Prediction of water quality parameters of Karoon River (Iran) by artificial intelligence-based models

This paper describes the application of multi-layer perceptron (MLP), radial basis network and adaptive neuro-fuzzy inference system (ANFIS) models for computing dissolved oxygen (DO), biochemical oxygen demand (BOD) and chemical oxygen demand (COD) levels in the Karoon River (Iran). Nine input water quality variables including EC, PH, Ca, Mg, Na, Turbidity, PO₄, NO₃ and NO₂, which were measured in the river water, were employed for the models. The performance of these models was assessed by the coefficient of determination R ², root mean square error and mean absolute error. The results showed that the computed values of DO, BOD and COD using both the artificial neural network and ANFIS models were in close agreement with their respective measured values in the river water. MLP was also better than other models in predicting water quality variables. Finally, the sensitive analysis was done to determine the relative importance and contribution of the input variables. The results showed that the phosphate was the most effective parameters on DO, BOD and COD.     

Monitoring the seasonal dynamics of physicochemical parameters from Atoyac River basin (Puebla), Central Mexico: multivariate approach

Spatiotemporal variations of ten physicochemical parameters in the water quality of Atoyac River basin, Central Mexico, were obtained from 22 sampling sites (66 samples in total) located all along the basin for three different seasons (dry, rainy and winter). Multivariate statistical techniques such as correlation matrix, factor analysis (FA) and cluster analysis (CA) were used as a tool to understand the process. Physicochemical parameters such as temperature (T), pH, conductivity (λ), dissolved oxygen (DO), spectral absorption coefficient (SAC), oxidation–reduction potential (ORP), turbidity, 5-day biochemical oxygen demand (BOD₅), chemical oxygen demand (COD) and total suspended solids (TSS) were analyzed. Extremely high values of pH (10.24), conductivity (1870 µS/cm) and reduced redox potential (?370.1 mV) were observed in the dry season, whereas elevated TSS of 2996 mg/L was detected during the rainy season. The results elucidated high influence from the adjoining industrial, agricultural and urban zones, making the river unsuitable for life. FA generated varifactors, which accounted for cumulative % of 75.04 (dry), 76.22 (rainy) and 79.96 (winter) clearly grouping the external factors responsible for these significant values indicating the source of contamination. Cluster analysis facilitated the ease of classifying the sampling sites based on the similarities of physicochemical parameters. This study carried out in different seasons using multivariate statistical techniques would definitely prove to be an efficient tool for the restoration and establishing the real-time monitoring stations along this important river basin of Mexico.     

Simplified method for calculating inflow into a deep excavation with consideration of the effects of cutoff walls

Water inflow into deep excavations is a key parameter in the evaluation of environmental impact, and a simplified method is needed to calculate the inflow with adequate accuracy, especially for a deep excavation with a cutoff wall around it. This paper presents a model for calculating inflow based on the method of fragments, assuming two-dimensional sectional flow. The effects of cutoff wall thickness (w), wall penetration, and excavation width on the inflow are well considered in this model. Explicit formulas for the form factor are given. The accuracy and scope of this method are discussed. The method can be applied in both flooding and no flooding cases. Extensive verification shows that, for the flooding case and when one of four conditions is satisfied, the error of the proposed method will be within 10%. These conditions are b/d?≥?0.8 or w/d?≥?0.03 or s/d?≥?0.2 or s₁/d?≥?0.2, for which b is the half width of the excavation, d is the thickness of the soil layer underlying the wall, and s and s₁ are the wall penetration depths on the non-excavation side and excavation side, respectively. For the case of no flooding and when the ratio of the distance from the recharge boundary to the aquifer thickness (a/T) is?≥2, the error does not exceed 15%.     

