(173ae) Tracking and Predicting Regional Greenhouse Gas Emissions: A Case Study of Connecticut

Global warming triggers wildfires, rising sea levels, and habitat loss for many species. This is driven by elevated greenhouse gases concentrations in the atmosphere. Emissions tracking and prediction frameworks are important to understand the source contributions and find mitigation strategies. Compared with developing a nationwide emissions inventory, however, a regional-level emissions inventory database is crucial for local governments to identify the specific activities and processes responsible for the emissions and implement tailored policies to reduce GHG emissions. This study develops a GHG inventory estimation and projection platform tailored specifically for the local area. This accounting method was applied to three metropolitan statistical areas (MSAs) in Connecticut, with the aim of discovering the emissions characteristics at different locations and forecasting future emissions shift and trends. However, the challenge of developing an inventory at this regional level lies in the availability of open and granular data. To address this issue, we propose a hybrid accounting approach, utilizing a bottom-up estimation based on activity data, which is supplemented by a top-down method by scaling up statewide emissions. We found that emissions in Hartford, Bridgeport, and New Haven MSA totaled 10.95, 7.4, and 6.97 million metric tons of carbon dioxide (MMTCO₂e) in 2021, respectively. Among all the emissions from seven sectors (transportation, electricity consumption, solid waste, stationary combustion, agriculture and wastewater), fossil fuel related activities, such as on-road transportation and residential heating, were the primary source of emissions. In each MSA, the transportation sector contributed to the highest amount of emissions, followed by stationary combustion. Besides, the percentage of electricity consumption in each MSA exceeded that of the entire state, illustrating the higher electricity consumption in urban area, which are heavily centralized in those three MSAs. The per capita emissions were 10.95, 7.41, and 9.6 MTCO₂e in Hartford, Bridgeport, and New Haven, respectively while the average per capita emissions for the entire state was about 9.59 metric tons. Surprisingly, Hartford's per capita emissions exceeded that of the entire state, suggesting the need for specific emissions reduction measures in this area. Additionally, the results show that trees play a significant role in sequestering carbon from the air, with 0.62, 1.98, and 0.74 MMTCO₂e in the three areas, respectively. This illustrates the need to maintain our forestry and promote greenery. After the establishment of the inventory processing and calculation procedure, historical data and emissions factors were utilized to estimate past emissions. We applied the auto-regressive integrated moving average (ARIMA) model to predict the future GHG emissions across all sectors up to 2035. Comprising autoregressive, integrated, and moving average terms, this model forecasts future values based on historical data. The model parameters (p,d,q) were determined through parameter estimation and model testing to minimize the root mean squared error. The resultant prediction outcomes offer valuable insights for policymakers to formulate strategies and set emission reduction targets. We expect that our framework would serve as a valuable tool for regional emissions accounting in support of nationwide net-zero goals.