Machine Learning in Weather Prediction and Climate Analyses—Applications and Perspectives

Comparison of RNN-LSTM, TFDF and stacking model approach for weather forecasting in Bangladesh using historical data from 1963 to 2022

Forecasting the weather in an area characterized by erratic weather patterns and unpredictable climate change is a challenging endeavour. The weather is classified as a non-linear system since it is influenced by various factors that contribute to climate change, such as humidity, average temperature, sea level pressure, and rainfall. A reliable forecasting system is crucial in several industries, including transportation, agriculture, tourism, & development. This study showcases the effectiveness of data mining, meteorological analysis, and machine learning techniques such as RNN-LSTM, TensorFlow Decision Forest (TFDF), and model stacking (including ElasticNet, GradientBoost, KRR, and Lasso) in improving the precision and dependability of weather forecasting. The stacking model strategy entails aggregating multiple base models into a meta-model to address issues of overfitting and underfitting, hence improving the accuracy of the prediction model. To carry out the study, a comprehensive 60-year meteorological record from Bangladesh was gathered, encompassing data on rainfall, humidity, average temperature, and sea level pressure. The results of this study suggest that the stacking average model outperforms the TFDF and RNN-LSTM models in predicting average temperature. The stacking average model achieves an RMSLE of 1.3002, which is a 10.906% improvement compared to the TFDF model. It is worth noting that the TFDF model had previously outperformed the RNN-LSTM model. The performance of the individual stacking model is not as impressive as that of the average model, with the validation results being better in TFDF.

Short-Term Precipitation Radar Echo Extrapolation Method Based on the MS-DD3D-RSTN Network and STLoss Function

Short-term precipitation forecasting is essential for agriculture, transportation, urban management, and tourism. The radar echo extrapolation method is widely used in precipitation forecasting. To address issues like forecast degradation, insufficient capture of spatiotemporal dependencies, and low accuracy in radar echo extrapolation, we propose a new model: MS-DD3D-RSTN. This model employs spatiotemporal convolutional blocks (STCBs) as spatiotemporal feature extractors and uses the spatial-temporal loss (STLoss) function to learn intra-frame and inter-frame changes for end-to-end training, thereby capturing the spatiotemporal dependencies in radar echo signals. Experiments on the Sichuan dataset and the HKO-7 dataset show that the proposed model outperforms advanced models in terms of CSI and POD evaluation metrics. For 2 h forecasts with 20 dBZ and 30 dBZ reflectivity thresholds, the CSI metrics reached 0.538, 0.386, 0.485, and 0.198, respectively, representing the best levels among existing methods. The experiments demonstrate that the MS-DD3D-RSTN model enhances the ability to capture spatiotemporal dependencies, mitigates forecast degradation, and further improves radar echo prediction performance.

Assessment of meteorological parameters in predicting seasonal temperature of Dhaka city using ANN

In this study, an attempt has been made to investigate the possibility of a machine learning model, Artificial Neural Network (ANN) for seasonal prediction of the temperature of Dhaka city. Prior knowledge of temperature is essential, especially in tropical regions like Dhaka, as it aids in forecasting heatwaves and implementing effective preparedness schemes. While various machine learning models have been employed for the prediction of hot weather across the world, research specially focused on Bangladesh is limited. Additionally, the application of machine learning models needs to be curated to suit the particular weather features of any region. Therefore, this study approaches ANN method for prediction of the temperature of Dhaka exploring the underlying role of related weather variables. Using the daily data for the months of February to July collected from the National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data (0.25° × 0.25° global grid) for the years 2011-2020, this study focuses on finding the combination of weather variables in predicting temperatures. The densely populated city, Dhaka, has faced severe consequences due to extreme climate conditions in recent years, and this study will pave a new dimension for further research regarding the topic.

