Reanalysis of the instrumental climate record by means of data assimilation involves quality control, bias correction, and specification of error statistics for the input observations. Conversely, reanalysis provides valuable feedback information about the quality of the data that can help quantify and reduce the uncertainties in the instrumental record. ERA-CLIM contributions in this area include:

- Development of methods and techniques for removing time-dependent biases in observations during data assimilation, using prior information about likely breakpoints in the data series;
- Improvement and homogenisation of the upper-air record from early radiosonde observations;
- Improvement of sub-surface and near-surface ocean data collections;
- Use of ensemble techniques to represent uncertainties in homogenised data record.