Forecasting China's carbon emission intensity and total carbon emissions based on the WOA-Stacking integrated model

China ranks first globally in carbon emissions. Accurate carbon emissions forecasting is crucial for devising effective mitigation strategies and has thus become a focal point of academic research. This study explores carbon emission drivers across the economy, demography, industry, and energy domains. We empirically analyze China's 1990-2021 emissions data to forecast levels for 2022-2025. Introducing a novel stacking integrated learning model refined with a whale optimization algorithm (WOA), this research employs World Bank data, China statistics, and BP data to verify the model's validity. Findings reveal that the proposed WOA-Stacking integrated model significantly outperforms in forecasting carbon emissions. At current rates, China's carbon emission intensity in 2025 is predicted to be 3.88% lower than 2020, likely missing its 14th Five-Year Plan target. Additionally, the key factors affecting carbon emission intensity are total electricity consumption, per capita energy consumption, trade openness, urbanization rate, and the previous period's carbon emission intensity data. The GDP, fixed asset investment, total energy consumption, urban population, and total population drive total carbon emissions. Therefore, China must improve policies targeting these factors to accelerate future emission mitigation.