Multiple serial correlations in global air temperature anomaly time series

Serial correlations within temperature time series serve as indicators of the temporal consistency of climate events. This study delves into the serial correlations embedded in global surface air temperature (SAT) data. Initially, we preprocess the SAT time series to eradicate seasonal patterns and linear trends, resulting in the SAT anomaly time series, which encapsulates the inherent variability of Earth's climate system. Employing diverse statistical techniques, we identify three distinct types of serial correlations: short-term, long-term, and nonlinear. To identify short-term correlations, we utilize the first-order autoregressive model, AR(1), revealing a global pattern that can be partially attributed to atmospheric Rossby waves in extratropical regions and the Eastern Pacific warm pool. For long-term correlations, we adopt the standard detrended fluctuation analysis, finding that the global pattern aligns with long-term climate variability, such as the El Niño-Southern Oscillation (ENSO) over the Eastern Pacific. Furthermore, we apply the horizontal visibility graph (HVG) algorithm to transform the SAT anomaly time series into complex networks. The topological parameters of these networks aptly capture the long-term correlations present in the data. Additionally, we introduce a novel topological parameter, Δσ, to detect nonlinear correlations. The statistical significance of this parameter is rigorously tested using the Monte Carlo method, simulating fractional Brownian motion and fractional Gaussian noise processes with a predefined DFA exponent to estimate confidence intervals. In conclusion, serial correlations are universal in global SAT time series and the presence of these serial correlations should be considered carefully in climate sciences.

Insights from modelling sixteen years of climatic and fumonisin patterns in maize in South Africa

Mycotoxin contamination of agricultural commodities is a global public health problem that has remained elusive to various mitigation approaches, particularly in developing countries. Climate change and its impact exacerbates South Africa's vulnerability to mycotoxin contamination, and significantly threatens its's food systems, public health, and agro-economic development. Herein we analyse sixteen years (2005/2006-2020/2021) of annual national meteorological data on South Africa which reveals both systematic and erratic variability in critical climatic factors known to influence mycotoxin contamination in crops. Within the same study period, data on fumonisin (FB) monitoring show clear climate-dependent trends. The strongest positive warming trend is observed between 2018/2019 and 2019/2020 (0.51 °C/year), and a strong positive correlation is likewise established between FB contamination and temperature (r ranging from 0.6 to 0.9). Four machine learning models, viz support vector machines, eXtreme gradient boosting, random forest, and orthogonal partial least squares, are generalized on the historical data with suitable performance (RMSE as low as 0.00). All the adopted models are able to predict future FB contamination patterns with reasonable precision (R2 ranging from 0.34 to 1.00). The most important model feature for predicting average FB contamination (YA) is the historical pattern of average FB contamination in maize within the region (ΣFBs_avg). The two most significant features in modelling maximum FB contamination (YM) are minimum temperature from the CMIP6 data (Pro_tempMIN) and observed precipitation from the CRU data (O_prep). Our study provides strong evidence of the impact of climate change on FB in South Africa and reiterates the significance of machine learning modelling in predicting mycotoxin contamination in light of changing climatic conditions, which could facilitate early warnings and the adoption of relevant mitigation measures that could help in mycotoxin risk management and control.

Enhancing Radar Echo Extrapolation by ConvLSTM2D for Precipitation Nowcasting

Precipitation nowcasting in real-time is a challenging task that demands accurate and current data from multiple sources. Despite various approaches proposed by researchers to address this challenge, models such as the interaction-based dual attention LSTM (IDA-LSTM) face limitations, particularly in radar echo extrapolation. These limitations include higher computational costs and resource requirements. Moreover, the fixed kernel size across layers in these models restricts their ability to extract global features, focusing more on local representations. To address these issues, this study introduces an enhanced convolutional long short-term 2D (ConvLSTM2D) based architecture for precipitation nowcasting. The proposed approach includes time-distributed layers that enable parallel Conv2D operations on each image input, enabling effective analysis of spatial patterns. Following this, ConvLSTM2D is applied to capture spatiotemporal features, which improves the model's forecasting skills and computational efficacy. The performance evaluation employs a real-world weather dataset benchmarked against established techniques, with metrics including the Heidke skill score (HSS), critical success index (CSI), mean absolute error (MAE), and structural similarity index (SSIM). ConvLSTM2D demonstrates superior performance, achieving an HSS of 0.5493, a CSI of 0.5035, and an SSIM of 0.3847. Notably, a lower MAE of 11.16 further indicates the model's precision in predicting precipitation.

