Quality aspects of Fengyun3 D∕E radio occultation bending angle products

<p>This study systematically evaluated the quality of ionosphere-corrected bending angles from Fengyun3 (FY3) <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M3" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">D</mi><mo>/</mo><mi mathvariant="normal">E</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="23pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="8f2aa91624c9c59738e9e168ab021603"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="amt-19-659-2026-ie00004.svg" width="23pt" height="14pt" src="amt-19-659-2026-ie00004.png"/></svg:svg></span></span> satellites (equipped with GPS and BDS receivers) using ERA5 data as references and MetOp products as comparisons. The quality of subsequent retrieved optimized bending angles, refractivity, and temperature were also analysed. Ionosphere-corrected bending angle were assessed via two approaches: outlier detection across 10–80 km and <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M4" display="inline" overflow="scroll" dspmath="mathml"><mrow><mtext>bias</mtext><mo>/</mo><mtext>noise</mtext></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="54pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="446f805ac264e5f3015fd50a785c7847"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="amt-19-659-2026-ie00005.svg" width="54pt" height="14pt" src="amt-19-659-2026-ie00005.png"/></svg:svg></span></span> quantification. Overall quality evaluation showed that FY3 ionosphere-corrected bending angles were consistent with MetOp below 40 km. Above 40 km, FY3 bending angles exhibited larger errors than MetOp. In outlier detection, MetOp had less than 5 % bad profiles, followed by FY3D (<span class="inline-formula">&lt;10 <i>%</i></span>), while FY3E (GPS/BDS) had about 20 % bad profiles. FY3E-GPS bending angles are prone to have large outliers in the height range of 35–50 km. For bias and noise quantification, the daily mean biases and noise levels of FY3 satellites were higher than those of MetOp. Specifically, FY3E-GPS showed notable large daily mean biases of about <span class="inline-formula">−0.4 µrad</span> and most of these biases are in setting RO events. FY3D and FY3E-BDS ranked second, with biases of approximately <span class="inline-formula">−0.1 µrad</span>. MetOp had the smallest biases, at around <span class="inline-formula">−0.05 µrad</span>. Regarding noises, FY3D, FY3E-GPS exhibited comparable noise levels, at roughly 2.5 <span class="inline-formula">µrad</span>; FY3E-BDS had lower noises of 1.5 <span class="inline-formula">µrad</span>. MetOp noises are smallest at about 1.0 <span class="inline-formula">µrad</span>. Due to the larger biases and noises at high altitudes, FY3's optimized bending angles were strongly corrected by background bending angles. Refractivity and temperature were also influenced by the strong correction of optimized bending angle. In summary, FY3 ionosphere-bending angles show high quality below 40 km. However, at high altitudes, further efforts are required for improving FY3 data's utility in numerical weather prediction and climate studies, especially for stratospheric applications.</p>