- Title
- Evaluation of VIC, ANN and Empirical Models for Estimating Daily Reference Evapotranspiration
- Creator
- Adamala, S.; Srivastava, A.; Bachina, H. B.; Palakuru, M.
- Relation
- Indian Journal of Ecology Vol. 47, p. 29-36
- Relation
- https://indianecologicalsociety.com/full-journals/
- Publisher
- Indian Ecological Society
- Resource Type
- journal article
- Date
- 2020
- Description
- The reference gross evapotranspiration (ETo) is estimated using different models for Mohanpur climatic location in India. The various models considered are water budget based variable infiltration capacity (VIC) model and empirical based Turc, FAO-24 Pan, Hargreaves, linear artificial neural network (LNN), quadratic artificial neural network (QNN), and cubic artificial neural network (CNN). The performance of different models was evaluated using five indices such as root mean squared error (RMSE), mean absolute error (MAE), coefficient of determination (R2), ratio of average output to the average target ETo values (Rratio), and index of agreement (d). The QNN models gave better performance in terms of low RMSE and MAE, and high R2 and d values as compared to other models. Rratiovalue very near to one in case of QNN indicates that neither over-estimation nor under-estimation of ETo values. Though, the performance of physically based VIC model is not comparable with the QNN model, the VIC models results also quite encouraging as it's a water balance based approach.
- Identifier
- http://hdl.handle.net/1959.13/1520395
- Identifier
- uon:57462
- Identifier
- ISSN:0304-5250
- Language
- eng
- Reviewed
- Hits: 15
- Visitors: 15
- Downloads: 0