Caspian Sea level prediction using satellite altimetry by artificial neural networks

The demand for accurate predictions of sea level fluctuations in coastal management and ship navigation activities is increasing. To meet such demand, accessible high-quality data and proper modeling process are critically required. This study focuses on developing and validating a neural methodology applicable to the short-term forecast of the Caspian Sea level. The input and output data sets used contain two time series obtained from Topex/Poseidon and Jason-1 satellite altimetry missions from 1993 to 2008. The forecast is performed by multilayer perceptron network, radial basis function, and generalized regression neural networks. Several tests of different artificial neural network (ANN) architectures and learning algorithms are carried out as alternative methods to the conventional models to assess their applicability for estimating Caspian Sea level anomalies. The results derived from the ANN are compared with observed sea level values and with the forecasts calculated by a routine autoregressive moving average (ARMA) model. Different ANNs satisfactorily provide reliable results for the short-term prediction of Caspian Sea level anomalies. The root mean square errors of the differences between observations and predictions from artificial intelligence approaches can be significantly reduced by about 50 % compared with ARMA techniques.     

Rainfall-Runoff Modeling using Clustering and Regression Analysis for the River Brahmaputra Basin

In this research, k-means, agglomerative hierarchical clustering and regression analysis have been applied in hydrological real time series in the form of patterns and models, which gives the fruitful results of data analysis, pattern discovery and forecasting of hydrological runoff of the catchment. The present study compares with the actual field data, predicted value and validation of statistical yields obtained from cluster analysis, regression analysis with ARIMA model. The seasonal autoregressive integrated moving average (SARIMA) and autoregressive integrated moving average (ARIMA) models is investigated for monthly runoff forecasting. The different parameters have been analyzed for the validation of results with casual effects. The comparison of model results obtained by K-means & AHC have very close similarities. Result of models is compared with casual effects in the same scenario and it is found that the developed model is more suitable for the runoff forecasting. The average value of R² determined is 0.92 for eight ARIMA models. This shows more accuracy of developed ARIMA model under these processes. The developed rainfall runoff models are highly useful for water resources planning and development.     

Short-run forecasting of United States coal production

Summary This paper focuses on short-range modelling and forecasting of aggregate US monthly coal production. The 1976–83 time-series data suggest a multiplicative autoregressive integrated moving average (ARIMA) model to replicate national level monthly coal production. The identified model required 12-month seasonal differencing and has an autoregressive component of lag 1 and a moving average component of lag 12. Model predictions for 1984 were very reasonable when compared with actual production: cyclical patterns were correctly replicated and the deterministic increasing trend was properly identified. The estimated model was enhanced by updating it with data for 1984. Intervention analysis was used to determine the impact of labour negotiations in coal production. Information relative to the identified ARIMA model was then used to model the intervening event of labour negotiations. Intervention modelling produced forecasts for 1984 superior to those identified by the ARIMA model. The mean predicted 1984 US monthly coal production of 1976–84 ARIMA and intervention models were 96.05 and 99.65% of the observed value of 74 178 thousand short tons per month, respectively. Simplicity of the ARIMA and intervention models, the realiability of their predictions, and the ease of updating make them very attractive when compared with large scale econometric models for use in short-term coal production forecasting.     

Combined gamma and M-test-based ANN and ARIMA models for groundwater fluctuation forecasting in semiarid regions

The shortage of surface water in arid and semiarid regions has led to the more use of the groundwater resources. In these areas, the groundwater is essential for activities such as water supply and irrigation. One of the most important stages in sustainable yield of groundwater resources is awareness of groundwater level. In this study, we have applied artificial neural networks (ANN) and autoregressive integrated moving average (ARIMA) models for groundwater level forecasting to 4 months ahead in Shiraz basin, southwestern Iran. Time series analysis was conducted according to the Box–Jenkins method. Meanwhile, gamma and M-test were considered for determining the optimal input combination and length of training and testing data in the ANN model. The results indicated that performance of multilayer perceptron neural network (4, 14, 1) and ARIMA (2, 1, 2) is satisfactory in the groundwater level forecasting for one month ahead. The performance comparison shows that the ARIMA model performs appreciably better than the ANN.     

