空气质量预测模拟技术演变与发展研究

近年来我国的环境问题日益引人注目,大气污染问题显得尤为突出。空气质量预测技术方法能够提前预测区域空间内的大气污染物浓度,其发展十分迅速。本文介绍了一些国内外常用的空气质量预测技术方法的原理及案例,对其结果与局限性进行归纳总结,并对其未来发展提出展望。空气质量预测技术方法分为两类:数值模拟和统计学习,数值模拟通常可以分为第一代、第二代和第三代空气质量模型,统计学习可以分为简单经验统计和机器学习。两类方法的目的都是尽可能真实、准确地实现特定时间、特定区域范围内大气污染物浓度的预测,但两类方法的原理算法差异较大。该研究系统梳理了空气质量预测技术方法演变的历程和发展的现状,展望了其发展趋势,分析了建立、健全空气质量预测技术标准规范体系的重要性,提出了相关参考建议。

Evolution and Development of Air Quality Prediction and Simulation Technology

In recent years, China''s environmental problems have become increasingly noticeable, and the problem of air pollution has become particularly prominent. The air quality prediction method can predict the concentration of pollutants in a certain area in advance, and its development is very rapid. This article introduces the principles and cases of some commonly used air quality prediction techniques, summarizes their results and limitations, and proposes prospects for future development. Air quality prediction methods are divided into two categories: numerical simulation and statistical learning. Numerical models can usually be divided into first, second and third generations. Statistical learning can be divided into simple empirical statistical methods and machine learning. The purpose of the two types of methods is to achieve the prediction of air pollutants in a specific time and within a specific area as truthfully and accurately as possible, but the principles and algorithms of the two types of methods are quite different. This research systematically sorts out the evolution progress and development status of air quality prediction technology, looks forward to the development trend of air quality prediction technology, analyzes the importance of establishing and perfecting air pollutant concentration prediction standards and regulations, and gives relevant reference suggestions.