Characterisation of the physicochemical qualities of a typical rural-based river: ecological and public health implications

The physicochemical qualities of a typical rural-based river were assessed over a 12-month period from August 2010 to July 2011 spanning the spring, summer, autumn and winter seasons. Water samples were collected from six sampling sites along Tyume River and analysed for total nitrogen, orthophosphate, biochemical oxygen demand (BOD), temperature, pH, dissolved oxygen (DO), electrical conductivity (EC), total dissolved solids (TDS) and turbidity. BOD regimes did not differ significantly between seasons and between sampling points and ranged from 0.78 to 2.76 mg/L across seasons and sampling points, while temperature ranged significantly (P < 0.05) between 6 and 28 °C. Turbidity varied significantly (P < 0.05) from 6 to 281 nephelometric turbidity units while TDS (range 24–209 ppm) and conductivity (range 47.6–408 mg/L) also varied significantly (P < 0.05) across sampling points with a remarkable similarity in their trends. Orthophosphate concentrations varied from 0.06 to 2.72 mg/L across seasons and sampling points. Negative correlations were noted between temperature and the nutrients, DO and temperature (r = ?0.56), and TDS and DO (r = ?0.33). Positive correlations were noted between TDS and temperature (r = 0.41), EC and temperature (r = 0.15), and DO and pH (r = 0.55). All nutrients were positively correlated to each other. Most measured parameters were within prescribed safety guidelines. However, the general trend was that water quality tended to deteriorate as the river flows through settlements, moreso in rainy seasons.     

A preliminary study on the environmental condition of the coral reef habitat

The present investigation was carried in coral reef areas of Palk Bay, Madapam. Palk Bay corals were found disturbed by human beings due to oil pollution, waste discharge from processing units and discharge of domestic household wastes from the nearby Mandapam town. Environmental conditions of the coral reef habitat in the Palk Bay, Mandapam was analyzed for a period of six months at fortnightly intervals for March to August 2001 on the coral reef areas. Temperature, pH, Dissolved Oxygen (DO), Biological Oxygen Demand (BOD) and Total Organic Carbon (TOC) were analyzed at five different stations of Palk Bay, Madapam. The maximum mean values of temperature, pH, DO, BOD and TOC for the four experimental and control stations were 32.6°C, 8.7, 7.02ppm, 8.29ppm and 2.13% and 31.5°C, 8.3, 7.98ppm, 5.71ppm and 0.48% respectively. The sampling stations were selected on the basis of their importance.     

The transfer of oxygen isotopic signals from precipitation to drip water and modern calcite on the seasonal time scale in Yongxing Cave,central China

Stable isotope data of precipitation (δ¹⁸O_p and deuterium excess), drip water (δ¹⁸O_d), and modern calcite precipitates (δ¹⁸O_c and δ¹³C_c) from Yongxing Cave, central China, are presented, with monthly sampling intervals from June 2013 to September 2016. Moderate correlations between the monthly variation of δ¹⁸O_p values (from ??11.5 to ??0.7‰) and precipitation amount (r = ??0.59, n?=?34, p?<?0.01) and deuterium excess (r?=?0.39, n?=?31, p?<?0.01) imply a combined effect of changes in precipitation amount and atmospheric circulation. At five drip sites, the δ¹⁸O_d values have a much smaller variability (from ??9.1 to ??7.5‰), without seasonal signals, probably a consequence of the mixing in the karst reservoir with a deep aquifer. The mean δ¹⁸O_d value (??8.4‰) for all drip waters is significantly more negative than the mean δ¹⁸O_p value (??6.9‰) weighted by precipitation amount, but close to the wet season (May to September) mean value (??8.3‰), suggesting that a threshold of precipitation amount must be exceeded to provide recharge. Calculation based on the equilibrium fractionation factor indicates that the δ¹⁸O_c values are not in isotopic equilibrium with their corresponding drip waters, with a range of disequilibrium effects from 0.4 to 1.4‰. The δ¹⁸O_c and δ¹³C_c values generally increase progressively away from the locus of precipitation on glass plates. The disequilibrium effects in the cave are likely caused by progressive calcite precipitation and CO₂ degassing related to a high gradient of CO₂ concentration between drip waters and cave air. Our study provides an important reference to interpret δ¹⁸O_c records from the monsoon region of China.     

