Sea Ice Products derived from the NIMBUS 5 Nimbus-E Microwave Spectrometer (NEMS) data from 1972-1973

The Nimbus-E Microwave Spectrometer (NEMS) aboard the Nimbus-5 satellite was a five-channel microwave spectrometer with two water-vapour channels (22.235 and 31.4 GHz) and three oxygen channels (53.65, 54.9, and 58.8 GHz). The instrument observed at nadir with a footprint approximately 180 km in diameter at the Earth’s surface. The NEMSN5L2 (Nimbus-5 NEMS Level-2) product is collocated with ERA5 atmospheric variables, including total column water vapour, cloud liquid water, and wind speed, which are subsequently used for atmospheric correction of the NEMSN5L2 dataset. Data are available for several discrete periods between 17 December 1972 and 31 October 1973: 17 December 1972 to 7 March 1973, 15 April to 1 May 1973, 17 June to 30 June 1973, 12 August to 31 August 1973, and 14 October to 31 October 1973. In total, 150 days with sufficient global coverage, defined as more than 100 observations per hemisphere per day, are available. One isolated day in May was excluded because of insufficient coverage. Of the five available channels, only the 22.235 GHz and 31.4 GHz brightness temperatures are used here.The experimental Nimbus satellite programme was highly successful and provided important heritage for modern satellite missions. Although NEMS predates contemporary multifrequency microwave radiometers, elements of modern processing methodologies can be applied to these data to derive geophysical products such as sea ice extent. In particular, the 22.235 GHz and 31.4 GHz channels are sensitive to atmospheric water vapour and liquid water, respectively, enabling atmospheric correction of brightness temperatures and consistent noise reduction over both sea ice and open water. The same processing methodology is applied for the retrieval of sea ice concentration (SIC), following the approach described by Kolbe et al. (2025) for the NIMBUS-6 SCAMS instrument. The NEMS dataset therefore extends the microwave satellite climate data record, including sea ice concentration and derived extent, into an important and otherwise sparsely observed period in the early 1970s.The dataset includes the filtered sea ice concentration and the raw, unfiltered estimates, alongside three uncertainty fields: total, smearing, and algorithm standard errors. The smearing uncertainty reflects spatial averaging caused by the coarse sensor footprint, while the algorithm uncertainty arises from natural variability in the brightness temperature tie-points. Quality information is provided through a status flag. Brightness temperatures are included in both uncorrected and atmospherically corrected form, the latter derived using collocated ERA5 total column water vapour, cloud liquid water, and wind speed. Geolocation is provided as latitude/longitude coordinates and projected coordinates on a Lambert Azimuthal Equal-Area grid, with timestamps expressed as days since 1 January 1900.