PLASMA GRADIENT EFFECTS ON DOUBLE-PROBE MEASUREMENTS IN THE MAGNETOSPHERE

The effects on double-probe electric field measurements induced by ele ctron density and temperature gradients are investigated. We show that on some occasions such gradients may lead to marked spurious electric fields if the probes are assumed to lie at the same probe potential w ith respect to the plasma. The use of a proper bias current will decre ase the magnitude of such an error. When the probes are near the plasm a potential, the magnitude of these error signals, delta E, can vary a s delta E similar to T-e(Delta n(e)/n(e)) + 0.5 Delta T-e where T-e is the electron temperature, Delta n(e)/n(e) the relative electron densi ty variation between the two sensors, and Delta T-e the electron tempe rature difference between the two sensors. This not only implies that the error signals will increase linearly with the density variations b ut also that such signatures grow with T-e, i.e., such effects are 10 times larger in a 10-eV plasma than in a 1-eV plasma. This type of err or is independent of the probe separation distance provided the gradie nt scale length is much larger than this distance. The largest errors occur when the probes are near to the plasma potential. At larger posi tive probe potentials with respect to the plasma potential, the error becomes smaller if the probes are biased, as is usually the case with spherical double-probe experiments in the tenuous magnetospheric plasm as. The crossing of a plasma boundary (like the plasmapause or magneto pause) yields an error signal of a single peak. During the crossing of a small structure (e.g., a double layer) the error signal appears as a bipolar signature. Our analysis shows that errors in double-probe me asurements caused by plasma gradients are not significant at large sca le (much greater than 1 km) plasma boundaries, and may only be importa nt in cases where small-scale (< 1 km), internal gradient structures e xist. Bias currents tailored for each plasma parameter regime (i.e., v ariable bias current) would improve the double-probe response to gradi ent effects considerably.