Review on the Application of Remote Sensing Data and Machine Learning to the Estimation of Anthropogenic Heat Emissions

Abstract

Anthropogenic heat is the heat generated by human activities such as industry, construction, transport, and metabolism. Accurate estimates of anthropogenic heat are essential for studying the impacts of human activities on the climate and atmospheric environment. Commonly applied methods for estimating anthropogenic heat include the inventory method, the energy balance equation method, and the building model simulation method. In recent years, the rapid development of computer technology and the availability of massive data have made machine learning a powerful tool for estimating anthropogenic heat fluxes and assessing its effects. Multi-source remote sensing data have also been widely used to obtain more details of the spatial and temporal distribution characteristics of anthropogenic heat. This paper reviews the main approaches for estimating anthropogenic heat emissions. The typical algorithms of the abovementioned three methods are introduced, and their advantages and limitations are also evaluated. Moreover, the recent progress in the application of remote sensing data and machine learning are discussed as well. Based on big data and machine learning techniques, the research on feature engineering and model fusion will bring about major changes in data analysis and modeling of anthropogenic heat. More in-depth research of this issue is recommended to provide important support for curbing global warming, mitigating air pollution, and achieving the national goals of carbon peak and a carbon neutrality strategy.

Keywords

• anthropogenic heat
• the inventory method
• the energy balance equation method
• the building model simulation method
• remote sensing data
• machine learning