A study on sustainable aviation fuel demand and carbon emissions forecasting in China based on deep learning LSTM model
YANG Lishan
WANG Honglei
Abstract：To address the critical challenges of uncertain market demand and ambiguous policy planning in China's sustainable aviation fuel(SAF)sector,this study proposes a novel framework that centers on national fuel requirements and policy intensity.By integrating historical data,we develop a fuel consumption forecasting and scenario analysis model based on the Long Short-Term Memory(LSTM)network,augmented with Monte Carlo simulations to quantify future uncertainties.The results show that despite continuous improvements in operational efficiency,growth in aviation traffic will drive a 46%increase in industry fuel consumption between 2025 and 2050,underscoring the limitation of relying solely on technological measures for emission reduction.Under an accelerated low-carbon scenario,SAF demand could reach 31.85 million tonnes(accounting for 50%of total fuel use)by 2050,contributing to a 92%reduction in carbon emissions relative to the business-as-usual case.In contrast,insufficient policy support would lead to a 24-percentage-point shortfall in emission abatement.Monte Carlo simulations further reveal substantial uncertainty in future demand while indicating a tangible potential for high-growth pathways.This study systematically outlines China's medium-to long-term SAF demand trajectory and associated emission reduction potential,offering critical quantitative evidence and decision support for setting science-based blending targets and designing phased industrial policies.
Keywords：sustainable aviation fuel(SAF)demand forecastingcarbon emissions forecastingLSTM model
Publication Date：2026-01-28
Online Publishing Date：2026-03-19(First online date of this platform, not the publication date of the document)
Pages：10( 130-139 )