G2.26   Uncertainty in O3 cross sections used in the spectral fit for DOAS, MAX-DOAS and Pandora data analysis

Gap detailed description

The uncertainty in the O3 absorption cross-sections is one of the main systematic error sources in the remote sensing of atmospheric O3 using UV-visible spectroscopy techniques. Even though the uncertainty can be considered as a systematic error source, the actual error depends on atmospheric temperature, and thus it can be considered as a pseudo-random error, as mentioned in the deliverable D4.3 ‘Uncertainty Budget’ of the EC FP7 project NORS (see http://nors.aeronomie.be/projectdir/PDF/NORS_D4.3_UB.pdf). Presently the uncertainty in the total O3 due to uncertainty in absorption cross sections is assumed to be around one to a few percent (WMO GAW report 218, NORS_D4.3_UB.pdf). In general, when the uncertainties related to O3 cross-sections and their temperature dependencies are well characterized, this effect can be included in the error budget of O3 observations.

The recent WMO IGACO-O3/UV activity ACSO (Absorption Cross Sections of O3 , http://igaco-o3.fmi.fi/ACSO/), performed a thorough evaluation of the existing cross sections and their impact on ground based and satellite O3 retrievals. In particular cross sections studied were Bass and Paur (1985), Brion, Daumont Malicet (1995) and Serdyuchenko et al. (2014). The outcome of the ACSO study was that the latest Serdyuchenko et al. cross sections are recommended to be used for ground-based Brewer and Dobson instruments. However, these cross sections were not recommended to be used for satellite retrievals due to deficiency in the signal-to-noise ratio close to 300nm. From the perspective of satellite validation, it would be beneficial if the same cross-sections were used by both satellites and ground-based instruments. However, if different absorption cross sections are used in the satellite validation, it is important to understand what type of differences they cause in the validation. Related to GAIA-CLIM, it is to be noted that neither Pandora nor any other DOAS or MAX-DOAS instruments were included in the ACSO study.

The uncertainties in the O3 absorption cross sections are partially addressed in the GAIA-CLIM project. 

Activities within GAIA-CLIM related to this gap

A literature study leading to a summary of the findings including a recommendation of how this should be applied with regard to DOAS, MAX-DOAS and Pandora instruments.

Gap remedy(s)

Remedy #1

Specific remedy proposed

It would certainly be beneficial to study what impact the differences in the O3 cross sections recommended for Dobson and Brewer instruments and the ones used for satellite retrievals have on the retrieved O3 amount when applied within the DOAS data analysis.  This will be predominently a literature study but will also include consultation with the Brewer and Dobson community.

Measurable outcome of success

If the difference in the end product (total O3 column) is quantifiable with regard to which of the different O3 cross-sections have been used within the retrieval, then this can be applied to better compare the O3 data measured by satellites with ground-based data sets while both satellite and ground-based observations still use their preferred O3 cross-sections for the data analysis.

Achievable outcomes

Technological / organizational viability: medium.

Indicative cost estimate: medium (>1million)/ low (<1 million).  While the cost estimate for a basic sensitivity study would be low, the cost estimate for applying more sophisticated measures such as funding new lab measurements would see the cost estimate rise to medium.


The study suggested here will help to understand the uncertainties caused by different sets of O3 cross-sections used within the data analysis and how this impacts on the overall measurement uncertainty, and therefore directly addresses this gap.


2 years.

Gap risks to non-resolution

Identified future risk / impact

Probability of occurrence if gap not remedied

Downstream impacts on ability to deliver high quality services to science / industry / society

Higher uncertainty and/or bias in O3 data sets due to differences in the O3 cross sections used in the analysis


Less reliable comparisons between O3 satellite and ground-based DOAS/MAX-DOAS/Pandora data sets.


Work package: