A novel model to predict significant wave height based on long short-term memory network

https://onlinelibrary.wiley.com/doi/pdf/10.1111/nrm.12189

Neural network modeling of hydrological systems: A review of implementation techniques

Multiple imputation of rainfall missing data in the Iberian Mediterranean context

Using Random forest and Gradient boosting trees to improve wave forecast at a specific location

Forecasting, hindcasting and feature selection of ocean waves via recurrent and sequence-to-sequence networks

A Coupled Numerical and Artificial Neural Network Model for Improving Location Specific Wave Forecast

Precipitation is a major factor controlling the hydrology of a region. It has been experienced that the precipitation data are sometimes lost by virtue of instrument malfunctioning, human error(s), climatic conditions which becomes an impediment in the hydrologic and hydraulic designs where rainfall series is a major input. Such missing values could be retrieved using estimated value by employing a large number of conventional hard computing techniques such as normal ratio method, inverse distance value method to name a few. The present work deals with estimation of missing daily rainfall values at 11 rain gauge stations in Pune District of India using soft computing technique of artificial neural networks and by coefficient of correlation weighing method. According to the correlations of averages of all stations, different groups comprising of 3–4 stations each were formed. Models were developed for each group to estimate daily rainfall values at any one station as output while data of other stations in ...

Infilling of missing daily rainfall records using artificial neural network

Artificial Neural Networks (ANNs) have been used increasingly in recent years for numerous hydrological applications because of their ability to model the non-linear relationships. The major drawba...

A novel approach for knowledge extraction from Artificial Neural Networks

Estimation of evaporation is of vital importance in all walks of life. Traditionally, evaporation is estimated using several direct and indirect methods. Most of these methods have disadvantages like extensive data requirement, due to which they are difficult to apply. Chaudhari et al. (Int J Hydrol Sci Technol 2(4):373–390, 2012) [2] successfully developed evaporation models for Nashik climatic centre, located in Maharashtra, India, using the soft computing tool of Artificial Neural Networks. In their work, they showed that ANN models work reasonably well in terms of estimation accuracy for pan evaporation modelling. Among the various parameters used by Chaudhari et al. (Int J Hydrol Sci Technol 2(4):373–390, 2012) [2], two of them were the minimum and maximum daily temperatures as two separate inputs for calibration of the ANN models. This study proposes a novel approach of using average daily temperature as a single input instead of using the minimum and maximum temperature. The models were developed for both average and minimum and maximum temperature, and a detailed analysis was performed for evaluating both the sets. For this purpose, various objective and subjective assessments were carried out, most of which showed that the models trained using average temperature were giving better results. Additionally, the authors were also able to portray the importance of using the correct error measures as per the physical process being modelled and the error measures to be used while judging evaporation models.

Shalaka Shah

Papers

A Coupled Numerical and Artificial Neural Network Model for Improving Location Specific Wave Forecast

Infilling of missing daily rainfall records using artificial neural network

A novel approach for knowledge extraction from Artificial Neural Networks

Evaluation of Pan Evaporation Model Developed Using ANN

Behavioural Characteristics of Multilevel Decomposition Systems of Neuro-Wavelet in Wave Forecasting☆