Monthly reservoir inflow forecasting using a new hybrid SARIMA genetic programming approach

Forecasting reservoir inflow is one of the most important components of water resources and hydroelectric systems operation management. Seasonal autoregressive integrated moving average (SARIMA) models have been frequently used for predicting river flow. SARIMA models are linear and do not consider the random component of statistical data. To overcome this shortcoming, monthly inflow is predicted in this study based on a combination of seasonal autoregressive integrated moving average (SARIMA) and gene expression programming (GEP) models, which is a new hybrid method (SARIMA–GEP). To this end, a four-step process is employed. First, the monthly inflow datasets are pre-processed. Second, the datasets are modelled linearly with SARIMA and in the third stage, the non-linearity of residual series caused by linear modelling is evaluated. After confirming the non-linearity, the residuals are modelled in the fourth step using a gene expression programming (GEP) method. The proposed hybrid model is employed to predict the monthly inflow to the Jamishan Dam in west Iran. Thirty years’ worth of site measurements of monthly reservoir dam inflow with extreme seasonal variations are used. The results of this hybrid model (SARIMA–GEP) are compared with SARIMA, GEP, artificial neural network (ANN) and SARIMA–ANN models. The results indicate that the SARIMA–GEP model (R ²=78.8, VAF =78.8, RMSE =0.89, MAPE =43.4, CRM =0.053) outperforms SARIMA and GEP and SARIMA–ANN (R ²=68.3, VAF =66.4, RMSE =1.12, MAPE =56.6, CRM =0.032) displays better performance than the SARIMA and ANN models. A comparison of the two hybrid models indicates the superiority of SARIMA–GEP over the SARIMA–ANN model.     

不同时间尺度的中长期水文预报研究

为研究中长期水文预报时间尺度对预报精度的影响,选取最近邻抽样回归模型与基于小波分析的组合模型对长江干流典型断面不同时间尺度的径流序列进行中长期径流预报。将1980~2012年的逐日径流资料经过时间聚集方法转换成三天、周、旬、半月、月、双月、季、半年、九月、年等10个不同时间尺度,对高场、寸滩、宜昌、螺山、汉口、大通6个典型断面的径流进行拟合和预报。结果表明:随着预报时间尺度增加,预报精度呈现先降低后提高的趋势,其中,在月时间尺度上预报效果最差,三天和年尺度上预报效果相对较好。     

Suspended sediment load prediction of river systems: GEP approach

This study presents gene expression programming (GEP), an extension of genetic programming, as an alternative approach to modeling the suspended sediment load relationship for the three Malaysian rivers. In this study, adaptive neuro-fuzzy inference system (ANFIS), regression model, and GEP approaches were developed to predict suspended load in three Malaysian rivers: Muda River, Langat River, and Kurau River [ANFIS (R ²?=?0.93, root mean square error (RMSE)?=?3.19, and average error (AE)?=?1.12) and regression model (R ²?=?0.63, RMSE?=?13.96, and AE?=?12.69)]. Additionally, the explicit formulations of the developed GEP models are presented (R ²?=?0.88, RMSE?=?5.19, and AE?=?6.5). The performance of the GEP model was found to be acceptable compare to ANFIS and better than the conventional models.     

基于Ricker子波核的支持向量回归方法及其在地震勘探记录去噪处理中的应用

针对地震勘探中强随机噪声的去噪问题,引进支持向量回归方法,提出并证明一种新的Ricker子波核函数。支持向量回归采用核映射的基本思想,基于结构风险最小化原则,将回归问题转化为一个二次规划问题。对单道记录或多道记录中任选道的仿真实验表明,与传统的基于径向基核函数的支持向量回归及褶积滤波方法相比,使用本方法去噪后的同相轴更为清晰,波形恢复得更好,信噪比也较高,因此有可能将其应用于地震勘探记录的去噪处理中。     

Evaluating the performance of artificial intelligence methods for estimation of monthly mean soil temperature without using meteorological data

Soil temperature has an important role in agricultural, hydrological, meteorological and climatological studies. In the present research, monthly mean soil temperature at four different depths (5, 10, 50 and 100 cm) was estimated using artificial neural networks (ANN), adaptive neuro-fuzzy inference system (ANFIS) and gene expression programming (GEP). The monthly mean soil temperature data of 31 stations over Iran were employed. In this process, the data of 21 and 10 stations were used for training and testing stages of used models, respectively. Furthermore, the geographical information including latitude, longitude and altitude as well as periodicity component (the number of months) was considered as inputs in the mentioned intelligent models. The results demonstrated that the ANN and ANFIS models had good performance in comparison with the GEP model. Nevertheless, the ANFIS generally performed better than ANN model.     

