Novel hybrid linear stochastic with non-linear extreme learning machine methods for forecasting monthly rainfall a tropical climate.

A novel hybrid approach is presented that can more accurately predict monthly rainfall in a tropical climate by integrating a linear stochastic model with a powerful non-linear extreme learning machine method. This new hybrid method was then evaluated by considering four general scenarios. In the first scenario, the modeling process is initiated without preprocessing input data as a base case. While in other three scenarios, the one-step and two-step procedures are utilized to make the model predictions more precise. The mentioned scenarios are based on a combination of stationarization techniques (i.e., differencing, seasonal and non-seasonal standardization and spectral analysis), and normality transforms (i.e., Box-Cox, John and Draper, Yeo and Johnson, Johnson, Box-Cox-Mod, log, log standard, and Manly). In scenario 2, which is a one-step scenario, the stationarization methods are employed as preprocessing approaches. In scenario 3 and 4, different combinations of normality transform, and stationarization methods are considered as preprocessing techniques. In total, 61 sub-scenarios are evaluated resulting 11013 models (10785 linear methods, 4 nonlinear models, and 224 hybrid models are evaluated). The uncertainty of the linear, nonlinear and hybrid models are examined by Monte Carlo technique. The best preprocessing technique is the utilization of Johnson normality transform and seasonal standardization (respectively) (R2 = 0.99; RMSE = 0.6; MAE = 0.38; RMSRE = 0.1, MARE = 0.06, UI = 0.03 &UII = 0.05). The results of uncertainty analysis indicated the good performance of proposed technique (d-factor = 0.27; 95PPU = 83.57). Moreover, the results of the proposed methodology in this study were compared with an evolutionary hybrid of adaptive neuro fuzzy inference system (ANFIS) with firefly algorithm (ANFIS-FFA) demonstrating that the new hybrid methods outperformed ANFIS-FFA method.

Directed Self-Assembly of Polarizable Ellipsoids in an External Electric Field.

The interplay between shape anisotropy and directed long-range interactions enables the self-assembly of complex colloidal structures. As a recent highlight, ellipsoidal particles polarized in an external electric field were observed to associate into well-defined tubular structures. In this study, we systematically investigate such directed self-assembly using Monte Carlo simulations of a two-point-charge model of polarizable prolate ellipsoids. In spite of its simplicity and computational efficiency, we demonstrate that the model is capable of capturing the complex structures observed in experiments on ellipsoidal colloids at low volume fractions. We show that, at sufficiently high electric field strength, the anisotropy in shape and electrostatic interactions causes a transition from three-dimensional crystal structures observed at low aspect ratios to two-dimensional sheets and tubes at higher aspect ratios. Our work thus illustrates the rich self-assembly behavior accessible when exploiting the interplay between competing long- and short-range anisotropic interactions in colloidal systems.

Application of multi-objective genetic algorithm for optimal combination of resources to achieve sustainable agriculture based on the water-energy-food nexus framework.

Understanding the systemic approach and its potential for decision-making is important for resource management, especially in agriculture in which increasing food demands and environmental and social issues are the main challenges. Therefore, multiple-criteria decision-making methods have a vital role in the optimum combination of resources. Computational models are commonly used to assist resource management decision-making; however, while water-energy-food nexus (WEFN) are increasingly well modeled, the inclusion of social issues has lagged behind. This paper outlines a model based on a multi-objective genetic algorithm (MOGA) that conceptualizes and proceduralizes balancing the goals of sustainable agricultural development highlighting impacts and interactions between social variables and the WEFN index in agriculture. The model was developed using a bottom-up approach, informed through farmer interviews, and secondary data in the Miandarband plain, west Iran. The Compromise Programming (CP) method, which is widely used to solve MOGA models, was applied to optimization algorithms in three-dimensional spaces. The model represents field conditions and provides a tool for policymakers and sustainable resource management. The modeling framework applied to the study area for the comparison of WFEN, life cycle assessment (LCA), and social dimension in current and optimum cultivation patterns. The proposed optimal cultivation pattern in minimum CP will reduce water and energy consumption by 2.56 % and 12.71 % while reducing environmental impacts by 6.82 %, and it will improve the social status of farmers. Results suggest that changes in the basic elements of objective functions will lead to a balance between cultivation patterns that depends on policies and socio-economic conditions. Moreover, proposed cultivation patterns may be sustainable but their viability varies across the periods and also in different human ecologies. However, by analyzing the feedback of the model and interactions between different dimensions, this work highlights that policymakers can decide sustainable agriculture how should be occur by comparing different solutions.

