Oleg Makarynskyy

Sea water level forecasting using genetic programming and comparing the performance with Artificial Neural Networks

Water level forecasting at various time intervals using records of past time series is of importance in water resources engineering and management. In the last 20 years, emerging approaches over the conventional harmonic analysis techniques are based on using Genetic Programming (GP) and Artificial Neural Networks (ANNs). In the present study, the GP is used to forecast sea level variations, three time steps ahead, for a set of time intervals comprising 12h, 24h, 5 day and 10 day time intervals using observed sea levels. The measurements from a single tide gauge at Hillarys Boat Harbor, Western Australia, were used to train and validate the employed GP for the period from December 1991 to December 2002. Statistical parameters, namely, the root mean square error, correlation coefficient and scatter index, are used to measure their performances. These were compared with a corresponding set of published results using an Artificial Neural Network model. The results show that both these artificial intelligence methodologies perform satisfactorily and may be considered as alternatives to the harmonic analysis.

Neuro-fuzzy and neural network techniques for forecasting sea level in Darwin Harbor, Australia

Accurate predictions of sea level with different forecast horizons are important for coastal and ocean engineering applications, as well as in land drainage and reclamation studies. The methodology of tidal harmonic analysis, which is generally used for obtaining a mathematical description of the tides, is data demanding requiring processing of tidal observation collected over several years. In the present study, hourly sea levels for Darwin Harbor, Australia were predicted using two different, data driven techniques, adaptive neuro-fuzzy inference system (ANFIS) and artificial neural network (ANN). Multi linear regression (MLR) technique was used for selecting the optimal input combinations (lag times) of hourly sea level. The input combination comprises current sea level as well as five previous level values found to be optimal. For the ANFIS models, five different membership functions namely triangular, trapezoidal, generalized bell, Gaussian and two Gaussian membership function were tested and employed for predicting sea level for the next 1h, 24h, 48h and 72h. The used ANN models were trained using three different algorithms, namely, Levenberg-Marquardt, conjugate gradient and gradient descent. Predictions of optimal ANFIS and ANN models were compared with those of the optimal auto-regressive moving average (ARMA) models. The coefficient of determination, root mean square error and variance account statistics were used as comparison criteria. The obtained results indicated that triangular membership function was optimal for predictions with the ANFIS models while adaptive learning rate and Levenberg-Marquardt were most suitable for training the ANN models. Consequently, ANFIS and ANN models gave similar forecasts and performed better than the developed for the same purpose ARMA models for all the prediction intervals.

Prediction of Short-Term Operational Water Levels Using an Adaptive Neuro-Fuzzy Inference System

Sea level estimates are important in many coastal applications and port activities. This paper investigates the ability of a neuro-fuzzy (NF) model to predict sea level variations at a tide gauge site in the Hillarys Boat Harbour, Western Australia. In the first part of the study, previously recorded sea levels were used as input to estimate current sea levels. The results showed an acceptable level of NF model accuracy. In the second part of the study, NF models were implemented to forecast sea levels averaged over 12- and 24-h time periods, three time steps ahead. The NF forecasts were compared with those of artificial neural networks (ANNs) for the same data set. The results show that the NF approach performed better than the ANN in half-daily 12-, 24-, and 36-h sea level predictions. The traditional linear regression and autoregressive models were also tested for comparison, and they demonstrated their inferiority to the results of other techniques.

Use of Soil Moisture Variability in Artificial Neural Network Retrieval of Soil Moisture

Passive microwave remote sensing is one of the most promising techniques for soil moisture retrieval. However, the inversion of soil moisture from brightness temperature observations is not straightforward, as it is influenced by numerous factors such as surface roughness, vegetation cover, and soil texture. Moreover, the relationship between brightness temperature, soil moisture and the factors mentioned above is highly non-linear and ill-posed. Consequently, Artificial Neural Networks (ANNs) have been used to retrieve soil moisture from microwave data, but with limited success when dealing with data different to that from the training period. In this study, an ANN is tested for its ability to predict soil moisture at 1 km resolution on different dates following training at the same site for a specific date. A novel approach that utilizes information on the variability of soil moisture, in terms of its mean and standard deviation for a (sub) region of spatial dimension up to 40 km, is used to improve the current retrieval accuracy of the ANN method. A comparison between the ANN with and without the use of the variability information showed that this enhancement enables the ANN to achieve an average Root Mean Square Error (RMSE) of around 5.1% v/v when using the variability information, as compared to around 7.5% v/v without it. The accuracy of the soil moisture retrieval was

Wind Speed Prediction by Using Different Wavelet Conjunction Models

Three wavelet conjunction models, wavelet-genetic programming (WGEP), wavelet-neuro-fuzzy (WNF) and wavelet-neural network (WNN) were introduced in this paper for predicting hourly and daily wind speed values with three lag times. Hourly wind speed measurements from Darwin Airport synoptic station and daily wind speed data from Tabriz Station (North-western Iran) were used as inputs to the wavelet conjunction models to predict 1-, 2- and 3-hour and 1-, 2- and 3-days ahead wind speeds. First, conventional GEP, NF and ANN models were applied to the wind speed time series. Then WGEP, WNF and WANN conjunction models were also used for the same purpose and their results were compared with those of the conventional GEP, NF and ANNs. The correlation coefficient, root mean squared error, scatter index and mean absolute error were used to evaluate the performance of the models. Inter-comparisons of model results indicated that the use of wavelet conjunction models increased the performance of the conventional GEP, ANFIS and ANN in forecasting hourly and daily wind speeds.