An application of artificial intelligence for rainfall-runoff modeling

This study proposes an application of two techniques of artificial intelligence (AI) for rainfall-runoff modeling: the artificial neural networks (ANN) and the evolutionary computation (EC). Two different ANN techniques, the feed forward back propagation (FFBP) and generalized regression neural network (GRNN) methods are compared with one EC method, Gene Expression Programming (GEP) which is a new evolutionary algorithm that evolves computer programs. The daily hydrometeorological data of three rainfall stations and one streamflow station for Juniata River Basin in Pennsylvania state of USA are taken into consideration in the model development. Statistical parameters such as average, standard deviation, coefficient of variation, skewness, minimum and maximum values, as well as criteria such as mean square error (MSE) and determination coefficient (R ²) are used to measure the performance of the models. The results indicate that the proposed genetic programming (GP) formulation performs quite well compared to results obtained by ANNs and is quite practical for use. It is concluded from the results that GEP can be proposed as an alternative to ANN models.     

Prediction of Resilient Modulus from Post-compaction Moisture Content and Physical Properties Using Support Vector Regression

The present study assesses the use of support vector machine regression to predict the variation of resilient modulus with post-compaction moisture content of soils commonly encountered in Oklahoma, Pennsylvania and Wisconsin. Results show the prediction model using the support vector regression (SVR) approach is a function of degree of saturation, moisture content and plasticity index. The developed model is compared to current models in the literature. Results indicate the proposed SVR model gives more accurate values than current regression models. This model will better predict changes in the bearing capacity of pavements due to seasonal variations of moisture content.     

Using of gene expression programming and climatic data for forecasting flow discharge by considering trend,normality, and stationarity analysis

In this research, the main hydrological characteristics (such as trend, stationarity, and normalization of hydrological data) of the Kasilian watershed are considered from 1970 to 2009. For forecasting of discharge, gene expression programming (GEP) method is applied. Normality and stationarity of time series are necessary for application of GEP method. For this purpose, third edition of Mann-Kendall trend test and skewness test are used for detection of trend and normalization of data, respectively. Also, five methods are applied for detection of stationarity of data. Modified Mann-Kendall trend test and Theil and Sen’s median slope method illustrate that annual and monthly precipitation data have slight decreasing trend, annual and monthly discharge data have insignificant decreasing trend, and annual and monthly temperature data have an increasing trend. Skewness test illustrates that annual, monthly, and daily discharge and precipitation data are not normal. By using logarithm function, skewness is minimized and symmetry of data is improved. After normalization of time series by logarithm function, five methods are applied for testing of stationarity of time series. These methods show that different normalized time series are stationarity and stationarity of time series is improved by elimination of periodic properties of data. For forecasting of daily discharge by GEP method, 85% of data are used for training and 15% of data are used for testing. By using data of 3 days ago, the GEP has the best efficiency. Coefficient of correlation (CC), root mean square error (RMSE), mean absolute error (MAE), and mean absolute relative error (MARE) are 0.9, 0.495 lit/s, 0.288 lit/s, and 0.053, respectively.     