Effects of land-based sources on water quality in the Omerli reservoir (Istanbul,Turkey)

The Omerli reservoir is located on the northeastern side of the Istanbul City. It is one of the most important sources of drinking water with a surface area of 23.1 km² and a volume of 220 × 10⁶ m³ in the Marmara Region. Water quality characteristics of the lake investigated from May 2002 to April 2003 enabled us to identify the effect of waste loads on water quality of Omerli Lake. The following parameters were measured in the lake water: temperature (16.1°C), conductivity (250 μS/cm), secchi disk depth (1.9 m), dissolved oxygen (DO) (9.36 mg/L), nutrients [ortho-phosphate (78.9 μg/L), nitrate + nitrite (707.5 μg/L) and ammonia-nitrogen (264 μg/L)], chlorophyll a (9.43 μg/L), total organic carbon (3.33 mg/L), total suspended solids (4.54 mg/L), total poly aromatic hydrocarbon (t-PAH) (0.69 μg/L) and copper (24.5 μg/L). T-PAH (16.5 mg/g-dry-w), Cu (96.5 μg/g-dry-w), organic carbon (org-C) (2.0%) and N (0.44%) were determined in the surface sediment. The values of chlorophyll a and DO in the upper layer were relatively high and low secchi disk depths indicates eutrophic state. There are five channels discharging water, including waste water, into the lake. All channels were sampled during six occasions in order to cover variations between seasons. The following parameters were measured: total organic carbon, total phosphorus, total kjeldahl nitrogen, ammonia and total suspended solids with the flow rates. The Göçbeyli stream has the highest flow into the lake (1.5 × 10⁸ m³/year) but most of the nutrients were discharged from the Pa?aköy channel. It is accounting for 81% of ammonia and 80% of total phosphorus into the lake. Three scenarios were run using AQUATOX model: (1) all existing inflows are discharging into the lake (present situation); (2) none is discharging; (3) all are discharging except the Pa?aköy channel. The first scenario produced concentrations consistent with measurements in the lake. In all stations, a phytoplankton peak value was predicted during November and December 2002, and January 2003. In the second scenario, as expected, a significant decrease in the concentrations was predicted. In the third scenario, a small improvement in the water quality was obtained. To significantly improve the state of the lake, instead of entering Pa?aköy channel, wastewater should bypass the lake.     

Water quality parameters along rivers

Water samples have been collected from a part of Surma River along different points and analyzed for various water quality parameters during dry and monsoon periods. Effects of industrial wastes, municipal sewage, and agricultural runoff on river water quality have been investigated. The study was conducted within the Chattak to Sunamganj portion of Surma River, which is significant due to the presence of two major industries-a paper mill and a cement factory. The other significant feature is the conveyors that travel from India to Chattak. This study involves determination of physical, biological and chemical parameters of surface water at different points. The river was found to be highly turbid in the monsoon season. But BOD and fecal coliform concentration was found higher in the dry season. The water was found slightly acidic. The mean values of parameters were Conductivity 84–805μs; DO: dry- 5.52 mg/L, monsoon-5.72 mg/L; BOD: dry-1 mg/L, monsoon-0.878 mg/L; Total Solid: dry-149.4 mg/L, monsoon- 145.7 mg/L. A model study was also conducted and values of different model parameters were estimated.     

A comparative analysis among computational intelligence techniques for dissolved oxygen prediction in Delaware River

Most of the water quality models previously developed and used in dissolved oxygen (DO) prediction are complex. Moreover, reliable data available to develop/calibrate new DO models is scarce. Therefore, there is a need to study and develop models that can handle easily measurable parameters of a particular site, even with short length. In recent decades, computational intelligence techniques, as effective approaches for predicting complicated and significant indicator of the state of aquatic ecosystems such as DO, have created a great change in predictions. In this study, three different AI methods comprising: (1) two types of artificial neural networks (ANN) namely multi linear perceptron (MLP) and radial based function (RBF); (2) an advancement of genetic programming namely linear genetic programming (LGP); and (3) a support vector machine (SVM) technique were used for DO prediction in Delaware River located at Trenton, USA. For evaluating the performance of the proposed models, root mean square error (RMSE), Nash–Sutcliffe efficiency coefficient (NS), mean absolute relative error (MARE) and, correlation coefficient statistics (R) were used to choose the best predictive model. The comparison of estimation accuracies of various intelligence models illustrated that the SVM was able to develop the most accurate model in DO estimation in comparison to other models. Also, it was found that the LGP model performs better than the both ANNs models. For example, the determination coefficient was 0.99 for the best SVM model, while it was 0.96, 0.91 and 0.81 for the best LGP, MLP and RBF models, respectively. In general, the results indicated that an SVM model could be employed satisfactorily in DO estimation.     