Structure of the crust and upper mantle of the Black and Caspian Seas

Many geophysical characteristics of the Caspian and Black Seas' deep basins are similar, having: suboceanic type of the crust, low average seismic velocity, absence of earthquakes and relatively small variation of magnetic anomalies. However, the sediments in the Caspian Sea deep basin are folded whereas in the Black Sea they are approximately horizontal. The Caspian Sea also has a far greater thickness of sediment accumulation.

The deep basins of the Caspian, Black and Mediterranean seas represent a sequence having similar crustal structures but with a decreasing thickness of sediments and consolidated layer, in that order. It is possible that the intensive sinking and accumulation of sediments began earliest in the Caspian Sea and spreaded continuously to the Black Sea and then the Mediterranean Sea. The Caspian and Black Sea deep basins have existed for long time (perhaps from Paleozoic time or even earlier) as areas with a specific and related type of evolution.     

Modification of the pattern informatics method for forecasting large earthquake events using complex eigenfactors

Recent studies have shown that real-valued principal component analysis can be applied to earthquake fault systems for forecasting and prediction. In addition, theoretical analysis indicates that earthquake stresses may obey a wave-like equation, having solutions with inverse frequencies for a given fault similar to those that characterize the time intervals between the largest events on the fault. It is therefore desirable to apply complex principal component analysis to develop earthquake forecast algorithms. In this paper we modify the Pattern Informatics method of earthquake forecasting to take advantage of the wave-like properties of seismic stresses and utilize the Hilbert transform to create complex eigenvectors out of measured time series. We show that Pattern Informatics analyses using complex eigenvectors create short-term forecast hot-spot maps that differ from hot-spot maps created using only real-valued data and suggest methods of analyzing the differences and calculating the information gain.     

Changes in the water regime of the caspian sea

The article deals with issues of structure and dynamics of the Caspian Sea water balance. On the base of historical, paleogeomorphological and other data the evolution history of the Caspian Sea and its basin has been observed for different time intervals down to 400 thous. years ago. Presented are computerized data on water balance components in the current centenary obtained from instrumental observations, revealed are causes of the sea-level fluctuations within that time interval and anthropogenic factor contribution to this process. Based on the analysis of this material, an attempt has been undertaken to present a scenarion of a possible sea-level position of the Caspian Sea with the expected versions of climatic changes at the end of the XX and beginning of the XXI centuries.     

The driving forces behind the Caspian Sea mean water level oscillations

Closed basins such as the Caspian Sea rapidly respond to variations in atmospheric and geological events. This study has been conducted to deduce the role of natural and anthropogenic influences on the Caspian Sea mean water level fluctuations for the period of 1998–2005. It is recognized that climate is the primary mechanism of the Caspian Sea mean water level variations based on the relatively small differences of the Caspian Sea hydrologic budget residuals and the Caspian Sea mean water level fluctuations. This is further supported by the similarity in water-level variations of the Caspian Sea with those of Lake Van and Lake Urmia. On the other hand, the Caspian Sea needed to lose some of its water to attain water balance equilibrium in 2000 and 2001. The year 2000 showed anomalous seismic activity particularly in the southwestern part of the Caspian Sea. Two significant earthquakes with magnitudes of 6.8 and 6.5 M_w occurred on November 25, 2000. The focal mechanisms of these earthquakes and numerous aftershocks indicated normal faulting and, therefore, caused the Caspian Sea lake level to decline in 2000. The contribution of submarine mud volcano eruptions to the Caspian Sea lake level could be insignificant based on the comparison of water budget residuals and the mean water level variations. Neither crustal deformation (based on the GPS measurements) nor the offshore oil and natural gas production activities in the Caspian Sea are responsible for noticeable changes to the level of the Caspian Sea.     

