There has been speculation that the most extreme release of radiation was associated with fires in the fuel pool of reactor 4. The wind and plume maps below support this hypothesis.
The aerial radiation mapping conducted March 17-19 by the DOE’s National Nuclear Security Administration (NNSA) reveals a pattern of radiation transport and deposition toward the northwest of Fukushima at about a 45 degree angle. The shape of the plume suggests a single incident, probably of several hour duration. It is worth exploring which reactor event caused this deposition pattern.
The days leading up to the aerial radiation survey were fraught with explosions, steam and fire. The issues at reactor 2 were ongoing during much the week; however the dramatic events at reactor 4 were discrete in nature and occurred over the span of approximately one day. A chronology of events that day is presented here and here. At the spent fuel pool of reactor 4 on March 15 (all times JST) there was an explosion ~6 am and then a fire beginning at ~9:30 am and lasting several hours. Boiling water in the spent fuel pool of reactor 4 was reported at ~5:30 pm. By 9 pm the radiation emanating from the reactor 4 spent fuel pool was so intense that the majority of the plant staff were evacuated. And then again a report of a fire on March 16 at 5:45 am, with indications that the prior day’s fire had not been completely extinguished. The radiation measured at the plant over time is illustrated in this data visualized by NPR (although it is a useful chronology, the magnitudes depend greatly on the locations within the plant; the gaps in the record are due to missing measurements):
Note the large radiation values experienced at the plant perimeter on the evening of March 15.
The Japan Nuclear Safety Commission (NSC) has recently begun releasing to the public its plume forecasts. These maps go back to the outset of the crisis and are a good source of data to help pinpoint the timing of the large release that created the deposition shown in the aerial radiation map. Throughout the morning of March 15 the winds were toward the south/southwest in the area of the Fukushima reactors. This is why Tokyo and surroundings received elevated (but low-level) radiation around 10:30 am (dosage timeseries here) associated with the morning fire at the reactor 4 spent fuel pool. Then after 1 pm, the winds shifted toward the northwest and remained that way until ~1 am Wednesday. Thus these northwest winds lasted ~12 hours. This period of northwest winds is shown below by the NSC forecast winds and plume for 6 pm JST on March 15. Winds then shifted toward the south/southeast (and increasingly offshore) for all of Wednesday and Thursday (March 16 and 17).
The northwestward winds in the afternoon and evening of March 15 correspond to the time of significantly enhanced emissions from the reactor 4 spent fuel pool – likely associated with an ongoing fire there. The deposition pattern is supported by the timing of these events. The subsequent evacuation zone modifications were designed to encompass this region of higher radiation and thus individual communities along the plume path were asked to evacuate on April 11. This guidance deviated from the prior circular evacuation zones that had been established up to then and was an important recognition of the role that coastal winds play in shaping transport.