Weather Temperature Time Series Analysis and Forecasting in the Mississippi Delta

Weather temperature is an important factor that impacts agricultural productivity. Accurate networked weather temperature data analysis and forecasting is crucial to characterization of weather temperature dynamic patterns for geospatial analysis for agricultural production management. This study focuses on evaluation and development of times series analysis and forecasting of weather temperature data collected in the Mississippi Delta and surrounding areas. The methods include deterministic and stochastic time series analysis for linear and nonlinear modeling, particularly the algorithms of Box-Jenkins and Autoregressive Conditional Heteroskedasticity. Artificial intelligence can significantly advance time series analysis and forecasting. So, this study further assesses machine learning and deep learning. The distributed optimal model and model ensemble developing from the methods of this study will be able to implement networked weather temperature time series data analysis and forecasting in the Mississippi Delta for characterization of weather dynamic patterns for geospatial agricultural productivity studies in the region. The methods from this study can be adopted to time series analysis and forecasting of other weather variables, such as precipitation.