The values of P-factor and R-factor were found to be 0.80 and 0.75 and 0.66 and 0.74, respectively, for model calibration and validation. Average annual soil loss from the catchment was 4.84 Mg ha −1 . The watershed indexed as SW18 resulted in highest soil loss in the range of 10-15 Mg ha −1 year −1 . Further, prioritization was done at the level of sub-watersheds using the data of simulated sediment yield, soil texture, land use, and slope for identifying vulnerable sub-watersheds that need immediate attention. The study inferred that sub-watersheds having index numbers SW17, SW18, and SW19 are highly vulnerable, and hence top priority should be given to these sub-watersheds for reduction in soil erosion through the implementation of suitable soil and water conservation measures.

No abstract