基于成分数据分析和能谱-面积分形模型的鄂西黄陵背斜化探异常识别

鄂西黄陵背斜地区虽已发现众多矿点,但是矿床规模较小,找矿潜力不明。由于区内植被茂盛、覆盖严重,致使地质异常信息被削弱或掩盖,对进一步找矿工作造成了严重干扰。本次研究使用成分数据分析和能谱-面积（S-A）分形模型识别与金矿化相关的异常,结果表明：（1）等距对数比（ILR）变换后的主成分分析结果相比对数（Log）变换更能够反应元素的共生组合规律,第一主成分（PC1）元素组合为Au-Cu-Ag,且Au元素占有最大的载荷,与金矿密切相关,PC1元素组合异常的分布模式与矿点的分布更加吻合;（2） S-A方法抑制了黄陵背斜区域水系沉积物的背景模式,增强了异常模式;（3）根据地质特征和以往勘查工作情况,圈定了三处找矿远景区,为该地区金矿下一步勘查工作提供了重要依据。     

Precipitation forecasting using classification and regression trees (CART) model: a comparative study of different approaches

Interest in semiarid climate forecasting has prominently grown due to risks associated with above average levels of precipitation amount. Longer-lead forecasts in semiarid watersheds are difficult to make due to short-term extremes and data scarcity. The current research is a new application of classification and regression trees (CART) model, which is rule-based algorithm, for prediction of the precipitation over a highly complex semiarid climate system using climate signals. We also aimed to compare the accuracy of the CART model with two most commonly applied models including time series modeling (ARIMA), and adaptive neuro-fuzzy inference system (ANFIS) for prediction of the precipitation. Various combinations of large-scale climate signals were considered as inputs. The results indicated that the CART model had a better results (with Nash–Sutcliffe efficiency, NSE?>?0.75) compared to the ANFIS and ARIMA in forecasting precipitation. Also, the results demonstrated that the ANFIS method can predict the precipitation values more accurately than the time series model based on various performance criteria. Further, fall forecasts ranked “very good” for the CART method, while the ANFIS and the time series model approximately indicated “satisfactory” and “unsatisfactory” performances for all stations, respectively. The forecasts from the CART approach can be helpful and critical for decision makers when precipitation forecast heralds a prolonged drought or flash flood.     

Compositional data analysis in the study of integrated geochemical anomalies associated with mineralization

Geochemical data are typical compositional data which should be opened prior to univariate and multivariate data analysis. In this study, a frequency-based method (robust principal component analysis, RPCA) and a frequency-space-based method (spectrum–area fractal model, S–A) are applied to explore the effects of the data closure problem and to study the integrated geochemical anomalies associated with polymetallic Cu mineralization using a stream sediment geochemical dataset collected from the Zhongteng district, Fujian Province (China). The results show that: (1) geochemical data should be opened prior to RPCA to avoid spurious correlation between variables; (2) geochemical pattern is a superimposition of multi-processes and should be decomposed; and (3) the S–A fractal model is a powerful tool for decomposing the mixed geochemical pattern.     

http://www.sciencedirect.com/science/article/pii/S1674987112000849

New methods are presented for processing and interpretation of shallow marine differential magnetic data,including constructing maps of offshore total magnetic anomalies with an extremely high resolution of up to 1-2 nT,mapping weak anomalies of 5-10 nT caused by mineralization effects at the contacts of hydrocarbons with host rocks,estimating depths to upper and lower boundaries of anomalous magnetic sources,and estimating thickness of magnetic layers and boundaries of tectonic blocks. Horizontal dimensions of tectonic blocks in the so-called "seismic gap" region in the central Kuril Arc vary from 10 to 100 km,with typical dimensions of 25-30 km.The area of the "seismic gap" is a zone of intense tectonic activity and recent volcanism.Deep sources causing magnetic anomalies in the area are similar to the "magnetic belt" near Hokkaido. In the southern and central parts of Barents Sea,tectonic blocks with widths of 30-100 km,and upper and lower boundaries of magnetic layers ranging from depths of 10 to 5 km and 18 to 30 km are calculated.Models of the magnetic layer underlying the Mezen Basin in an inland part of the White Sea-Barents Sea paleorift indicate depths to the lower boundary of the layer of 12-30 km.Weak local magnetic anomalies of 2-5 nT in the northern and central Caspian Sea were identified using the new methods,and drilling confirms that the anomalies are related to concentrations of hydrocarbon.Two layers causing magnetic anomalies are identified in the northern Caspian Sea from magnetic anomaly spectra.The upper layer lies immediately beneath the sea bottom and the lower layer occurs at depths between 30-40 m and 150-200 m.     

