Sensitivity analysis of the potential impact of discrepancies in stratosphere–troposphere exchange on inferred sources and sinks of CO ₂

Abstract. The upper troposphere and lower stratosphere (UTLS) represents a transition region between the more dynamically active troposphere and more stably stratified stratosphere. The region is characterized by strong gradients in the distribution of long-lived tracers, which are sensitive to discrepancies in transport in models. We evaluate the GEOS-Chem model in the UTLS using carbon dioxide (CO2) and ozone (O3) observations from the HIAPER (The High-Performance Instrumented Airborne Platform for Environmental Research) Pole-to-Pole Observations (HIPPO) campaign in March 2010. GEOS-Chem CO2 / O3 correlation suggests that there is a discrepancy in mixing across the tropopause in the model, which results in an overestimate of CO2 and an underestimate of O3 in the Arctic lower stratosphere. We assimilate stratospheric O3 data from OSIRIS and used the assimilated O3 fields together with the HIPPO CO2 / O3 correlations to obtain a correction to the modeled CO2 profile in the Arctic UTLS (primarily between the 320 and 360 K isentropic surfaces). The HIPPO-derived correction corresponds to a sink of 0.13 Pg C month−1 in the Arctic. Imposing this sink during March–August 2010 results in a reduction in the CO2 sinks inferred from GOSAT observations for temperate North America, Europe, and tropical Asia of 20, 12, and 50%, respectively. Conversely, the inversion increased the source of CO2 from tropical South America by 20%. We found that the model also underestimated CO2 in the upper tropical and subtropical troposphere, which may be linked by mixing across the subtropical tropopause. Correcting for the bias relative to HIPPO in the tropical upper troposphere, by imposing a source of 0.33 Pg C, led to a reduction in the source from tropical South America by 44%, and produced a flux estimate for tropical Asia that was in agreement with the standard inversion (without the imposed source and sink). However, the seasonal transition from a source to a sink of CO2 for tropical Asia was shifted from April to June. It is unclear whether the discrepancies found in the UTLS are due to errors in mixing associated with the large-scale dynamics or are due to the numerical errors in the advection scheme. However, our results illustrate that discrepancies in the CO2 distribution in the UTLS can affect CO2 flux inversions and suggest the need for more careful evaluation of model transport errors in the UTLS.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.