Comparison of ozone profiles between Superconducting Submillimeter-Wave Limb-Emission Sounder and worldwide ozonesonde measurements

We compared ozone profiles measured by the Superconducting Submillimeter-Wave Limb-Emission Sounder (SMILES) with those taken at worldwide ozonesonde stations. To assess the quality of the SMILES version 2.3 ozone data for 16-30 km, 601 ozonesonde profiles were compared with the coincident SMILES ozone profiles. The agreement between SMILES and ozonesonde measurements was generally good within 5%-7% for 18-30 km at middle and high latitudes but degraded below 18 km. At low latitudes, however, the SMILES ozone data showed larger values (∼6%-15% for 20-26 km) than those at middle and high latitudes. To explain this bias, we explored some possible issues in the ozonesonde measurement system. One possibility is due to a pressure bias in radiosonde measurements with a pressure sensor, but it would be within a few percent. We also examined an issue of the ozonesonde's response time. The response time was estimated from ozonesonde measurements with ascending and descending profiles showing clear difference, by using the time lag correction method to minimize the difference between them. Our estimation shows 28 s on average which is a similar value derived by prelaunch preparation. By applying this correction to the original profiles, we found a negative bias of the ascending ozonesonde measurement more than 7% at 20 km in the equatorial latitude where the vertical gradient of ozone is steep. The corrected ozonesonde profiles showed better agreement with the SMILES data. We suggest that the response time of ozonesondes could create a negative bias, particularly in the lower stratosphere at equatorial latitudes. Key Points SMILES ozone profiles are compared with worldwide ozonesondes This study presents the quality of the SMILES ozone product Suggest a possible bias of ozonesonde measurements