Machine learning models for ecological footprint prediction based on energy parameters

Study of biocapacity areas to reduce ecological footprint deficits: A case study of Turkey

Our world has had difficulty meeting humans' needs in recent years. To ensure that the world can sustain its inhabitability and self-sufficiency in terms of natural resources, it is required to make the total amount of biocapacity areas equal to or higher than the ecological footprint. An analytical study has been carried out to remedy the biocapacity deficit by utilizing this information for Turkey and then these areas are optimized with heuristic optimization techniques. As a result, Artificial Bee Colony provides better objective function results (fewer errors) compared to Particle Swarm Optimization and Global Optimization Method Based on Clustering and Parabolic Approximation in terms of minimum, maximum, average value, and standard deviation. The rates of change according to the current situation of the biocapacity areas in 2016 are 277.97 %, 30.28 %, -29.28 %, 14.97 %, and -44.85 % for cropland, grazing land, forestland, fishing grounds, and built-up land, respectively. Depending on the population growth, these rates should additionally change by 83.24 %, -0.69 %, 3.97 %, 6.22 %, and -14.24 % respectively in 2050. The developed model can be used in industry or within the frame of government development policy and thus the balance between ecological footprint and biocapacity can be kept under control.

Exposing and understanding synergistic effects in co-pyrolysis of biomass and plastic waste via machine learning

During co-pyrolysis of biomass with plastic waste, bio-oil yields (BOY) could be either induced or reduced significantly via synergistic effects (SE). However, investigating/ interpreting the SE and BOY in multidimensional domains is complicated and limited. This work applied XGBoost machine-learning and Shapley additive explanation (SHAP) to develop interpretable/ explainable models for predicting BOY and SE from co-pyrolysis of biomass and plastic waste using 26 input features. Imbalanced training datasets were improved by synthetic minority over-sampling technique. The prediction accuracy of XGBoost models was nearly 0.90 R² for BOY while greater than 0.85 R² for SE. By SHAP, individual impact and interaction of input features on the XGBoost models can be achieved. Although reaction temperature and biomass-to-plastic ratio were the top two important features, overall contributions of feedstock characteristics were more than 60 % in the system of co-pyrolysis. The finding provides a better understanding of co-pyrolysis and a way of further improvements.

An overview of deterministic and probabilistic forecasting methods of wind energy

In recent years, a variety of wind forecasting models have been developed, prompting necessity to review the abundant methods to gain insights of the state-of-the-art development status. However, existing literature reviews only focus on a subclass of methods, such as multi-objective optimization and machine learning methods while lacking the full particulars of wind forecasting field. Furthermore, the classification of wind forecasting methods is unclear and incomplete, especially considering the rapid development of this field. Therefore, this article aims to provide a systematic review of the existing deterministic and probabilistic wind forecasting methods, from the perspectives of data source, model evaluation framework, technical background, theoretical basis, and model performance. It is expected that this work will provide junior researchers with broad and detailed information on wind forecasting for their future development of more accurate and practical wind forecasting models.

Hybrid Machine Learning Model for Electricity Consumption Prediction Using Random Forest and Artificial Neural Networks

Predicting electricity consumption is notably essential to provide a better management decision and company strategy. This study presents a hybrid machine learning model by integrating dimensionality reduction and feature selection algorithms with a backpropagation neural network (BPNN) to predict electricity consumption in Thailand. The predictive models are developed and tested using an actual dataset with related predictor variables from public sources. An open geospatial data gathered from a real service as well as geographical, climatic, industrial, household information are used to train, evaluate, and validate these models. Machine learning methods such as principal component analysis (PCA), stepwise regression (SWR), and random forest (RF) are used to determine the significant predictor variables. The predictive models are constructed using the BPNN with all available variables as baseline for comparison and selected variables from dimensionality reduction and feature selection methods. Along with creating a predictive model, the most related predictors of energy consumption are also selected. From the comparison, the hybrid model of RF with BPNN consistently outperforms the other models. Thus, the proposed hybrid machine learning model presented from this study can predict electricity consumption for planning and managing the energy demand.

Extreme fire weather is the major driver of severe bushfires in southeast Australia

In Australia, the proportion of forest area that burns in a typical fire season is less than for other vegetation types. However, the 2019-2020 austral spring-summer was an exception, with over four times the previous maximum area burnt in southeast Australian temperate forests. Temperate forest fires have extensive socio-economic, human health, greenhouse gas emissions, and biodiversity impacts due to high fire intensities. A robust model that identifies driving factors of forest fires and relates impact thresholds to fire activity at regional scales would help land managers and fire-fighting agencies prepare for potentially hazardous fire in Australia. Here, we developed a machine-learning diagnostic model to quantify nonlinear relationships between monthly burnt area and biophysical factors in southeast Australian forests for 2001-2020 on a 0.25° grid based on several biophysical parameters, notably fire weather and vegetation productivity. Our model explained over 80% of the variation in the burnt area. We identified that burnt area dynamics in southeast Australian forest were primarily controlled by extreme fire weather, which mainly linked to fluctuations in the Southern Annular Mode (SAM) and Indian Ocean Dipole (IOD), with a relatively smaller contribution from the central Pacific El Niño Southern Oscillation (ENSO). Our fire diagnostic model and the non-linear relationships between burnt area and environmental covariates can provide useful guidance to decision-makers who manage preparations for an upcoming fire season, and model developers working on improved early warning systems for forest fires.

Predicting ecological footprint based on global macro indicators in G-20 countries using machine learning approaches

Paying attention to human activities in terms of land grazing infrastructure, crops, forest products, and carbon impact, the so-called ecological impact (EF) is one of the most important economic issues in the world. For the present study, global database data were used. The ability of the penalized regression (RR) approaches (including Ridge, Lasso and Elastic Net) and artificial neural network (ANN) to predict EF indices in the G-20 countries over the past two decades (1999-2018) was illustrated and compared. For this purpose, 10-fold cross-validation was used to evaluate the predictive performance and determine the penalty parameter for PR models. According to the results, the predictive performance compared to linear regression improved somewhat using the penalized methods. Using the elastic net model, more global macro indices were selected than Lasso. Although Lasso selected only a few indicators, it had better predictive performance among PR ns models. In addition to relative improvement in the predictive performance of PR methods, their interest in selecting a subset of indicators by shrinking coefficients and creating a parsimonious model was evident and significant. As a result, PR methods would be preferred, using variable selection and interpretive considerations to predictive performance alone. On the other hand, ANN models with higher determination coefficients (R²) and lower RMSE values performed significantly better than PR and OLS and showed that they were more accurate in predicting EF. Therefore, ANN could provide considerable and appropriate predictions for EF indicators in the G-20 countries.