Abstract:The deposits of the Toruń Basin are dominated by a few-metre thick sand series which fill up buried valley-like depressions. In many cases they underlie the Weichselian till which builds up the ice marginal streamway (pradolina) terraces or they are exposed at the basin slopes. As the results of the geological and sedimentological studies, as well as of the dating of the deposits at the sites in the Toruń Basin indicate, the deposits include two fluvial series accumulated before the advancement of the Leszno Phase ice sheet, i.e. in Middle Weichselian and at the beginning of Late Weichselian. The oldest fluvial series connected with the Saalian Glaciation was found at the mouth section of the Drwęca Valley. The fluvial system of the Toruń Basin during Middle Weichselian and at the beginning of Late Weichselian developed in two phases of the sand-bed braided river. During the first one the river channel were dominated by large mid-riverbed sandbars, while during the second phase the water flow was smaller and, as a result, low transverse sandbars and two-dimensional dunes developed. Other active river channel also showed low-energy flows, more intensive meandering than in the case of the braided rivers, as well as sandy side-bars. Analysis of the rounding and frosting of the quartz grains indicate that the studied series of the Weichselian sandy deposits represent alluvia of a river which were fed from two diverse sources. The first one might have represented the alluvia of a warm river which transformed its load, while the other one might have mainly carried the underlying Quaternary deposits.

Dis tri bu tions of equiv a lent doses (D E ) ob tained by the Sin gle Aliquot Re gen er a tive-dose (SAR) OSL method ap plied for large aliquots of coarse quartz grains ex tracted from flu vial sed i ments are pre sented and ana lysed with re spect to a flu vial palaeoenvironment. The Nowe D¹bie and £ochowo flu vial suc ces sion from the west ern part of Toruñ Ba sin (east ern part of Noteae-Warta streamway, Toruñ-Eberswalde ice-mar ginal val ley) was ana lysed. The flu vial depositional con di tions con trolling the ex tent of day light bleach ing are re con structed by sedimentological stud ies. The re la tion be tween the amount of bleach ing and sedimentological prop er ties of flu vial de pos its in di cate that rip ple cross-lam i nated sands that ac cu mu lated on the floodplain and hor i zon tally-bed ded sands de pos ited in shal low chan nels are more ap pro pri ate for OSL dat ing than sands de rived from the deep chan nel. Along with lu mi nes cence re sults ob tained for the river de pos its, data mea sured from an ancient pot tery sam ple, en sur ing com plete re set of the OSL sig nal, are pre sented, com pared to and dis cussed. On this base the poorly bleached sed i ment sam ples are iden ti fied and an ad just ment fac tor is in tro duced for cor rect ing their OSL dat ing re sults in or der to avoid age in ver sion. The ap pli ca tion of the ad just ment fac tor is found to min i mize over es ti ma tion of OSL ages of sed i ments.Key words: OSL age in ver sion, quartz, op ti cal bleach ing, sand-bed braided river, ice-mar ginal streamway/val ley.

20Knowledge of the kinetic parameters E (thermal activation energy) and s (frequency factor) of charge-trapping defects in the quartz crystal lattice is of paramount importance to assess the thermal stability of associated luminescence signals used for dosimetry and dating. Since methods proposed for constraining thermoluminescence (TL) kinetics usually make use of the 25 signal response to thermal treatments, accurate temperature control is required to obtain valid E and s values. In an attempt to check the extent to which consistent kinetic parameters could be obtained using routine luminescence measurement equipment, we have investigated three methods (isothermal decay, initial rise and the Hoogenstraaten method) in an intercomparison study involving eight laboratories using Risø and Freiberg Instruments systems. 30The target signal was the so-called 110 °C TL peak of a sample of Oligocene coastal dune quartz sand from the Fontainebleau sand formation (France).TL glow curves recorded with heating rates in the range 0.02-5.0 K s -1 showed peak positions varying by up to 60 °C between systems at the highest heating rates, as caused by temperature calibration errors and/or thermal lag. Kinetic parameters derived from the complete data set 35show a large spread, covering the ranges ~0.5-1.2 eV and 10 5 -10 18 s -1 for E and s. In most cases, interlaboratory variations exceeded those of replicate measurements within individual laboratories. Signal lifetimes at 20 °C derived from the isothermal decay (~57 min) and initial rise methods (at low heating rates; ~60-80 min) most closely match the directly measured value (~70 min). Finally, we discuss the consequences of these findings for dosimetry and 40 2 dating using luminescence signals and possible ways to reduce systematic errors in laboratory measurements of kinetic parameters.