Application of geoecological monitoring data for multipurpose geochemical mapping: a case history of North Caspian

This paper discusses some techniques and results of investigating of seafloor geochemical field in the Russian part of the North Caspian Sea. Conclusions on the present (2001–2002) level of heavy metal pollution and on the nature of hydrocarbon gase anomalies are presented.     

First results of investigation into the relation between magnetic properties of bottom sediments in the Northern Caspian and variations in the Caspian Sea level in the Late Neo-Pleistocene

The first data of investigation into the relation between changes in magnetic properties of the Northern Caspian sediments and variations in the Caspian Sea level in the Late Neo-Pleistocene are presented. It is shown that there is a certain correlation between magnetic characteristics of sediments and variations in the Caspian Sea level that cause changes in the lithological and faunistic composition of sediments.     

Nonlinear dynamics of the sea level time series in the eastern English Channel

Coastal flooding due to surge events represents natural hazards with huge potential consequences for coastal regions. Sea level time series display variations on a large range of timescales, with a deterministic component associated with tidal variations and a stochastic component primarily associated with meteorological forcing, the non-tidal residual. The deterministic component can be evaluated using a model taking into account astronomical forcing and topographic information. The measured sea level is the sum of a slowly varying mean sea level component, the tidal term and the stochastic term. Here, we consider hourly time series recorded in the ports of Boulogne-sur-Mer, Calais, and Dunkirk, in the eastern English Channel. Measured data and modeled data, both provided by the SHOM (“Service hydrographique et océanographique de la marine,” hydrographic and oceanographic services of the French Navy), are analyzed using Fourier spectral analysis. The statistics of return times of extreme events are also estimated directly from the time series and compared between modeled and measured data. It is found that return times from tidal or measured time series are quite different for large thresholds and that they also have a very different Fourier power spectrum, the measured data having a power-law regime which is not found in the modeled tidal data. It is also shown, using Hilbert–Huang transform, that non-tidal residual time series are intermittent and possess multifractal scaling properties. Finally, water level non-tidal residual relationship is explored, and it is shown that the larger mean values of the surge (negative and positive parts) are obtained for the medium level of the tidal value.     

Comparison of cokriging and adaptive neuro-fuzzy inference system models for suspended sediment load forecasting

The present research was carried out by using artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), cokriging (CK) and ordinary kriging (OK) using the rainfall and streamflow data for suspended sediment load forecasting. For this reason, the time series of daily rainfall (mm), streamflow (m³/s), and suspended sediment load (tons/day) data were used from the Kojor forest watershed near the Caspian Sea between 28 October 2007 and 21 September 2010 (776 days). Root mean square error, efficiency coefficient, mean absolute error, and mean relative error statistics are used for evaluating the accuracy of the ANN, ANFIS, CK, and OK models. In the first part of the study, various combinations of current daily rainfall, streamflow and past daily rainfall, streamflow data are used as inputs to the neural network and neuro-fuzzy computing technique so as to estimate current suspended sediment. Also, the accuracy of the ANN and ANFIS models are compared together in suspended sediment load forecasting. Comparison results reveal that the ANFIS model provided better estimation than the ANN model. In the second part of the study, the ANN and ANFIS models are compared with OK and CK. The comparison results reveal that CK was a better estimation than the OK. The ANFIS and ANN models also provided better estimation than the OK and CK models.     