Determination of Minor Contents of Smectite Layers in the Dispersed Dioctahedral K-bearing Micaceous Minerals of the Illite,Aluminoceladonite, and Glauconite Composition

A theoretically substantiated technique has been elaborated for the precise determination of minor contents of smectite layers in the dioctahedral mixed-layer structures with a disordered alternation of K-bearing micaceous and ethylene glycol-saturated smectite layers. It is shown that observed positions of the possible mixed-layer mica–smectite structures in the d(003)_obs–d(007)_obs or 2θ(003)_obs–2θ(007)_obs diagrams are distributed along a set of parallel straight lines. Data points of each straight line correspond to the mixed-layer structures with one and the same content of smectite layers. A complete matching of theoretical model was recorded in the positions of d(003)_obs–d(007)_obs and 2θ(003)_obs–2θ(007)_obs pairs of the basal reflections determined in diffractograms simulated for 160 two-component mixed-layer mica–smectite structures differing in the thickness of alternating layers and their relative content. We studied samples of the dioctahedral K-bearing micas of the glauconite, illite, and aluminoceladonite composition with different thickness of micaceous layers. Positions of the mixed-layer structures of the studied samples on the 2θ(003)_obs–2θ(007)_obs and d(003)_obs–d(007)_obs plots make it possible to visually determine the content of smectite layers in each structure accurate to 0.5%. The content of smectite layers in the studied samples varied from 2 to 15%. We obtained equations relating the thickness of smectite and micaceous layers for a specified content of smectite layers. They make it possible to determine the thickness correlation between specified mica and smectite layers. Analogous equations can be used to calculate the Wsm value in each sample for specified experimental values of d(001)_mc and d(001)_sm and d(003)_obs or d(007)_obs. The W_sm values coincide within an error limit of 0.2% with those based on the visual estimates in plots.     

Relationships Between Long-Term Trend of Satellite-Derived Chlorophyll-<Emphasis Type="Italic">a</Emphasis> and Hypoxia Off the Changjiang Estuary

The Changjiang Estuary is one of the largest estuaries in the world, where hypoxia frequently occurs during the summer season. Recent routine surveys in the observed area found that the low dissolved oxygen (DO) in the summer bottom water was not rapidly expanding with increasing nutrient loads in Changjiang diluted waters. Based on the remote sensing data and in situ data, we examined the historic seasonal hypoxia observations for the bottom waters of the Changjiang Estuary and investigated the short- and long-term effects that runoff had on variations in DO and chlorophyll-a (Chl-a). Our analysis indicated that the recent areal variation in hypoxia was due to the changing Chl-a distribution and stratification conditions. The correspondence between hypoxia and surface Chl-a concentration showed that remotely sensed Chl-a larger than 3.0 mg L^?1 was an essential condition for the formation of hypoxia off the Changjiang Estuary. The trend of Chl-a concentration was significantly impacted by the Three Gorges Dam (TGD), and the inter-annual variation of Chl-a was weakly affected by global-scale climate variability. After the TGD impoundment, the sediment loading in the Changjiang runoff and suspended sediments in Changjiang Estuary in August decreased, and the high Chl-a concentration moved landward. These shifted the hypoxia from its optimal forming conditions.     