Combining deterministic modelling with artificial neural networks for suspended sediment estimates

The presented manuscript presents a novel approach to the applied ocean and coastal engineering problem of sediment concentration estimates.The approach has been developed on the basis of numerical modelling and with application of artificial neural networks.It has been demonstrated that the proposed methodology can be generalised onto near-by locations.Further generalisation must be tested before applying. Estimates of suspended sediment concentrations and transport are an important part of any marine environment assessment study because these factors have a direct impact on the life cycle and survival of marine ecosystems. This paper proposes to implement a combined methodology to tackle these estimates. The first component of the methodology comprised two numerical current and wave models, while the second component was based on the artificial intelligence technique of neural networks (ANNs) used to reproduce values of sediment concentrations observed at two sites. The ANNs were fed with modelled currents and waves and trained to produce area-specific concentration estimates. The trained ANNs were then applied to predict sediment concentrations over an independent period of observations. The use of a data set that merged together observations from both the mentioned sites provided the best ANN testing results in terms of both the normalised root mean square error (0.13) and the mean relative error (0.02).

Forecasting daily stream flows using artificial intelligence approaches

This study compares three different artificial intelligence approaches, namely, gene expression programming (GEP), adaptive neuro-fuzzy inference system (ANFIS) and artificial neural networks (ANNs), in daily stream flow forecasting of Alavian Dam Station on the Soofi-Cahi River in the Northwestern Iran. The study demonstrates that the optimal results were obtained from the triple-input models, including stream flows of current and 2 previous days, and that the GEP model performed better than the ANN and ANFIS models in daily stream flow forecasting. It was found that the optimal GEP model with coefficient of determination R 2 = 0.924, root mean square error = 0.908 m3/s and scatter index = 0.354 in the test stage was superior in forecasting daily stream flow to the optimal ANN and ANFIS models.

Forecasting Sea Water Levels at Mukho Station, South Korea Using Soft Computing Techniques

The accuracy of three different data-driven methods, namely, Gene Expression Programming (GEP), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), is investigated for hourly sea water level prediction at the Mukho Station in the East Sea (Sea of Japan). Current and four previous level measurements are used as input variables to predict sea water levels up to 1, 24, 48, 72, 96 and 120 hours ahead. Three statistical evaluation parameters, namely, the correlation coefficient, the root mean square error and the scatter index are used to assess how the models perform. Investigation results indicate that, when compared to measurements, for +1h prediction interval, all three models perform well (with average values of R = 0.993, RMSE = 1.3 cm and SI = 0.04), with slightly better results produced by the ANNs and ANFIS, while increasing the prediction interval degrades model performance.

Forecasting Water Level Fluctuations of Urmieh Lake using Gene Expression Programming and Adaptive Neuro- Fuzzy Inference System

Forecasting lake level at various prediction intervals is an essential issue in such industrial applications as navigation, water resource planning and catchment management. In the present study, two data driven techniques, namely Gene Expression Programming and Adaptive Neuro-Fuzzy Inference System, were applied for predicting daily lake levels for three prediction intervals. Daily water-level data from Urmieh Lake in Northwestern Iran were used to train, test and validate the used techniques. Three statistical indexes, coefficient of determination, root mean square error and variance accounted for were used to assess the performance of the used techniques. Technique inter-comparisons demonstrated that the GEP surpassed the ANFIS model at each of the prediction intervals. A traditional auto regressive moving average model was also applied to the same data sets; the obtained results were compared with those of the data driven approaches demonstrating superiority of the data driven models to ARMA.

Genetic Programming for Sea Level Predictions in an Island Environment

Accurate predictions of sea-level are important for geodetic applications, navigation, coastal, industrial and tourist activities. In the current work, the Genetic Programming (GP) and artificial neural networks (ANNs) were applied to forecast half-daily and daily sea-level variations from 12 hours to 5 days ahead. The measurements at the Cocos (Keeling) Islands in the Indian Ocean were used for training and testing of the employed artificial intelligence techniques. A comparison was performed of the predictions from the GP model and the ANN simulations. Based on the comparison outcomes, it was found that the Genetic Programming approach can be successfully employed in forecasting of sea level variations.