Figure 5

(a) Amplitude of the electric field \(\overline{\mathrm{E}}(\mathrm{y})\) produced by an LWA in a vacuum and sampled at a point \({\mathrm{p}}_{0}=\mathrm{p}(0,{\mathrm{y}}_{\mathrm{j}},{\mathrm{z}}_{\mathrm{k}})\) normalised to the value of the electric-field amplitude calculated at \({\mathrm{p}}_{\mathrm{if}}=\mathrm{p}\left(0,{\mathrm{y}}_{\mathrm{if}},{\mathrm{z}}_{\mathrm{k}}\right)\); being \({\mathrm{z}}_{\mathrm{k}}>0\) and \({\mathrm{y}}_{\mathrm{j}}>{\mathrm{y}}_{\mathrm{if}}=1.5\uplambda\). (b) Amplitude of the electric field E for \(\mathrm{z}=5\uplambda\) in three different media with growing conductivity. The LWA was designed such that the attenuation vector would have been parallel to the separation surface for \(\upsigma =0.05\) S/m, \(\upmu ={\upmu }_{0}\), and \(\upvarepsilon ={\upvarepsilon }_{0}\).