Applying Satellite Data in Evaluating Accuracy of Self-Report Environmental Releases Policies: Evidence from the United States Toxic Release Inventory

Information disclosure programs, by design, could motivate plants to behave better under the pressure of transparent environmental performance. However, the effectiveness of such programs is also severely questioned due to potential inaccuracy, under-reporting, and lack of monitoring of self-reporting data. As such, empirical examination of the accuracy of self-reporting is crucial. It has yet been limited by the availability of the monitoring network nearby, and further obscured by the complicated source-receptor relationship. We instead use satellite observations with high spatial-temporal coverage to examine the accuracy of the U.S.s self-reporting Toxic Release Inventory (TRI). We find general accordance between satellite retrievals and self-reported total onsite releases of ammonia, the largest source of air releases in TRI. Surrounding the point sources of industrial plants, we observe a significant linearly decreasing effect across the top 50, 100, 200, 400, and 500 largest sources in the US from 2007 to 2018. This suggests that even without close monitoring and validation, firms self-reported releases are likely to reflect their environmental performance. In addition, we estimate the likelihood of individual plants misreporting releases for a specific year. Our work proposes the possibility of using satellite observations to further improve transparency of firms environmental performance, and thus strengthening the effectiveness of environmental information disclosure programs.