International Journal of Agricultural Engineering

S. N. Bansude ¹, G. L. Chunale ², A. A. Shinde ³, Pravendra Kumar ⁴

Affiliations
1 Department of Soil and Water Conservation Engineering, Dr. Panjabrao Deshmukh Krushi Vidyapeeth, Akola (M.S.), IN
2 Zonal Agricultural Research Station, Shenda Park (M.P.K.V.), Karveer (M.S.), IN
3 Jain Irrigation Systems Ltd., Ramthal Lift Irrigation Scheme, Bagalkot (Karnataka), IN
4 Department of Soil and Water Conservation Engineering, G.B. Pant University of Agriculture and Technology, Pantnagar(Uttarakhand), IN

Abstract

In the present study, two mathematical models namely (i) Lag and route model and (ii) Muskingum model which are based on unit-step and transfer functions approach were developed for runoff prediction from Shenda park watershed treating the watershed as lumped, linear and time-invariant system. The hydrological data of the study watershed were collected from zonal station of National Agricultural Research Project, Shenda Park, Kolhapur (M.S.) for the years 2000 to 2008. Out of twelve single storm events, nine storm events were included in the analysis for parameters estimation and remaining three storm events were considered for prediction purposes. The model parameters, viz., lag time and (τ) and storage co-efficient (K) of Lag and route model were estimated by the methods of cumulants (Singh, 1988) and moments (Nash, 1957) whereas the model parameter storage constant (K) for Muskingum model was estimated by using method suggested by Jawed (1973). Performance evaluation of these two developed model in determining direct runoff hydrograph ordinates were evaluated using various statistical indices such as correlation co-efficient (R), special correlation co-efficient (Rs.), co-efficient of efficiency (CE) and ischolar_main mean square error (RMSE). The results showed that both the developed model can be used for prediction of the direct run off hydrograph from the study watershed, however, direct runoff hydrographs obtained through Muskingun models are much closer to actual observed direct runoff hydrograph than that of Lag and route model.

Keywords

Lag and Route Model, Muskingum Model, Unit-Step Transfer Functions.

Full Text

     

Shreya Nivesh ¹, Pravendra Kumar ¹

Affiliations
1 Department of Soil and Water Conservation Engineering, College of Technology, Govind Ballabh Pant University of Agriculture and Technology, Pantnagar, U.S. Nagar (Uttarakhand), IN

Abstract

Soil erosion by water is the most serious form of land degradation resulting in loss of crop productivity by 0.2-10.9 q/ha (66% total production loss) for cereals, 0.1-6.3 q/ha for oilseeds (21% total production loss) and 0.04-4.4 q/ha for pulses (13% total production loss) estimated across states, which has a direct bearing on food security of the country. Therefore, a major challenge still remaining is the accurate prediction of the catchment sediment yield responses to the rainfall-runoff events. One viable approach to this challenge is the use of suitable statistical and soft-computing techniques for the efficient management of watersheds and ecosystems. The present study deals with the development of adaptive neuro-fuzzy inference system (ANFIS) and multiple linear regression (MLR) models to estimate the suspended sediment load from Vamsadhara river catchment comprising of 7820 km², situated between Mahanadi and Godavari river basins in south India. Considering the active monsoon period, 70% data were used for model calibration and remaining 30% data were used for model validation. Results revealed that the Neuro-Fuzzy models are in good agreement with the observed values and present better performance in comparison to the statistical models.

Keywords

Adaptive Neuro-Fuzzy Inference System, Multiple Linear Regression, Calibration, Validation, Soft-Computing.

Full Text