Abstract

To support international climate treaties, there is a pressing need to develop ‘top-down’ verification of trace gas emissions on global and national scales. Emissions estimates can be made using atmospheric measurements, provided that appropriate monitoring networks and computer models exist. Numerous models have been developed at institutions across the world to simulate pollutant transport, many of which use ‘analyzed’ meteorological fields from weather forecasting centres. These transport models are combined with atmospheric measurements using statistical methods such as Bayesian optimization to refine or provide independent estimates of inventory emissions. Relatively simple models can reliably determine global emissions of greenhouse gases and ozone depleting species. These estimates are often found to be significantly different to those calculated from the bottom-up. In recent years, increasingly sophisticated modeling tools and denser monitoring networks have allowed us to infer emissions on national scales in certain regions, with promising results and often large discrepancies with inventory estimates. However, much uncertainty remains on these fine scales, and improvements to monitoring networks, bottom-up inventories, transport models and techniques for uncertainty quantification will be required to make top-down methods suitable for national emissions verification.