the files contains RCS (dBsm) and auxiliary information (Latitude, Longitude, Time) that permits to generate Figures 2-5 and Extended Data Figure 2. The code implementation for the processing of echoes is explained: The input array IN, which is two-dimensional, contains the 128 fast time (I, Q) for N consecutive individual echoes (IEs). The first step is to use FFT in the fast time dimension to transform the IN array into the RGM array. RGM is the radargram with two dimensions: one is 128 range bins and the other dimension the N IEs. RGM = scale_plrm.*... % IEs radargram (abs(fftshift(fft(IN,PAR.N_bins_SAR.*PAR.rangeOversampling,1),1)).^2)./... PAR.N_bins_SAR.^2; %%% radargram For each IE, the maximum peak in range is selected in the radargram, which we call Pu. Then we take log10 to output in dBsm, as follows: RCS=10*log10(Pu )+... scale_RCS+L2_atmo; % this is the Radar Cross Section Scale_RCS and scale_plrm are the scaling factors documented in Dinardo and Lucas (2024).”