Roadheader performance prediction using genetic programming (GP) and gene expression programming (GEP) techniques

Roadheading machines play a vital role in excavation operation in tunneling and mining industries notably when selective mining is required. Roadheaders are more effective in soft to medium rock formations due to a higher cutting rate in such strata. A precise prediction of machine’s performance is a crucial issue, as it has considerable effects on excavation planning, project’s cost estimation, machine specification selection as well as safety of the project. In this research, a database of machine performance and some geomechanical parameters of rock formations from Tabas coal mine project, the largest and fully mechanized coal mine in Iran, has been established, including instantaneous cutting rate (ICR), uniaxial compressive strength, Brazilian tensile strength, rock quality designation, influence of discontinuity orientation (Alpha angle) and specific energy. Afterward, the parameters were analyzed through genetic programming (GP) and gene expression programming (GEP) approaches to yield more accurate models to predict the performance of roadheaders. As statistical indices, coefficient of determination, root mean square error and variance account were used to evaluate the efficiency of the developed models. According to the obtained results, it was observed that developed models can effectively be implemented for prediction of roadheader performance. Moreover, it was concluded that performance of the GEP model is better than the GP model. A high conformity was observed between predicted and measured roadheader ICR for GEP model.     

Linear genetic programming for time-series modelling of daily flow rate

In this study linear genetic programming (LGP), which is a variant of Genetic Programming, and two versions of Neural Networks (NNs) are used in predicting time-series of daily flow rates at a station on Schuylkill River at Berne, PA, USA. Daily flow rate at present is being predicted based on different time-series scenarios. For this purpose, various LGP and NN models are calibrated with training sets and validated by testing sets. Additionally, the robustness of the proposed LGP and NN models are evaluated by application data, which are used neither in training nor at testing stage. The results showed that both techniques predicted the flow rate data in quite good agreement with the observed ones, and the predictions of LGP and NN are challenging. The performance of LGP, which was moderately better than NN, is very promising and hence supports the use of LGP in predicting of river flow data.     

Modeling of Oblique Load Test on Batter Pile Group Based on Support Vector Machines and Gaussian Regression

This paper evaluates the potential of two machine learning approaches i.e. Support vector machine (SVR) and Gaussian processes (GP) regression to model the oblique load capacity of batter pile groups. Linear regression was used to compare the performance of both SVR and GP based regression approaches to model the oblique load. Data set used consists of 147 samples obtained from the laboratory experiments. Out of the total sample size, 105 randomly selected samples were used for training whereas remaining 42 were used for testing the models. Input data set consist of angle of oblique load, pile length, sand relative density, number of vertical piles, number of batter piles where as oblique load was considered as output. Two kernel functions i.e. Polynomial and radial based kernel function were used with both SVR and GP regression. A comparison of results suggest that radial basis function based SVR approach works well in comparison to GP and linear regression based approaches and it could successfully be employed in modelling the oblique load capacity of batter pile groups. Parametric analysis and sensitivity analysis suggest that loading angle, pile length and number of batter pile were important in prediction of oblique load.     

GPM与TRMM降水数据在中国大陆的精度评估与对比

为评估TRMM 3B42（TRMM）和新一代GPM IMERG（GPM）卫星降水产品精度,基于国内824个气象站点日降水数据,选用相关系数（R）、相对误差（E_R）和公正先兆评分（S_ET）等指标,对比分析了二者在中国大陆和九大流域内逐日、逐月尺度的观测精度。研究表明:①在日尺度上,中国大陆内的GPM降水数据精度整体优于TRMM,二者的R、E_R和S_ET分别达到了0.73、2.03%、0.36和0.70、3.75%、0.33;②GPM和TRMM日降水数据在海河流域、淮河流域、长江流域、珠江流域、东南诸河流域呈现较高的观测精度,在松辽流域、黄河流域、西南诸河片区精度次之,在内陆河片区相对最低;③在月尺度上,中国大陆内的GPM冬季降水精度明显好于TRMM,这是由于GPM提高了对弱降水和固态降水的观测能力。总体上,GPM降水产品在中国各大流域精度较好且优于TRMM,表明其在流域降水研究及水文模拟中将有较好的应用前景。     

Climate change impacts on extreme floods I: combining imperfect deterministic simulations and non-stationary frequency analysis