Factors of <Superscript>18</Superscript>O/<Superscript>16</Superscript>O fractionation in corundum estimated from the calculation of isotopic shifts on vibration frequencies

The β-factors of corundum were estimated on the basis of DFT calculations of vibrational frequency changes due to ¹⁶O–¹⁸O isotope substitution in a harmonic approximation using an all-electron Gaussian-type basis set and the B3LYP hybrid functional (the CRYSTAL program). Calculations were performed accounting for eight phonon wave vectors within the first Brillouin zone. The results are approximated by the relation 1000ln β _crn = 9.19874x–0.12326x ² + 0.00213x ³ (x = 10⁶/T(K)², 400 < T(K) < 1300), which can be used in isotope geochemical studies in combination with the known temperature effects on the β-factors of other phases.     

Distribution of inhomogeneities in the interstellar plasma in the directions of three distant pulsars from observations with the RadioAstron ground–space interferometer

The RadioAstron ground–space interferometer has been used to measure the angular sizes of the scattering disks of the three distant pulsars B1641–45, B1749–28, and B1933+16. The observations were carried out with the participation of the Westerbork Synthesis Radio Telescope; two 32-m telescopes at Torun, Poland and Svetloe, Russia (the latter being one antenna of the KVAZAR network); the Saint Croix VLBA antenna; the Arecibo radio telescope; the Parkes, Narrabri (ATCA), Mopra, Hobart, and Ceduna Australian radio telescopes; and the Hartebeesthoek radio telescope in South Africa. The full widths at half maximum of the scattering disks were 27 mas at 1668 MHz for B1641–45, 0.5 mas at 1668 MHz for B1749–28, and 12.3 at 316 MHz and 0.84 mas at 1668 MHz for B1933+16. The characteristic time scales for scatter-broadening of the pulses on inhomogeneities in the interstellar plasma τ_sc were also measured for these pulsars using various methods. Joint knowledge of the size of the scattering disk and the scatter-broadening time scale enables estimation of the distance to the effective scattering screen d. For B1641–45, d = 3.0 kpc for a distance to the pulsar D = 4.9 kpc, and for B1749–28, d = 0.95 kpc for D = 1.3 kpc. Observations of B1933+16 were carried out simultaneously at 316 and 1668 MHz. The positions of the screen derived using the measurements at the two frequencies agree: d ₁ = 2.6 and d ₂ = 2.7 kpc, for a distance to the pulsar of 3.7 kpc. Two screens were detected for this pulsar from an analysis of parabolic arcs in the secondary dynamic spectrum at 1668 MHz, at 1.3 and 3.1 kpc. The scattering screens for two of the pulsars are identified with real physical objects located along the lines of sight toward the pulsars: G339.1–04 (B1641–45) and G0.55–0.85 (B1749–28).     

Mathematical expression of discharge capacity of compound open channels using MARS technique

In this paper, analytical methods, artificial neural network (ANN) and multivariate adaptive regression splines (MARS) techniques were utilised to estimate the discharge capacity of compound open channels (COC). To this end, related datasets were collected from literature. The results showed that the divided channel method with a coefficient of determination (R ²) value of 0.76 and root mean square error (RMSE) value of 0.162 has the best performance, among the various analytical methods tested. The performance of applied soft computing models with R ²=0.97 and RMSE = 0.03 was found to be more accurate than analytical approaches. Comparison of MARS with the ANN model, in terms of developed discrepancy ratio (DDR) index, showed that the accuracy of MARS model was better than that of MLP model. Reviewing the structure of the derived MARS model showed that the longitudinal slope of the channel (S), relative flow depth (H _r ) and relative area (A _r ) have a high impact on modelling and forecasting the discharge capacity of COCs.     

