Skip navigation.

Round Robin

Printer-friendly versionSend to friend

Introduction to the Ozone_cci Round Robin

Interim and Round Robin evaluation activities consist of dedicated subsets of the full ECV evaluation, carried out on subsets of the available satellite and correlative data records. Activities and reporting shall comply with the Product Validation Plan. Care is given to produce accurate documentation of the algorithm, of the data, of the evaluation process and of the evaluation results. Selection criteria in the case of Round Robin exercises, and quality indicators in the case of interim evaluations, are based on the availability and appropriateness of the algorithm, data and documentation (including metadata):

Retrieval algorithm

Are the retrieval concept and implementation accepted by the community, verified through algorithm inter-comparison exercises, sanctioned by peer-reviewed documents, and up-to-date? Is the algorithm description available (e.g. through ATBD) and complete, with a sufficient level of detail to judge its validity and its potential capabilities in the CCI context? Are there particular features revealed by algorithm verifications and inter-comparisons, like e.g. negative values, biases, vertical oscillations, incoherent geographical structures?

Level-2 data

Is the data available? In parts only or the complete data record? Publicly or with restricted/controlled access? Is the data description sound and complete (e.g. PSD)? Is there a need for data extraction and reading routines and are they available? Is the processing chain well identified: name and reference of the processors, location, version of level-1 data, and version of level-1-to-2 processing? Is the data accompanied by relevant error bars/uncertainty estimates – calculated with community agreed methods – and associated metadata, including diagnostic quantities for every measurement: e.g., error covariance and averaging kernel matrices, a priori profile, and parameters required for data filtering like e.g. star brightness, occultation obliquity and solar zenith angle?

Quality assurance

Has the data been validated? If yes, how and how far? With respect to which correlative data of reference, traceable to international standards? Against which user requirements? Is the validation process well described, accompanied by relevant validation metadata, and compliant with community agreed practices, especially regarding data manipulations (e.g. re-gridding and smoothing operations)? Are the validation results well documented and leading to clear quality indicators? Is the terminology unambiguous and compliant with recommended standards? Apart from comparison results, are there particular features affecting the quality of the data, like e.g. negative values, biases, vertical oscillations, incoherent geographical structures? Does the published use of the data demonstrate its fitness for purpose in the context of the Ozone_cci project?

Round Robin evaluation of ozone profile data products from nadir-viewing instruments

Two scientific algorithms developed at KNMI (OPERA) and at RAL, and applicable to all sensors, are intercompared and characterised in a Round Robin exercise. Results of this evaluation supports the determination of the new Ozone-cci nadir profile algorithm baseline, which will likely be a combined RAL/OPERA algorithm.

Round Robin evaluation of MIPAS limb ozone profile data

MIPAS level-2 data are processed by ESA/DLR plus by several scientific institutions, namely: IFAC Florence, IMK/IAA, ISAC Bologna, Leicester University, and Oxford University. They participate in a Round Robin exercise, which supports the selection of the ‘best’ processor for generating the ECV limb ozone profile data from MIPAS.


For more details, please consult the Product Validation and Selection Report.