Flood quantiles are routinely used in hydrologic engineering to design hydraulic structures, optimize erosion control structure and map the extent of floodplains. As an increasing number of papers are pointing out cycles and trends in hydrologic time series, the use of stationary flood distributions leads to the overestimation or underestimation of the hydrologic risk at a given time. Several authors tried to address this problem by using probability distributions with time-varying parameters. The parameters of these distributions were assumed to follow a linear or quadratic trend in time, which may be valid for the short term but may lead to unrealistic long-term projections. On the other hand, deterministic rainfall-runoff models are able to successfully reproduce trends and cycles in stream flow data but can perform poorly in reproducing daily flows and flood peaks. Rainfall-runoff models typically have a better performance when simulation results are aggregated at a larger time scale (e.g. at a monthly time scale vs. at a daily time scale). The strengths of these two approaches are combined in this paper where the annual maximum of the time-averaged outputs of a hydrologic model are used to modulate the parameters of a non-stationary GEV model of the daily maximum flow. The method was applied to the Kemptville Creek located in Ontario, Canada, using the SWAT (Soil and Water Assessment Tool) model as rainfall-runoff model. The parameters of the non-stationary GEV model are then estimated using Monte Carlo Markov Chain, and the optimal span of the time windows over which the SWAT outputs were averaged was selected using Bayes factors. Results show that using the non-stationary GEV distribution with a location parameter linked to the maximum 9-day average flow provides a much better estimation of flood quantiles than applying a stationary frequency analysis to the simulated peak flows.     

黑河出山径流的非线性特征分析

应用非线性动力学的理论和方法,对黑河出山径流的非线性特征进行了分析.结果表明,黑河月出山径流的年内分布、年平均流量的一次峰、谷变化符合单重或双重威布尔分布,并具有自相似性质.黑河出山径流多年变化在相空间中的运动轨迹收缩到一个约为4.32维的吸引子上,而描述流量的动力方程需要8个独立变量.黑河出山径流的非线性特征还表现在对内部结构为非线性函数的输入输出模型的良好应用上,如GRNN神经网络模型、非线性回归模型等.     

Effects of changes in vegetation on precipitation in the northern Tianshan Mountains evaluated using multiple time scales

This study used a combination of the wavelet cross-correlation technique and numerical analysis of vegetative feedback to study the role of climate–vegetation feedback from 1981 to 2009 in the northern Tianshan Mountains, Xinjiang Province, China. The study area included the Irtysh River, the Bortala and Ili River valleys, the northern slopes of the Tianshan Mountains, and the western Junggar Basin. The feedback effects of changes in vegetation on precipitation appeared to vary in these five regions when different time scales are used to examine them. The most useful time scale was generally found to be 4–6 months. Time lag was another characteristic of this process, and the optimal time lag was 3–4 months. Nevertheless, optimal time scale and time lag did not differ significantly in these five regions. In this way, the correct time scale of the effects of variations in vegetation on precipitation in this cold, arid area was found. This time scale and time lag can be assessed through wavelet cross-correlation analysis. Then numerical analysis can be used to improve the accuracy of the analysis.     

Correlation of Pile Axial Capacity and CPT Data Using Gene Expression Programming

Numerous methods have been proposed to assess the axial capacity of pile foundations. Most of the methods have limitations and therefore cannot provide consistent and accurate evaluation of pile capacity. However, in many situations, the methods that correlate cone penetration test (CPT) data and pile capacity have shown to provide better results, because the CPT results provide more reliable soil properties. In an attempt to obtain more accurate correlation of CPT data with axial pile capacity, gene expression programming (GEP) technique is used in this study. The GEP is a relatively new artificial intelligent computational technique that has been recently used with success in the field of engineering. Three GEP models have been developed, one for bored piles and two other models for driven piles (a model for each of concrete and steel piles). The data used for developing the GEP models are collected from the literature and comprise a total of 50 bored pile load tests and 58 driven pile load tests (28 concrete pile load tests and 30 steel pile load tests) as well as CPT data. For each GEP model, the data are divided into a training set for model calibration and an independent validation set for model verification. The performances of the GEP models are evaluated by comparing their results with experimental data and the robustness of each model is investigated via sensitivity analyses. The performances of the GEP models are evaluated further by comparing their results with the results of number of currently used CPT-based methods. Statistical analyses are used for the comparison. The results indicate that the GEP models are robust and perform well.     

