Temperature forecasting plays an important role in human production and operational activities. Traditional temperature forecasting mainly relies on numerical forecasting models to operate, which takes a long time and has higher requirements for the computing power and storage capacity of computers. In order to reduce computation time and improve forecast accuracy, deep learning-based temperature forecasting has received more and more attention. Based on the atmospheric temperature, dew point temperature, relative humidity, air pressure, and cumulative wind speed data of five cities in China from 2010 to 2015 in the UCI database, multivariate time series atmospheric temperature forecast models based on recurrent neural networks (RNN) are established. Firstly, the temperature forecast modeling of five cities in China is established by RNN for five different model configurations; secondly, the neural network training process is controlled by using the Ridge Regularizer (L2) to avoid overfitting and underfitting; and finally, the Bayesian optimization method is used to adjust the hyper-parameters such as network nodes, regularization parameters, and batch size to obtain better model performance. The experimental results show that the atmospheric temperature prediction error based on LSTM RNN obtained a minimum error compared to using the base models, and these five models obtained are the best models for atmospheric temperature prediction in the corresponding cities. In addition, the feature selection method is applied to the established models, resulting in simplified models with higher prediction accuracy.

The occurrence of pollution indicator bacteria (total and faecal coliform) has been used as a sanitary parameter for evaluating the quality of drinking water. It is known that these indicators are associated with disease causing organisms which are of great concern to public health. This study assessed the relationship between coliform bacteria and water geochemistry in surface and ground water systems in the Tarkwa mining area using logistic regression models. In surface water sources, higher values of chloride (OR = 0.891, p<005), phosphates (OR = 0.452, p<0.05), pH (OR = 0.174, p<0.05) and zinc (OR = 0.001, p<0.05) were associated with lower odds of faecal coliform contamination. In groundwater sources, higher values of phosphates (OR = 0.043, p<0.001), total dissolved solids (OR = 0.858, p<0.05), turbidity (OR = 0.996, p<0.05) and nickel (OR = 6.09E-07, p<0.05) implied non-contamination by faecal coliform. However, higher values of electrical conductivity (OR = 1.097, p<0.05), nitrates (OR = 1.191, p<0.05) and total suspended solids (OR = 1.023, p<0.05) were associated with higher odds of faecal coliform contamination of groundwater sources. Nitrates and total suspended solids, in this case, were completely mediated by the heavy metals. For total coliform in surface water systems, higher values of magnesium (OR = 1.070, p<0.05) was associated with higher odds of total coliform contamination while higher values of phosphates (OR = 0.968, p<0.05) was associated with lower odds of total coliform contamination although the presence of heavy metals completely mediated these relationships. For ground water systems, higher values of pH (OR = 0.083, p<0.05), phosphates (OR = 0.092, p<0.05), turbidity (OR = 0.950, p<0.05) and chloride (OR = 0.860, p<0.05) were associated with lower odds of total coliform contamination. However, higher values of total suspended solids (OR = 1.054, p<0.05) and nitrates (OR = 1.069, p<0.05) implied contamination of total coliform in ground water sources. The relationship between nitrates and total coliform were mediated by the heavy metals. This study establishes the need to monitor, manage and remediate surface and ground water sources for potential disease causing microbes in ways that takes into consideration the factors that create different conditions in the two water systems. This study validates the usefulness of statistical models as tools for preventing surface and ground water contamination.