Advances and applications of machine learning and deep learning in environmental ecology and health

Machine learning (ML) and deep learning (DL) possess excellent advantages in data analysis (e.g., feature extraction, clustering, classification, regression, image recognition and prediction) and risk assessment and management in environmental ecology and health (EEH). Considering the rapid growth and increasing complexity of data in EEH, it is of significance to summarize recent advances and applications of ML and DL in EEH. This review summarized the basic processes and fundamental algorithms of the ML and DL modeling, and indicated the urgent needs of ML and DL in EEH. Recent research hotspots such as environmental ecology and restoration, environmental fate of new pollutants, chemical exposures and risks, chemical hazard identification and control were highlighted. Various applications of ML and DL in EEH demonstrate their versatility and technological revolution, and present some challenges. The perspective of ML and DL in EEH were further outlined to promote the innovative analysis and cultivation of the ML-driven research paradigm.

Prediction of monthly precipitation using various artificial models and comparison with mathematical models

Precipitation (PP) prediction is an interesting topic in the meteorology or hydrology field since it is directly related to agriculture, the management of water resources in hydrologic basins, and water scarcity. Selecting the right model to predict precipitation has always been a challenge because it could help researchers to use the proper model for their purposes. Accordingly, the performance of five artificial models (feed-forward neural network, cascade forward neural network, Elman neural network, multi-layer perceptron neural network, and radial basis neural network) and three mathematical models (Poisson regression model (PRM), quadratic model, and multiple linear regression) were evaluated for their ability to predict the monthly precipitation in Mediterranean coastal cities located in Eastern part of Mediterranean Sea for the first time. Twenty-seven Mediterranean coastal cities are considered case studies. For this aim, scenario 1 and scenario 2 with various input variables are proposed. Scenario 1 is developed using the number of months (MN), maximum temperature (Tmax), minimum temperature (Tmin), downward radiation (DR), wind speed (WS), vapor pressure (VP), and actual evapotranspiration (AE). Scenario 2 is developed by adding geographical coordinates (latitude, longitude, and altitude) to the global meteorological data to see the impact of geographical coordinates on the accuracy of the prediction of monthly precipitation. This study utilized the monthly data, which were obtained from TerraClimate for the period from 2010 to 2021. Based on the performance indexes, the PRM model performed best for the prediction of monthly precipitation in all selected locations compared to other models. Moreover, the results indicate that scenario 2 ([Formula: see text]) has shown higher prediction accuracy compared to scenario 1 ([Formula: see text]). In conclusion, PRM with the combination of [[Formula: see text]] had RMSE value that was lower by 12% relative to PRM with the combination of [[Formula: see text]]. Consequently, the PRM model can be recommended for modeling the complexity of interactions for precipitation-climate conditions-geographical coordinates and predicting precipitation.

