Machine learning based prediction for China's municipal solid waste under the shared socioeconomic pathways.

Area of Science:

• Environmental Science
• Data Science
• Waste Management

Background:

• Sustainable waste management relies on accurate municipal solid waste (MSW) generation forecasts.
• Big data analysis offers a novel approach for enhanced MSW prediction accuracy.

Purpose of the Study:

• To evaluate supervised machine learning models for MSW forecasting.
• To predict China's MSW generation from 2020-2060 under Shared Socioeconomic Pathways (SSPs).
• To explore socioeconomic drivers of MSW generation.

Main Methods:

• Utilized five supervised machine learning models: linear regression, polynomial regression, support vector machine, random forest, and extreme gradient boosting (XGBoost).
• Applied models to forecast China's MSW generation under five SSPs scenarios.
• Analyzed the relationship between MSW generation and socioeconomic indicators like population and GDP.

Main Results:

• Population and GDP were identified as dominant indicators for MSW prediction.
• The extreme gradient boosting (XGBoost) model demonstrated high effectiveness in MSW forecasting.
• China's MSW generation is projected to reach 464-688 megatons by 2060, a 4-6 fold increase from 2000 levels.
• SSP3, characterized by high population and climate challenges, is the only scenario indicating a potential MSW peak within the study period.
• GDP-driven increases in per capita MSW are the primary factor for rising waste generation.

Conclusions:

• Machine learning, particularly XGBoost, provides a robust method for accurate MSW forecasting.
• Significant MSW increases are anticipated in China, driven by economic growth and population.
• Policy interventions are necessary to mitigate future MSW generation.