Atmospheric modelers are still warily watching the spread of radionuclides from the stricken Fukushima nuclear plant as workers there struggle to bring the situation under control. "The wind is blowing everything out to sea, so they've been lucky so far," says Andreas Stohl of the Norwegian Institute for Air Research. Modelers' biggest problem at the moment is scant knowledge of how much radioactive material is actually escaping from Fukushima.
"The data flow is very incoherent," Stohl says. The weather is also causing concern as the forecasts currently show a change in wind direction on Sunday night and Monday that could blow contamination toward Tokyo. "Things may change in the meanwhile," says Gerhard Wotawa of Austria's Central Institute for Meteorology and Geodynamics in Vienna. "But if the worst happens [at the plant], then that could cause problems there [in Tokyo]."
Modelers have to first estimate how much material is coming from the Fukushima plant, known as the "source term," and then take current wind field data provided by, for example, the European Centre for Medium-Range Weather Forecasts. Then they plug all this into an atmospheric model to predict how the material will spread over the coming days. Stohl says information on the source term is so scarce that his model  has had to rely in part on media information. Wotawa, on the other hand, has access to data from radionuclide sensors belonging to the Comprehensive Test Ban Treaty Organization for his model .
CTBTO has a network  of different types of sensors scattered around the globe to detect signs of clandestine nuclear cheats. Of the 60 radionuclide stations currently operating, there are more than 20 on islands in the Pacific and around the Pacific Rim. Such sensors would be a valuable resource when dealing with a nuclear accident, but the organization's mandate says the information can be given only to signatory governments, not released publicly. CTBTO radionuclide stations test for radioactive particles; air filters are checked once daily with a gamma-ray detector to identify what radionuclides may be present and in what quantities. It takes 2 or 3 days for the information to get to the signatory governments. Wotawa says that there were problems with early data from one of the two stations in Japan this week, first because of a power failure following the earthquake and then because of contamination of the filter. Those problems have now been resolved and he is expecting the latest data tomorrow.
Based on the radionuclide data he has received so far, he believes none of the Fukushima reactors has experienced a full meltdown. Most of what has been detected have been volatile isotopes, including iodine, cesium, and xenon. A meltdown would have added less volatile elements to the mix, such as zirconium and barium, and these haven't been detected in large quantities.
Because of the source term uncertainties, all the dispersion models can do is give "qualitative advice on where the plume is going," says Stohl. Thus far, most has been blown out to sea and is unlikely to reach places like North America in significant concentrations because of three factors:
- Much of the matter has a short half-life and decays away over a few days.
- The plume often gets lifted up into the upper troposphere by anticyclones.
- The plume gets more and more diluted the farther it travels.
Stohl says the material will likely be measurable in the United States and Europe but not in any concentrations that will impact on human health.
The change in wind direction forecast for the end of Sunday is causing some concern, however. Stohl says a cold front is forecast to pass over Japan and the wind will start blowing from the north and that may continue for a couple of days. "If it blows over populated areas, there could be a problem," Wotawa says. "And in combination with rain the contamination would be more severe." With the amount of material currently coming from the reactors, this would require Tokyo residents to stay indoors for only a couple of days, Wotawa believes. And this is still only a tentative prediction. "It can be very difficult to make reliable predictions of processes such as the dispersion through the atmosphere in view of the variability of the wind speed and direction on the short term," says Luigi Monte of the Italian nuclear agency ENEA.