Paper Title
Electric Power Sector Co2 Estimations and Emissions-Mitigation Pathways

Abstract
Carbon dioxide (CO2) emissions from the electric power sector has contributed massively towards the hefty weight of United States (US) in global energy-related CO2. The Energy Information Administration (EIA) uses the National Energy Modelling System to forecast and make long-term projections on electric sector CO2 emissions in the US but the forecast inaccuracies of past projections are considerably high due to unrealized assumptions and scenarios on numerous peripheral variables, a uniform time effect used, and volatilities in patterns of activity data. Here, we propose and apply a volatility-consistent recurrent neural network technique devoid of exogenous variables and assumptions; that allows for varying time effects to forecast electric power sector CO2 emissions in the US for the 2005-2015 period. Based on the high-accuracy forecasts, we propose optimal emissions-mitigation pathways to achieve zero CO2-emissions target in the electric sector. The empirical results suggest the proposed model presents overwhelming improvement up to ~74-fold relative projections in selected EIA's Annual Energy Outlooks (AEOs). Projections from utilizing the proposed high-accuracy technique shows CO2 emissions will decrease steadily till 2034-2035 which straddles findings from AEO2018. Results from the optimal mitigation path also demonstrates that intensifying current policies are enough to completely curtail energy-related CO2 emissions from the sector by 2023. Keywords - Emissions-Mitigation Pathways, Electric Power Sector, Carbon Dioxide Emissions, Forecasting, LSTM RNN.