Short summary and metadata file of SHEBA EM measurements. Indirect measurements of ice thickness were completed with a Geonics EM31 electromagnetic induction (EM) device. The instrument determines the apparent conductivity of the underlying medium based on measurements of the secondary electromagnetic field induced by a transmitter coil in the ice and the seawater underneath. Owing to low summertime sea ice conductivities (<50 mS/m) and high seawater conductivities (>2500 mS/m) the signal is controlled by eddy currents generated at the ice-seawater interface. The two coplanar transmitter and receiver antenna coils are mounted at a spacing of 3.66 m and operate at 9.8 kHz. The entire instrument has been mounted in a polyethylene kayak hull to allow for towing by hand and snowmobile across the ice surface and melt ponds. To assess the amount and rate of ice ablation, EM data have been obtained along 5 main profiles (between 60 and 900 m in length) at 5 m spacing and at several additional sites. Apart from a set of measurements completed before the onset of the melt season and after the start of fall freeze-up, repeated sampling along each of the profiles at intervals ranging between one and three weeks provide data on the progression of melt throughout the summer. Ice thicknesses have been derived from apparent conductivity measurements through inversion of an empirical ice thickness-conductivity relationship based on a large data set collected over Arctic first- and multi-year summer sea ice (Haas et al., 1997, Haas, unpublished). For further validation, concurrent ice thickness and EM measurements have been obtained at 74 ablation gauge sites. Furthermore, based on ice-core and upper ocean conductivity measurements a one-dimensional, two-layer conductivity model has been employed to assess sources of error and reliability of the data set. The accuracy of the EM31 instrument is specified at better than 1 mS/m. Based on the first derivative of the empirical thickness-conductivity relations, the sensitivity of the method has been derived as 0.015 m for 3 m thick level ice and 0.09 m for 5 m thick level ice. Uncertainties in the distance between the instrument and the ice surface and in coil orientation increase the total measurement error to approximately 0.05 m for 2 m thick ice when compared against profile drill-hole ice thickness measurements (at a point spacing and profile length that captures the relevant ice roughness features; Haas et al., 1997). Taking into account errors associated with drill-hole measurements and averaging over repeat profile measurements results in an error estimated at approximately 0.05 m (ice thickness 2 m) for the present ice ablation data set. As derivation of ice thicknesses in ridged ice sections with keel depths larger than approximately 7.5 m or level ice thicknesses >6 m requires 2-D modelling and futher data analysis, ablation measurements have been restricted to ice thicknesses smaller than these critical thresholds. The data structure of all files is the same, with the first column indicating the relative position along the profile in meters. The second column gives the apparent conductivity as determined with the EM-31 in mS/m. The third column gives the depth of melt puddles (if present). The fourth column provides the ice thickness (not including snow layer or meltwater layer in the presence of puddles). The ice thickness has been derived from the semi-empiric ice thickness (Z, in m) - apparent conductivity (C, in mS/m) inversion equation Z = 7.71 - ln(C - 79.5) / 0.913 which is based on a large data set of first-, second- and multi-year Arctic sea ice (Haas, 1997). Comments, such as the location of ablation gauges, are noted in the fifth column. The data block is preceded by a header of two lines indicating the location and day (line 1) and identifying the column variables and units (line 2). Thickness data have been obtained repeatedly from the start to the end of the ablation season along a total of 13 profiles at 7 different locations. Except where noted otherwise, the co-location error between the same point at different dates is less than <2 m for all SEA, TP and APL profiles, and less than <5 m for the ATL and MAI profiles. The file name indicates the day of the year (1998) of the profile measurements (digits 3-5) and a three letter code (digits 6- 8) identifies the location of the profile. These are: APL or ALn, where n is an integer from 1 to 3: Airport lead (a.k.a. Sarah's Lake). Three parallel profiles have been measured perpendicular to the lead's edge. Profile AL1 is at the center and originates at ablation gauge 300 (originally about 5 m away from the lead's edge). Profile AL2 is 10 m closer towards the ship and profile AL3 is 15 m further away from the ship. ATL: Atlanta. The profile starts at the ablation gauge site and extends along the snow depth profile towards Tuk. MAI: Main snow depth/ice thickness profile. This profile starts at Post No. 5 near the Seattle ablation gauge sites and continues along the main snow depth profile across The Ridge past Pittsburgh towards the Generator Hut. SEA or SEn, where n is an integer from 1 to 5: Seattle Lead. Four to five parallel profiles have been measured perpendicular to the lead's edge. Profile SE1 is near the center and runs past the ablation gauges 306 to 310 (see comments for exact locations). The other profiles are parallel to SE1 spaced between 5 and 20 m apart to either side. TP1 or TP2 or Pxy, where x and y are either 1 or 2: Ice surface topography profiles number 1 and 2. Profile 1 was located in level second-year ice parallel to the ocean hut line towards Sediment City. Profile 2 was located in deformed multi-year ice extending at a right angle from The Pass (through the active ridge separating the Des Groseilliers from the location of the main camp) towards the logistics huts (for details on these profiles see metadata file on sea- ice topography measurements). In the file name "EMdddPxy", x indicates the number of the surface topography profile and y indicates the measurement run along that profile (which was occasionally repeated running in the opposite direction). Note that the profiles for day 239 were obtained after melt ponds had started to freeze. Thus, ice thicknesses indicated for all profiles EM239... include the thickness of the meltwater layer in underlying ponds (i.e., in contrast with convention that applies to all other data sets). TUK: Long ice thickness profile originating at the Tuk ablation gauge site, traversing an area of deformed ice and after crossing of a ridge onto a neighbouring floe continuing along through mostly level 2nd-year ice towards a ridge at the end. Note that due to losses of markers along this profile, the co-location error for the four data files may be >50 m in spots such that no direct comparison of individual morphological features is warranted. Note that the profiles for day 243 were obtained after melt ponds had started to freeze. Thus, ice thicknesses indicated for profile EM243TUK include the thickness of the meltwater layer in underlying ponds (i.e., in contrast with convention that applies to all other data sets). References: Haas, C., S. Gerland, H. Eicken, H. Miller (1997) Comparison of sea-ice thickness measurements under summer and winter conditions in the Arctic using a small electromagnetic induction device, Geophysics, 62(3), 749- 757 Haas, C. (1997) Determination of sea-ice thickness with seismic and electromagnetic-inductive methods (in German). Ber. Polarforsch., 223, 161pp.