COSMIC BACKGROUND ANISOTROPIES AND GALACTIC DUST - A NEW METHOD FOR EXTRACTING THE POWER SPECTRAL INDEX FROM DATA

Various techniques for extracting the power spectrum from a map of cos mic background anisotropies (CBAs) have been discussed in the literatu re; the best signal-to-noise ratio is expected for a large and symmetr ic map. Unfortunately, many of the data sets already available do not provide a map; typically they consist of a limited number of long stri pes over the sky. We discuss a straightforward method for estimating t he power spectral index from these data sets. The method (called ''Del ta statistics'') is tested over the infrared IRAS map, where it allows us to avoid the effects of some artifacts, like extended stripes with different detector responsivity. We show that it is possible to exten d the analysis down to a differential brightness of 0.05 MJy sr(-1) at 100 mu m, with an angular resolution of 3'; at these low levels dust anisotropies have a spectral power index -2.5 less than or equal to n less than or equal to -1.5. While these ''local'' values of n are not in disagreement with a unique value of n if the cosmic variance is tak en into account, the selection of clean sky regions where fluctuations are at a minimum will peak up zones with hat power spectra, thereby s uggesting possible significant contributions to CBA at small angular s cales, The data provided by balloon experiments are particularly suita ble for Delta statistics analysis, being taken while the gondola is sl owly moving parallel to the horizon. The sky stripes explored by ULISS E and ARGO experiments are reanalyzed and provide values of n(ULISSE) = -2.2 +/- 0.7 around l = 20 and n(ARGO) - -1.4 +/- 0.4 around l = 200 . These values are independent of COBE normalization, calibration unce rtainties, and, to a great extent, of the details of the experiment, s uch as beam size, modulation amplitude, and type.