PLANCK is ESA's third medium-sized mission, to be launched simultaneously with the FIRST Cornerstone mission in early 2007. PLANCK will survey the whole sky at frequencies between 30 and 900 GHz in order to map the cosmic microwave background anisotropies. Finland participates in the international PLANCKLFI (Low Frequency Instrument) Consortium. Finnish companies and institutes build parts of the 70 GHz receiver. Turku and Helsinki Universities, as well as the Helsinki University of Technology, will participate in the processing and scientific use of the PLANCK data. The Finnish science participation is coordinated by K. Enqvist from Helsinki University, with E. Valtaoja as Tuorla's representative Associate in the consortium. The radio astronomers at Tuorla and Metsähovi are mainly interested in the so-called foreground sources, quasars and other AGN, whose contribution must be removed from PLANCK maps in order to map accurately the 3K background. As a by-product, a unique, complete radio-to-IR database of point sources will be generated for further studies of AGN. Also, simulations of large scale structure formation in a low density Universe are planned by P. Heinämäki, A. Chernin and M. Valtonen for comparison with the PLANCK data.

Hubble law and the Malmquist bias, both classical and cosmological

Problems with the extragalactic distance scale and the determination of the Hubble constant H_o have been studied by P. Teerikorpi, T. Ekholm, and M. Hanski, in collaboration with the French groups involved in the LEDA extragalactic database (G. Paturel from Lyon, L. Bottinelli, L. Gouguenheim, G. Theureau from Meudon). Extensive work on the value of the Hubble constant, based on the direct Tully-Fisher relation and the method of the normalised distances for the large KLUN sample of galaxies, was completed and published in 1997. After this effort an interesting challenge was to understand why the inverse Tully-Fisher relation, often regarded as an "unbiased" distance indicator, stubbornly continued to yield very high values of H_o, in comparison with the direct relation (about 90 versus 55 km/s/Mpc). And finally the reason was found in the form of what was termed "calibration sample bias".

It had been thought that the inverse TF relation method is immune to the nature of the calibration sample (eg. whether it is magnitude or volume limited), contrary to the direct relation. However, it was realized that this is true only in the ideal case when the inverse slope applicable for the field galaxy sample is the same as the calibrator slope. As these slopes are normally different, the fact that the available (Cepheid-based) calibrator galaxy sample is not volume-limited leads to a biased value of the Hubble constant derived from the field sample. A new approach involving a correction formula for this calibrator sample bias was found and now the values of Ho from the inverse and direct Tully-Fisher relations have been shown to be consistent which each other (around 55 km/s/Mpc).

There was a collaboration with G. Paturel (Lyon) on a study of how to derive H_o using Paturel's sosies (look alike) galaxy-distance indicator. In this work the new HIPPARCOS calibration of Cepheids (by Paturel) was implemented. The work was published in 1998 and also yielded H_o ~ 55 km/s/Mpc.

In an Ann. Rev. Astron.Astrophys. article in 1997 P. Teerikorpi reviewed some historical and modern work made on the various Malmquist-like biases (He attempted to coin the terms Malmquist bias of the first and second kinds to replace the existing confusing nomenclatura. This review was concerned with the local galaxy universe where space may be regarded as Euclidean and the flux of light follows the classical inverse square law as a function of distance. The question arose what the Malmquist bias is like in more general cosmological situations, and led to a study of the cosmological Malmquist bias in the Hubble diagram (log z vs. m) in Friedmann universes. This work, published in 1998, contains the basis for calculating the Malmquist bias (of the 1st kind) for gaussian standard candles and comparing observations with the predicted curves which are the standard Mattig curves of the Friedmann models. The Mattig curves remain valid when the Hubble diagram is studied in the sense of m vs. log z, but then the Malmquist bias of the 2nd kind has to be considered.

Large scale structure of the galaxy universe

Together with Yu. Baryshev (St.Petersburg) and the group of L. Pietronero (Rome), including F. Sylos Labini and M. Montuori, P. Teerikorpi has studied the problem of fractality in the space distribution of galaxies. This collaboration resulted in an article published in 1998, on the cosmological implications of the observed large-scale fractality. In particular, it was pointed out that the co-existence of the good, linear Hubble law (down to the outskirts of the Local Group) and the very inhomogeneous fractal structures with fractal dimension D ~ 2 in the same spatial scales represents an intriguing problem within the Friedmann universe where the linear Hubble law is the strict consequence of homogeneity.

Linear perturbation approximation calculations of the Hubble velocity field allowed one to make some conclusions on how the observed linear Hubble law could be explained in the frame of Friedmann models (a developed version of the Sandage-Tammann- Hardy test of 1972). If all matter, both luminous and dark, participate in fractals (at least up to the scale of 200 Mpc), then the density parameter Omega must be very small, at most around 0.01, otherwise one expects strong deviations from the linear Hubble law. If the luminous matter only is involved in fractals and one considers larger values of Omega, then it is required that there is a uniform distribution of dark matter taking care of the good Hubble law. The density of the uniform dark component is predicted to be high, e.g. if Omega = 1, then Omega_dark should be at least 0.99. This result is independent of early universe physics.

As such calculations depend on approximating fractality with spherically symmetric Tolman-Bondi models, Yu. Baryshev (St.Petersburg), A. Gromow (St. Petersburg), D. Suson (Texas University), and P. Teerikorpi have started a detailed investigation on theoretical problems in this approach.

Though the so-called gamma function method of Pietronero has gradually been accepted by students of the large-scale galaxy distribution, the debate on the maximum spatial scale of the fractality has continued. In order to check in a special case the Pietronero group's conclusion that there is a large scale (>200 Mpc) fractality, P. Teerikorpi and M. Hanski (in collaboration with Yu. Baryshev and the French KLUN-group) calculated the all-sky average radial space density distribution using the KLUN sample and a novel method based on TF-distance moduli. The result was published in 1998 and implies that the local galaxy universe around us is "thinning" outwards, at least up to 200 Mpc, as expected if the galaxy distribution is fractal with the fractal dimension D ~ 2.2. This same value of D was derived by Pietronero's group from the LEDA database sample using the general gamma function method.