Automated Street Light Adjustment System on Campus with AI-Assisted Data Analytics

The smart city concept has been popularized in the urbanization of major metropolitan areas through the implementation of intelligent systems and technology to serve the increasing human population. This work developed an automatic light adjustment system at Thammasat University, Rangsit Campus, Thailand, with a primary objective of optimizing energy efficiency, while providing sufficient illumination for the campus. The development consists of two sections: the device control and the prediction model. The device control functionalities were developed with the user interface to enable control of the smart street light devices and the application programming interface (API) to send the light-adjusting command. The prediction model was created using an AI-assisted data analytic platform to obtain the predicted illuminance values so as to, subsequently, suggest light-dimming values according to the current environment. Four machine-learning models were performed on a nine-month environmental dataset to acquire predictions. The result demonstrated that the three-day window size setting with the XGBoost model yielded the best performance, attaining the correlation coefficient value of 0.922, showing a linear relationship between actual and predicted illuminance values using the test dataset. The prediction retrieval API was established and connected to the device control API, which later created an automated system that operated at a 20-min interval. This allowed real-time feedback to automatically adjust the smart street lighting devices through the purpose-designed data analytics features.

Dengue Prediction in Latin America Using Machine Learning and the One Health Perspective: A Literature Review

Dengue fever is a serious and growing public health problem in Latin America and elsewhere, intensified by climate change and human mobility. This paper reviews the approaches to the epidemiological prediction of dengue fever using the One Health perspective, including an analysis of how Machine Learning techniques have been applied to it and focuses on the risk factors for dengue in Latin America to put the broader environmental considerations into a detailed understanding of the small-scale processes as they affect disease incidence. Determining that many factors can act as predictors for dengue outbreaks, a large-scale comparison of different predictors over larger geographic areas than those currently studied is lacking to determine which predictors are the most effective. In addition, it provides insight into techniques of Machine Learning used for future predictive models, as well as general workflow for Machine Learning projects of dengue fever.

Importance of Meteorological Parameters and Airborne Conidia to Predict Risk of Alternaria on a Potato Crop Ambient Using Machine Learning Algorithms

Secondary infections of early blight during potato crop season are conditioned by aerial inoculum. However, although aerobiological studies have focused on understanding the key factors that influence the spore concentration in the air, less work has been carried out to predict when critical concentrations of conidia occur. Therefore, the goals of this study were to understand the key weather variables that affect the hourly and daily conidia dispersal of Alternaria solani and A. alternata in a potato field, and to use these weather factors in different machine learning (ML) algorithms to predict the daily conidia levels. This study showed that conidia per hour in a day is influenced by the weather conditions that characterize the hour, but not the hour of the day. Specifically, the relative humidity and solar radiation were the most relevant weather parameters influencing the conidia concentration in the air and both in a linear model explained 98% of the variation of this concentration per hour. Moreover, the dew point temperature three days before was the weather variable with the strongest effect on conidia per day. An improved prediction of Alternaria conidia level was achieved via ML algorithms when the conidia of previous days is considered in the analysis. Among the ML algorithms applied, the CART model with an accuracy of 86% were the best to predict daily conidia level.

A Comprehensive Study of Artificial Intelligence Applications for Soil Temperature Prediction in Ordinary Climate Conditions and Extremely Hot Events

Soil temperature is a fundamental parameter in water resources and irrigation engineering. A cost-effective model that can accurately forecast soil temperature is urgently needed. Recently, many studies have applied artificial intelligence (AI) at both surface and underground levels for soil temperature predictions. In the present study, attempts are made to deliver a comprehensive and detailed assessment of the performance of a wide range of AI approaches in soil temperature prediction. In this regard, thirteen approaches, from classic regressions to well-established methods of random forest and gradient boosting to more advanced AI techniques, such as multi-layer perceptron and deep learning, are taken into account. Meanwhile, great varieties of land and atmospheric variables are applied as model inputs. A sensitivity analysis was conducted on input climate variables to determine the importance of each variable in predicting soil temperature. This examination reduced the number of input variables from 8 to 7, which decreased the simulation load. Additionally, this showed that air temperature and solar radiation play the most important roles in soil temperature prediction, while precipitation can be neglected in forecast AI models. The comparison of soil temperature predicted by different AI models showed that deep learning demonstrated the best performance with R-squared of 0.980 and NRMSE of 2.237%, followed by multi-layer perceptron with R-squared of 0.980 and NRMSE of 2.266%. In addition, the performance of developed AI models was evaluated in extremely hot events since heat warnings are essential to protect lives and properties. The assessment showed that deep learning and multi-layer perceptron methods still have the best prediction. However, their R-squared decreased to 0.862 and 0.859, and NRMSE increased to 6.519% and 6.601%, respectively.

