56 (3) The amount of 137Cs released directly into the ocean was estimated based on the concentrations at the northern and southern discharge channels of the Fukushima Daiichi NPP, which were monitored approximately twice per day. Firstly the daily concentrations were averaged and then the amount of 137Cs at the sea surface within a volume of 1.5 km × 1.5 km × 1 m was calculated assuming that 137Cs with averaged concentrations exists in the volume, because the distance between the northern and southern discharge channels is around 1.5 km. The calculated amounts were adjusted by multiplying the constant (7.5) obtained from a comparison of the total released amount of 137Cs during 120 hours between 1–6 April 2011 with the values from TEPCO monitoring [74], which states that the total released amount of 137Cs during this period was 0.94 PBq (1 PBq = 1015 Bq). Figure 37 illustrates the resulting temporal variation of the released amount of 137Cs to be used in the numerical simulations. This source term estimation leads to a total 137Cs release of 3.5 PBq for the period 26 March–30 June 2011. Therefore, atmospheric deposition data, integrated over three hour intervals, together with the direct release source term presented in Figure 37, have been used in order to make realistic simulations on the dispersion of 137Cs released from the Fukushima Daiichi NPP. 3.5.2. Results The time frame of calculations extends from 12 March until 30 June 2011. Time series of calculated 137Cs concentrations in surface water are provided by models for the points listed in Table 16, for which time series of measured concentrations were obtained by the Japanese company TEPCO (T1–T8). These measurements have been reported in regular press releases [76]. Three additional points were sampled [77] and have been included to compare model results with measurements taken at larger distances from Fukushima. The location of sampling points is shown in the map of Figure 38. Two sets of calculations have been carried out: ? Exercise 4a, with common parameters (as in Exercise 3): circulation, bathymetry, diffusion coe?cients, adsorption/desorption parameters and sediment characteristics. No suspended matter in the water column will be considered since, as has been shown before, it does not play a significant role; ? Exercise 4b, each model will use different water circulation and its own set of parameters. Thus, a range of expert estimations will be provided. Results of these exercises are described in detail in the following sections.