TOLNet validation of satellite ozone profiles in the troposphere: impact of retrieval wavelengths
Bibliographic record
Abstract
The Tropospheric Ozone Lidar Network (TOLNet) was used to validate retrievals of ozone (O 3 ) profiles in the troposphere from the TROPOspheric Monitoring Instrument (TROPOMI) ultraviolet (UV), Cross-track Infrared Sounder (CrIS) infrared (IR), and a combined UV + IR wavelength retrieval from TROPOMI/CrIS. Observations from six separate ground-based lidar systems and various locations of ozonesondes distributed throughout North America and in the Netherlands were used to quantify systematic bias and random errors for each satellite retrieval. Furthermore, TOLNet data were used to intercompare idealized UV, IR, and UV + IR convolved lidar profiles of O 3 in the troposphere during case studies representative of high-O 3 events. This study shows that the improved sensitivity and vertical resolution in UV + IR retrievals in the middle- and upper-troposphere resulted in tropospheric degree of freedom (DOF) values ∼ 33 % higher compared to UV- and IR-only retrievals. The increased DOFs in the UV + IR retrievals allowed for improved reproduction of mid- and upper-tropospheric O 3 enhancements and, to a lesser degree, near-surface pollution enhancements compared to single-wavelength satellite products. The validation of O 3 profiles in the troposphere retrieved with the UV-only, IR-only, and UV + IR Tikhonov regularised Ozone Profile retrievAl with SCIATRAN (TOPAS) algorithm developed at the Institute for Environmental Physics, University of Bremen, demonstrated the utility of using TOLNet as a satellite evaluation data set. TOPAS UV-only, IR-only, and UV + IR wavelength retrievals had systematic biases, quantified with normalized mean bias, throughout the troposphere of 11.2 ppb (22.1 %), −1.7 ppb (−0.3 %), and 3.5 ppb (7.8 %), respectively, which meet the tropospheric systematic bias requirements defined by the science teams for the TROPOMI and CrIS sensors. The primary drivers of systematic bias were determined to be solar zenith angle, surface albedo, and cloud fraction. Random errors, representative of uncertainty in the retrievals and quantified by root mean squared errors (RMSEs), were large for all three retrievals, with UV-only, IR-only, and UV + IR wavelength retrievals having RMSEs throughout the troposphere of 17.4 ppb (19.8 % of mean tropospheric column values), 10.5 ppb (12.6 % of mean tropospheric column values), and 14.0 ppb (14.6 % of mean tropospheric column values), respectively. TOPAS UV-only profiles did not meet the uncertainty requirements defined for TROPOMI for the troposphere; however, CrIS IR-only retrievals did meet the uncertainty requirements defined by this mission. The larger random errors reflect the challenge of retrieving daily O 3 profiles due to the limited sensitivity and vertical resolution of these retrievals in the troposphere. Tropospheric systematic biases and random error were lower in IR-only and combined UV + IR retrievals compared to UV-only products due to the increased sensitivity in the troposphere allowing the retrievals to deviate further from the a priori profiles. Observations from TOLNet demonstrated that the performance of the three satellite products varied by season and altitude in the troposphere. TOLNet was shown to result in similar validation statistics compared to ozonesonde data, which are a commonly used satellite evaluation data source, demonstrating that TOLNet is a sufficient source of satellite O 3 profile validation data in the troposphere, which is critical as this data source is the primary product identified for the tropospheric O 3 validation of the recently launched Tropospheric Emissions: Monitoring of Pollution (TEMPO) mission.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".