刘文清：人工智能助力环境监测从“感知”到“认知”，碳计量仍是关键瓶颈

Core Viewpoint - The current application of artificial intelligence (AI) in environmental monitoring is still in the "small scene" stage, and expanding its use requires promoting data openness and sharing [1] Group 1: AI Development Stages - AI development is categorized into three stages: computational intelligence, perceptual intelligence, and cognitive intelligence, with the environmental monitoring field currently in the initial stage of "perceptual intelligence" [1] - The complexity of atmospheric components, including pollutants and greenhouse gases, necessitates more than just ground monitoring stations to accurately reflect air quality [1] Group 2: Data Integration and Monitoring - AI can bridge data gaps by integrating satellite remote sensing data, ground monitoring data, and other multi-source information to dynamically display regional air quality changes [1] - The need for technological upgrades in environmental monitoring is emphasized, comparing it to advancements in medical imaging technologies [2] Group 3: Carbon Monitoring Challenges - Carbon measurement is crucial, with projections indicating that China's carbon emissions will be around 11 billion tons by 2030 and need to be reduced to 1 billion tons by 2060, requiring significant technological intervention [2] - Current carbon accounting methods are inadequate, relying on estimations from coal and electricity consumption, which do not meet precision requirements [2] Group 4: Monitoring Precision and Data Sharing - Monitoring precision is a significant challenge, as atmospheric CO2 concentration changes are minimal, necessitating high-resolution instruments to capture these variations [3] - Two key initiatives proposed to enhance AI application in environmental monitoring include breaking down data barriers for cross-department sharing and continuously upgrading monitoring technologies [3] - The future of AI in environmental monitoring is promising, with potential advancements in carbon measurement accuracy, data openness, and technological innovation [3]