Predicting Meteorological Variables on Local Level with SARIMA, LSTM and Hybrid Techniques

The choice of holiday destinations is highly depended on climate considerations. Nowadays, since the effects of the climate crisis are being increasingly felt, the need for accurate weather and climate services for hotels is crucial. Such a service could be beneficial for both the future planning of tourists’ activities and destinations and for hotel managers as it could help in decision making about the planning and expansion of the touristic season, due to a prediction of higher temperatures for a longer time span, thus causing increased revenue for companies in the local touristic sector. The aim of this work is to calculate predictions on meteorological variables using statistical techniques as well as artificial intelligence (AI) for a specific area of interest utilising data from an in situ meteorological station, and to produce valuable and reliable localised predictions with the most cost-effective method possible. This investigation will answer the question of the most suitable prediction method for time series data from a single meteorological station that is deployed in a specific location; in our case, in a hotel in the northern area of Crete, Greece. The temporal resolution of the measurements used was 3 h and the forecast horizon considered here was up to 2 days. As prediction techniques, seasonal autoregressive integrated moving average (SARIMA), AI techniques like the long short-term memory (LSTM) neural network and hybrid combinations of the two are used. Multiple meteorological variables are considered as input for the LSTM and hybrid methodologies, like temperature, relative humidity, atmospheric pressure and wind speed, unlike the SARIMA that has a single variable. Variables of interest are divided into those that present seasonality and patterns, such as temperature and humidity, and those that are more stochastic with no known seasonality and patterns, such as wind speed and direction. Two benchmark techniques are used for comparison and quantification of the added predictive ability, namely the climatological forecast and the persistence model, which shows a considerable amount of improvement over the naive prediction methods, especially in the 1-day forecasts. The results indicate that the examined hybrid methodology performs best at temperature and wind speed forecasts, closely followed by the SARIMA, whereas LSTM performs better overall at the humidity forecast, even after the correction of the hybrid to the SARIMA model. Lastly, different hybrid methodologies are discussed and introduced for further improvement of meteorological predictions.

Comparative Analysis of SCMOC and Models Rainstorm Forecasting Performance in Qinling Mountains and Their Surrounding Areas

Taking CMPA (CMA Multi-source Merged Precipitation Analysis System) analysis data as a reference, the research analyzes the forecast performance of ECMWF, CMA-Meso, and SCMOC (National Meteorological Center grid precipitation forecast guidance product) in 74 rainstorm cases in 2020 and 2021 in Qinling Mountains and their surrounding areas by using the dichotomy classical verification score comprehensive diagram and the object-oriented MODE spatial verification method, based on the circulation classification in rainstorm weather. The research conclusions are as follows: (1) based on the high- and low-altitude circulation situation and focused on the direct impact system, rainstorms in the Qinling Mountains and their surrounding areas can be divided into five patterns. (2) Point-to-point verification shows that SCMOC has obvious advantages in rainstorm forecast, but the disadvantage is that the Bias is relatively high. CMA-Meso has advantages in RST (weak weather system) decentralized rainstorm forecast. (3) MODE verification shows that the number of ECMWF and SCMOC independent objects is significantly lower than that of observation, the forecast area of regional rainstorm objects of SCMOC is significantly larger, the SCMOC scattered rainstorm objects are missed, and the number of independent precipitation objects of CMA-Meso is higher than that of the other two precipitation products. (4) The forecast object area and intensity of SCMOC and observation match best in the XFC (westerly trough) circulation situation, while ECMWF has the best results for the forecast of FGXFC (subtropical high westerly trough) rainstorms.