Study of internal wave generation and propagation features in non-tidal seas based on satellite synthetic aperture radar data

Despite the intense attention paid to internal wave (IW) investigation, the most experimentally studied and theoretically described are internal gravitational waves in shelf zones of oceans and tidal seas appearing during the interaction of tidal currents with the margin of the shelf. Information on surface manifestations of internal waves in enclosed seas, such as the Black and Caspian seas, is almost absent. In this paper, the results of study of the peculiarities of generation and propagation of nontidal internal waves are presented; the study is carried out on the basis of combined analysis in data of marine surface radiolocation and data of optical and infrared satellite-borne sensors. The experimental basis of investigation is radar images of the marine surface derived with the help of Synthetic Aperture Radars, onboard the Envisat and ERS-2 satellites. Additionally, the data of the following sensors belonging to the optical and infrared ranges were used for the purpose of radar image interpretation: MODIS (onboard the Aqua/Terra satellite), MERIS (Envisat), and AVHRR (NOAA). Surface manifestations of IWs in the northeastern part of the Black Sea and in the Caspian Sea have been found in radar imagery for the first time, their pattern of spatial and temporal variation has been reconstructed. The possible factors leading to generation of the observed nontidal IWs are determined and suppositions about the corresponding generation mechanisms are made. In particular, the IW manifestations recorded in the northeastern part of the Black Sea are localized in the vicinity of the boundaries of eddies or hydrological fronts; this fact evidences for the frontal mechanism of generation, at which IWs are radiated by a nonstationary (moving and/or inertially oscillating) front. The most probable main sources of generation of IWs detected in the Caspian Sea are longitudinal one-knot seiches, whose knot point is located near the Apsheron sill.     

Correlation of upper quaternary deposits and paleogeography of the Black and Caspian seas

The evolution of the Black and Caspian seas is considered based on the analysis of new stratigraphic and paleogeographic data. Three transgression stages (Karangatian, Surozh, Black-Sea) and two regression stages (Post-Karangatian and New-Euxinian) were characterized for the Black Sea, as well as four transgression stages (late Khazarian, early Khvalynian, late Khvalynian, and New-Caspian) and three regression stages (Atelian, Enotaevkan, Mangyshlakian), for the Caspian Sea. The analysis of data on the absolute age of deposits allowed correlation of paleogeographic events for the basins, between them and with the stages of the Last (Valdai) Glaciation: the Karangatian and late Khazarian transgressions were correlated with the Mikulinian Interglacial; the post-Karangatian and Atelian regressions, with the Kalininan glaciation; the early Khvalynian and Surozh transgressions, with the middle Valdai Interstadial; the New-Euxinian and Enotaevkan regressions, with the Ostashkovian glaciation; the Black Sea, late Khvalynian, and New-Caspian transgression, with the late glaciation — post-glaciation periods; the Mangyshlakian regression, with the Older Dryas (?). The last connections between the Caspian and Black seas are dated to the middle Valdai time when waters of the early Khvalynian basin drained down to the Surozh basin.     

Intrasecular variability in the level of the largest lakes of Russia

The features of the intrasecular changes in the level of the largest lakes of Russia (Baikal, Ladoga, Onego, and the inland Caspian Sea) are jointly considered for the first time on the basis of the long-term data measured in the 20th–21st centuries. The influence of climate and anthropogenic factors is considered to improve the forecast of the level of lakes.     

Univariate modelling of summer-monsoon rainfall time series: Comparison between ARIMA and ARNN

The present article reports studies to develop a univariate model to forecast the summer monsoon (June–August) rainfall over India. Based on the data pertaining to the period 1871–1999, the trend and stationarity within the time series have been investigated. After revealing the randomness and non-stationarity within the time series, the autoregressive integrated moving average (ARIMA) models have been attempted and the ARIMA(0,1,1) has been identified as a suitable representative model. Consequently, an autoregressive neural network (ARNN) model has been attempted and the neural network has been trained as a multilayer perceptron with the extensive variable selection procedure. Sigmoid non-linearity has been used while training the network. Finally, a three-three-one architecture of the ARNN model has been obtained and after thorough statistical analysis the supremacy of ARNN has been established over ARIMA(0,1,1). The usefulness of ARIMA(0,1,1) has also been described.