20 Aug 2025

Parameterized Statistical Distributions of Unique Origin-Destination Pairs for Major Waterborne Commodity Groups

Abstract: Modeling the spatiotemporal aspects of freight movements within a distributed network is crucial to forecasting transportation infrastructure needs, prioritizing investments, and estimating emissions. Commodity flow patterns and trends along the inland waterway transportation system are significant because of their importance for the economy, in line with priorities of the US Committee on the Marine Transportation System. Analyzing these inland waterway flows better informs multimodal freight transportation modeling. This exploratory research uncovers, describes, and summarizes patterns and trends of the US waterway transportation system by mining waterborne freight data. The purpose of this work is to identify parameterized statistical distributions that describe the relative dispersion of unique waterborne Origin-Destination (OD) pairs when sorted high to low by annual freight tonnage. Best-fit statistical distributions and associated parameters are identified for the leading commodities transported on waterways, and an 11-year time-series analysis of commodity-specific distribution parameters provide their evolution across time. Results show that the power law best explains the distribution of ranked ODs by tonnage for seven of the twelve commodities analyzed. The root-mean-square error (RMSE) of any given commodity modeled is less than 1%. These results provide insights into the underlying behavior of inland waterway freight transportation.

6 Nov 2024

Time-Series Forecasting Methods: A Review

Abstract: Time-series forecasting techniques are of fundamental importance for predicting future values by analyzing past trends. The techniques assume that future trends will be similar to historical trends. Forecasting involves using models fit on historical data to predict future values. Time-series models have wide-ranging applications, from weather forecasting to sales forecasting, and are among the most effective methods of forecasting, especially when making decisions that involve uncertainty about the future. To evaluate forecast accuracy and to compare among models fitted to a time series, three performance measures were used in this study: mean absolute error (MAE), mean square error (MSE), and root-mean-square error (RMSE).