AN INVESTIGATION INTO THE HORIZONTAL STRUCTURE OF SPREAD-ES

Observations by FROM and WHITEHEAD [(1986), Radio Sci. 21, 309] indica ted that spread-E(s) consisted of many small clouds covering a large a ngular range and all travelling with similar velocity, but each reflec tor only lasting for 2-3 s. Model data used with a similar analysis to that used by From and Whitehead have shown that the clouds may well b c long lasting. The noisy, short-lived appearance of the echoes is an artefact of the analysis procedure. These spurious results only occur when many echoes, extending over an angular range much larger than the beamwidth, are situated closer to each other than the angular resolut ion of the radar. An investigation shows that model data generated fro m long-lasting echoes with reflection points spaced closer than 3 km i ntervals, and travelling with the same velocity, produce similar resul ts to experimental spread-E(s) data. An example is also presented show ing the structure of spread-E(s) moving with a similar velocity to the overall drift of the E(s) layer. A better model for spread-E(s) then becomes one where a larger cloud, produced by the windshear mechanism, subsequently has many small irregularities of the order 1 km in exten t produced within it by some other disturbance. The irregularities mov e with the same velocity as the larger cloud, which is consistent with non-field-aligned irregularities produced by the gradient instability , but unlikely for irregularities caused by atmospheric waves. The con clusions of JONES [(1984), J. atmos. terr. Phys. 46, 1179] concerning short-lived reflectors in the E-region must also be open to question b y these results.