Observed Sentimental Alteration in the Public Water Pollution Complaints during Climatic Extremes and the COVID-19 Pandemic

Population growth and economic development increase water demand, while human activities degrade the quality of available water resources along the adjacent rivers. The U.S. state of Alabama has been suffering from floods causing the degraded water quality by scouring pollutants into the water. In recent decades, Alabama has been experiencing persistent precipitation deficits and unusual severe droughts, resulting in limited economic and water-based recreation activities within downstream states. Since 2020, The COVID-19 pandemic aroused a series policies like quarantine and lock down, which slowed down the economic development and reduced chances of people going outside to witness the water pollution accidents.

In this study, we conducted a sentiment analysis of over 9,900 water pollution complaints (2012-2020) from residents in Alabama. Overall, it is found that complaints are dominated by negative and objective complaints no matter what extremes events or environmental accidents. Results show that sentiment alteration during climate extremes and COVID period was detected. Potential causes of the sentimental alteration in the public water pollution complaint reports were explored. Results show more complaints during summer seasons, which can be explained as higher temperature and intensive precipitation at that time. More complaints are distributed in the counties that are higher socioeconomically developed, to be more specific, counties with more population and higher GDP level. The severity of antecedent extreme events can affect the sentiment of environmental pollution complaints related to on-going extreme events due to limited human judgements. Key words extracted from the complaints point out the pollution resources and locations, which provide important clues from local government to resolved problems.

This study provides an example of how unstructured data such as public complaints can be used as a technology to improve the water pollution and public health monitoring with the help of big data and artificial intelligent technologies. While the results of this study were based water pollution complaints from residents of Alabama state, it is applicable to other environmental pollutions (like air and land) and other regions with available long-term textual data.