Mapping the spatial and temporal variability of flood susceptibility using remotely sensed normalized difference vegetation index and the forecasted changes in the future.

Accurate runoff forecasting plays a considerable role in the appropriate water resource planning and management. The spatial and temporal evaluation of the flood susceptibility was explored in the Quebec basin, Canada. This study provides a new strategy for runoff modelling as one of the complicated variables by developing new machine learning techniques along with remote sensing. A novel scheme of the Group Method of Data Handling (GMDH) known as the generalized structure of GMDH (GSGGMDH) is developed to overcome this classical approach's limitation. A simple time series based scenario with exogenous variables including precipitation and Normalized Difference Vegetation Index (NDVI) was introduced for runoff forecasting. MODIS data included MOD13Q1 product was employed and a JavaScript code was developed to preprocess collected data in the Google Earth Engine (GEE) environment. Using different seasonal and non-seasonal lags of all input variables, the developed GSGMDH found the most optimum input combination for each station in terms of simplicity and accuracy, simultaneously (average values; SI = 0.554, RMSRE = 1.55, MAE = 5.076). The precipitation values are modelled with the CanEsm2 climate change model. To apply NDVI for runoff forecasting, a simple spatial-temporal GSGMDH based model was developed (average values; SI = 0.27; RMSRE = 8.27, MAE = 0.08). The forecasting results indicated that the months in which the maximum runoff occurred have changed, and these months have increased compared to the historic period. In the historical period, the frequency of maximum runoff was in April and March. Still, for the two forecasting periods (i.e. 2020-2039 and 2040-2059), the months in which the maximum runoff has occurred have changed, and their amount has been reduced and added to other months, especially February and August.

A generalized linear stochastic model for lake level prediction.

Endorheic lakes are one of the most important factors of an environment. Regarding their morphology, these lakes, in particular saline lakes, are much more sensitive and can either benefit or pose a threat to their surroundings. Thus, constant monitoring of such lakes' water level, modeling and analyzing them for future planning and management policies is vitally important. We proposed a generalized linear stochastic model (GLSM) for forecasting the weekly and monthly Urmia lake water levels, the sixth-largest saltwater lake on Earth. In this methodology, three approaches are defined to pre-process data. The first approach is merely based on the differencing method, while the second and third are a one-step (the combination of de-trending with standardization and spectral analysis) and two-step (the combination of the 2nd approach with normalization transform) preprocessing, respectively. A thorough comparison of the GLSM results with eminence nonlinear AI models (Adaptive Neuro-Fuzzy Inference Systems, ANFIS, Multilayer Perceptron, MLP, Gene Expression Programming, GEP, Support Vector Machine with Firefly algorithm, SVM-FFA, and Artificial Neural Networks ANN) showed that by using an appropriate method that delivers accurate information of the entailing terms in time series, it is possible to model Urmia lake level with acceptable precision. Concisely, the GSLM with coefficients of determination (R2) 99.957% and root mean squared error (RMSE) of 2.121% outperformed the SVM-FFA with R2 99.59%, RMSE 3.27%, ANN with R2 99.56%, RMSE 3.3%, ANFIS with R2 98.9%, RMSE 4.3%, GP with R2 99.89%, RMSE 3.47%, GEP with R2 94.75%, RMSE 4.15% for forecasting weekly time series. In forecasting monthly time series, the GLSM method with R2 99.517% and RMSE 6.91% also outperformed GEP R2 91.95%, RMSE 15.3%, ANFIS R2 92.85%, RMSE 47.55% models. Consequently, GSLM proved that by applying proper comprehensible linear techniques promising results can be obtained rather than using sophisticated AI methods.