Use of SWAT to determine the effects of climate and land use changes on streamflow and sediment concentration in the Purna River basin,India

A semi-distributed, physically based, basin-scale Soil and Water Assessment Tool (SWAT) model was developed to determine the key factors that influence streamflow and sediment concentration in Purna river basin in India and to determine the potential impacts of future climate and land use changes on these factors. A SWAT domain with a Geographical Information System (GIS) was utilized for simulating and determining monthly streamflow and sediment concentration for the period 1980–2005 with a calibration period of 1980–1994 and validation period of 1995 to 2005. Additionally, a sequential uncertainty fitting (SUFI-2) method within SWAT-CUP was used for calibration and validation purpose. The overall performance of the SWAT model was assessed using the coefficient of determination (R²) and Nash–Sutcliffe efficiency parameter (E_NS) for both calibration and validation. For the calibration period, the R² and E_NS values were determined to be 0.91 and 0.91, respectively. For the validation period, the R² and E_NS were determined to be 0.83 and 0.82, respectively. The model performed equally well with observed sediment data in the basin, with the R² and E_NS determined to be 0.80 and 0.75 for the calibration period and 0.75 and 0.65 for the validation, respectively. The projected precipitation and temperature show an increasing trend compared to the baseline condition. The study indicates that SWAT is capable of simulating long-term hydrological processes in the Purna river basin.     

Effect of Slope on <Emphasis Type="Italic">p</Emphasis>-<Emphasis Type="Italic">y</Emphasis> Curves for Laterally Loaded Piles in Soft Clay

Pile foundations are often subject to lateral loading due to various forces on a variety of structures like high rise buildings, transmission towers, power stations, offshore structures and highway and railway structures. The present investigation is to study the effect of slopes on p-y curves (where p is the static soil reaction and y is the pile deflection) due to static lateral loading in soft clay (Consistency index I_c = 0.42). A series of laboratory model tests were carried out on the instrumented model pile on sloping ground (slopes of 1V:1H, 1V:1.5H, 1V:2H, 1V:3H and 1V:5H) and with varying embedment length to diameter ratio (L/D) of 20, 25 and 30. From the experimental results, the bending moment curves along the pile shaft are double differentiated to obtain the soil resistance (p) and double integrated to obtain the deflection (y) using curve fitting method. New p-y curves for piles located on crest of soft clay with different sloping ground surface under static lateral loading are developed. Moreover, the effect of sloping angles on proposed p-y curves was studied.     

Abundances of carbon,nitrogen, oxygen,and other elements in the atmospheres of the giants 25 Mon and HR 7389

An analysis of high-resolution CCD spectra of the giant 25 Mon, which shows signs of metallicity, and the normal giant HR 7389 is presented. The derived effective temperatures, gravitational accelerations, and microturbulence velocities are T_eff = 6700 K, log g = 3.24, and ξ _t = 3.1 km/s for 25 Mon and T_eff = 6630 K, log g = 3.71, and ξ _t = 2.6 km/s for HR 7389. The abundances (log ε) of nine elements are determined: carbon, nitrogen, oxygen, sodium, silicon, calcium, iron, nickel, and barium. The derived excess carbon abundances are 0.23 dex for 25 Mon and 0.16 dex for HR 7389. 25 Mon displays a modest (0.08 dex) oxygen excess, with the oxygen excess for HR 7389 being somewhat higher (0.15 dex). The nitrogen abundance is probably no lower than the solar value for both stars. The abundances of iron, sodium, calcium (for HR 7389), barium, and nickel exceed the solar values by 0.22–0.40 dex for both stars. The highest excess (0.62 dex) is exhibited by the calcium abundance for 25 Mon. Silicon displays a nearly solar abundance in both stars—small deficits of ?0.03 dex and ?0.07 dex for 25 Mon and HR 7389, respectively. No fundamental differences in the elemental abundances were found in the atmospheres of 25 Mon and HR 7389. Based on their T_eff and log g values, as well as theoretical calculations, A. Claret estimated the masses, radii, luminosities, and ages of 25 Mon (M/M_⊙ = 2.45, log(R/R_⊙) = 0.79, log(L/L_⊙) = 1.85, t = 5.3 × 10⁸ yr) and HR 7389 (M/M_⊙ = 2.36, log(R/R_⊙) = 0.50, log(L/L_⊙) = 1.24, t = 4.6 × 10⁸ yr), and also of the stars 20 Peg (M/M_⊙ = 2.36, log(R/R_⊙) = 0.73, log(L/L_⊙) = 1.79, t = 4.9 × 10⁸ yr) and 30 LMi (M/M_⊙ = 2.47, log(R/R_⊙) = 0.73, log(L/L_⊙) = 1.88, t = 4.8 × 10⁸ yr) studied by the author earlier.     

