Polar regions are data sparse regions. Research ships operating in polar regions often record sea-ice conditions during their transects through ice infested waters. Such observations of the sea-ice conditions are often the only information that can be provided in addition to satellite-based estimates of the sea-ice conditions, such as sea-ice concentration or sea-ice thickness.

Such observations have been carried out and gathered using two protocols. For the Antarctic, this is the so-called ASPeCt protocol [Worby and Allison, 1999; Worby and Dirita, 1999; Worby et al., 2008]. For the Arctic, this is the so-called ASSIST/IceWatch protocol [Hutchings et al., 2018]. The latter builds on the ASPeCt protocol, incorporating surface melt conditions being more ubiquitous in the Arctic.

Ship-based observations of the sea-ice conditions are conducted manually, visually, i.e. by eye, regularly every hour taking into account an area around the ship of about one kilometer radius. Note that this area distorts to an elliptically shaped area as a function of observers' experience, ships' cruising speed and ice and visibility conditions.

Each observation comprises the total sea-ice concentration, and the concentration, level ice thickness, level ice snow depth, fraction and height of ridges, ice type, snow type, and floe size for the up to three thickest ice types. For the Arctic, melt-pond fraction and stage-of-melt are also part of the observables. In addition to the ships' position often auxiliary parameters such as visibility, wind speed and direction, or air and water temperature are recorded.

For development and evaluation of satellite-based sea-ice products, such ship-based observations are of great value. Because of this, within the ESA-CCI sea-ice ECV project (ESA-SICCI), phase 2, a standardized data set of such ship-based observations was generated for both polar regions. It comprises data from June 2002 through December 2015. This time period is motivated by the purpose to evaluate sea-ice concentration data retrieved from AMSR-E and AMSR2 brightness temperature measurements which, at the time the project was initiated, were planned to be retrieved until the end of 2015.

The data set incorporates observational data from various collections, e.g. a part of the original ASPeCt collection [Worby et al., 2008], which ended in May 2005. More information about all data sources is given below.

All data have been manually standardized to the same format (i.e., number of decimals, unit), using the same value to describe missing data, using the same temporal ordering, and filling gaps with the respective missing-data value. Double data entries have been removed. The data set is split into two ascii text files, one for the Arctic, one for the Antarctic. It has been successfully used to evaluate sea-ice concentration and thickness products of the ESA-SICCI phase 2 project.