不同蒸散发计算模型在新疆喀什地区的适用性及对比分析研究

为分析不同蒸散发模型在新疆喀什地区的适用性,结合国内常用两种蒸散发计算方法,应用区域实测蒸发数据,按照年、月、季三个时间尺度分析两种方法计算的适用性和精度。结果表明:VIC模型在年尺度上适用性好于P-M模型,但在夏季和秋季尺度上,P-M模型好于VIC模型,各模型在夏季蒸发计算上好于其他季节。研究成果对于喀什地区蒸发计算提供重要的方法参考。     

叶尔羌河源流区近50年来年径流变化及其对气候变化的响应

  基于新疆卡群水文站和塔什库尔干气象站1959～2005年的观测资料,运用小波分析和统计分析相结合的方法,从多时间尺度研究叶尔羌河源流区近50年来年径流的非线性变化趋势,以及径流对气候变化的响应。结果表明: 1）年径流、气温和降水的主要变化周期几乎一致,年径流和年降水量都存在24年的主要变化周期,而年平均气温则是23年。2）年径流量表现出具有时间尺度依赖性的非线性变化趋势,与区域气候变化密切相关。3）年径流变化是区域气候变化响应的结果,从8（2³）年、4（2²）年和2（2¹）年时间尺度上来看,年径流量与年平均气温和年降水量之间存在显著的线性相关关系。     

Multi-temporal scale modeling on climatic-hydrological processes in data-scarce mountain basins of Northwest China

Previous studies showed that the climatic processes drive the streamflow of the inland river in Northwest China. However, it is difficult to quantitatively assess the climatic-hydrological processes in the ungauged mountainous basins because of the scarce data. This research developed an integrated approach for multi-temporal scale modeling the climatic-hydrological processes in data-scarce mountain basins of Northwest China by combining downscaling method (DM), backpropagation artificial neural network (BPANN), and wavelet regression (WR). To validate the approach, we also simulated the climatic-hydrological processes at different temporal scales in a typical data-scarce mountain basin, the Kaidu River Basin in Northwest China. The main results are as follows: (i) the streamflow is related with regional climatic change as well as atmosphere-ocean variability, (ii) the BPANN model well simulated the nonlinear relationship between the streamflow and temperature and precipitation at the monthly temporal scale, and (iii) although the annual runoff (AR) appears to have fluctuations, there are significant correlations among AR, annual average temperature (AAT), annual precipitation (AP), and oscillation indices, which can be simulated by equations of WR at different temporal scales of years.     

Spatial and temporal variability and drivers of net ecosystem metabolism in western Gulf of Mexico estuaries

Net ecosystem metabolism (NEM) is becoming a commonly used ecological indicator of estuarine ecosystem metabolic rates. Estuarine ecosystem processes are spatially and temporally variable, but the corresponding variability in NEM has not been properly assessed. Spatial and temporal variability in NEM was assessed in four western Gulf of Mexico shallow water estuaries. NEM was calculated from high-frequency dissolved oxygen measurements. Interbay, intrabay, and water column spatial scales were assessed for NEM, gross primary production (GPP), and respiration (R) rate variability. Seasonal, monthly, and daily temporal scales in NEM, GPP, and R were also assessed. Environmental conditions were then compared to NEM to determine which factors were correlated with each temporal and spatial scale. There was significant NEM spatial variability on interbay, intrabay, and water column spatial scales. Significant spatial variability was ephemeral, so it was difficult to ascertain which environmental conditions were most influential at each spatial scale. Significant temporal variability in NEM on seasonal, monthly, and daily scales was found and it was correlated to temperature, salinity, and freshwater inflow, respectively. NEM correlated strongly with dissolved oxygen, temperature, and salinity, but the relationships where different in each bay. The dynamics of NEM on daily scales indicate that freshwater inflow events may be the main driver of NEM in the semiarid estuaries studied. The variable nature of NEM found here is further evidence that it is not valid to use single station monitoring deployments for assessment of whole estuarine ecosystem metabolic rates in large ecosystems. The relationship between NEM and temperature, salinity, and freshwater inflow events could drive predictive models assessing the potential influence of projected climate change and watershed development scenarios on estuarine metabolic rates.