Investigating the impact of the properties of pilot points on calibration of groundwater models: case study of a karst catchment in Rote Island,Indonesia

A robust configuration of pilot points in the parameterisation step of a model is crucial to accurately obtain a satisfactory model performance. However, the recommendations provided by the majority of recent researchers on pilot-point use are considered somewhat impractical. In this study, a practical approach is proposed for using pilot-point properties (i.e. number, distance and distribution method) in the calibration step of a groundwater model. For the first time, the relative distance–area ratio (d/A) and head-zonation-based (HZB) method are introduced, to assign pilot points into the model domain by incorporating a user-friendly zone ratio. This study provides some insights into the trade-off between maximising and restricting the number of pilot points, and offers a relative basis for selecting the pilot-point properties and distribution method in the development of a physically based groundwater model. The grid-based (GB) method is found to perform comparably better than the HZB method in terms of model performance and computational time. When using the GB method, this study recommends a distance–area ratio of 0.05, a distance–x-grid length ratio (d/X _grid) of 0.10, and a distance–y-grid length ratio (d/Y _grid) of 0.20.     

Prediction of strength parameters of sedimentary rocks using artificial neural networks and regression analysis

Accurate laboratory measurement of geo-engineering properties of intact rock including uniaxial compressive strength (UCS) and modulus of elasticity (E) involves high costs and a substantial amount of time. For this reason, it is of great necessity to develop some relationships and models for estimating these parameters in rock engineering. The present study was conducted to forecast UCS and E in the sedimentary rocks using artificial neural networks (ANNs) and multivariable regression analysis (MLR). For this purpose, a total of 196 rock samples from four rock types (i.e., sandstone, conglomerate, limestone, and marl) were cored and subjected to comprehensive laboratory tests. To develop the predictive models, physical properties of studied rocks such as P wave velocity (V_p), dry density (γ_d), porosity, and water absorption (A_b) were considered as model inputs, while UCS and E were the output parameters. We evaluated the performance of MLR and ANN models by calculating correlation coefficient (R), mean absolute error (MAE), and root-mean-square error (RMSE) indices. The comparison of the obtained results revealed that ANN outperforms MLR when predicting the UCS and E.     

The effect of different meteorological parameters on the temporal variations of reference evapotranspiration

Temporal changes of meteorological variables can affect reference evapotranspiration (ET₀). The goal of the present research is to analyze the changes of ET₀ and identify the impact of effective meteorological parameters to the changes of ET₀. For this purpose, daily meteorological data recorded in 30 synoptic stations of Iran during 1960–2014 were used. The annual and seasonal values of ET₀ were calculated by the recorded data. To calculate ET₀, FAO56 Penman–Monteith method (standard method) was used. The annual and seasonal trends of ET₀ and its eight effective parameters were analyzed. Then the contributions of effective parameters changes on ET₀ were determined. To analyze ET₀ trend at annual and seasonal scales, two common methods, Spearman’s Rho and Mann–Kendall tests, were used. The R ² = 0.99 showed that the results of the mentioned methods were similar and on the basis of T-statistic <0.057, their difference was not significant (95% confidence level). Therefore, only one method’s results (Spearman’s Rho) were reported. On the basis of Spearman’s Rho results, the annual and seasonal values of ET₀ had negative trend in most of arid and semi-arid stations while the trend of this parameter was positive in humid and very humid stations. At annual and seasonal scales, decreasing in wind speed (W), temperature (T), sunshine hours (n), minimum temperature (TN), dew point temperature (TD), maximum temperature (TX), saturation vapor pressure deficit (SVPD) and solar radiation (RS) was observed in 58, 54, 39, 43, 56, 65, 65 and 37% studied stations, respectively. In many scales, the results showed that TX and W were the most effective meteorological variables on ET₀ changes and then SVPD was located in second step in arid and semi-arid stations. In humid and very humid stations, W was the first effective parameter at all scales, except autumn.