1996A&A...306..108K

Color excesses in the uvbyβ-system for most A3-G0 stars brighter than B=11.mag 5 and above b=+70deg have been averaged in one degree latitude zones for reference and to see if the cosecant postulate has any meaning for the polar cap itself. Two latitude intervals have tolerable cosecant relations: [70<b<81, <E_b–y≥+0.025cosec(b)-0.012 all E_b–y] and [83<b<90, <E_b–y≥+1.805cosec(b)-1.800 all E_b–y] for the complete cap the correlation may be improved if an E_b–y selection is done according to what could be a simple distinction between intercloud and cloud lines of sight, E_b–y≃0.017: [70<b<90, <E_b–y≥+0.021cosec(b)-0.022, E_b–y<0.017], [70<b<90, <E_b–y≥+0.056cosec(b)-0.023, E_b–y>0.018] and [70<b<90, <E_b–y≥+0.105cosec(b)-0.063, E_b–y>0.030] between 81deg and 83deg the general average oscillates. Correlation coefficients for the five relations are 0.35, 0.98, 0.45, 0.50 and 0.21 respectively. E_b–y=0.030 is the typical reddening in a diffuse cloud. These results are somewhat surprising: averages for the complete zone do display a marginal cosecant variation in the range from 70deg to 82deg whereafter a marked drop to the polar reddening E(b-y)_89–90=0.004mag takes place. The physical relevance of the average latitude trend may however be questioned since each of the four quadrants show distinctly different trends despite each containing ∼1200 stars. In the quadrant 90<l<180 there is a sharp increase from E_b–y=0.008 to 0.028 when b is varied from 70deg to 76deg. Some 30deg longitude sectors, e.g. l:120-150, have significantly increasing average reddenings and some display a stronger decrease than the average cosecant slope, e.g. l:300-330deg. The three 30deg sectors of the 4th quadrant are all different. The plain average value of E_b–y is 0.0137mag and the standard deviation 0.0220, a value also valid for most of the latitude zones. The mean corresponds to A_B=5.59x0.0137=0.077mag. The cosecant fits to the one degree averages predict a polar reddening at b=90deg of E_b–y=0.013-0.016 for all stars and those with a reddening exceeding -0.015. These values correspond to a semi thicknesses of the disk A_V≃0.056-0.068. These values are remarkably close to the minimum projected value recently observed for a 2kpc path through the disk at a latitude +4deg A_V=0.073 at l=262deg, Jonch-Sorensen & Knude (1994). In the cap beyond ≃84deg the average reddening is significantly smaller than the csc extrapolation and the grand average: <E>_{(b–y,b>87deg)}=0.007, but excesses larger than 0.030 are also found at these high latitudes. A combination of the data presented here to previously published uvbyβ excesses may indicate a shift in the reddening's latitude dependence between ≃40deg and ≃45deg. A downwards jump of about 0.040 in E_b–y seems possible. For extra Galactic investigations there has sometimes been a discussion of a clear window above b=50 deg or a cosecant variation all the way to the pole with A_B,pole≃0.2. As the discussion shows the number A_B, pole_≃0.2 is valid for several subsamples of the polar stars related to the cloud reddening, but ≃(2)/(3) of the lines of sight have E_b–y<0.018 producing an average A_B,pole=0.0. A simple general reddening law is thus not aplicable for the polar cap and if an accuracy of A_B better than 0.10 is required the reddening must be measured. If the object is known to be reddened the relation valid for E_b–y>0.018 should be used.