Meteorology and Hazard Prediction
Posted March 15, 2011
With the after-effects of the earthquake and tsunami in Japan including a number of nuclear plants periodically leaking low levels of radiation and under some threat of larger releases, we've gotten a few e-mails from people concerned about family in other parts of the world, including one mother very worried about her child in Australia. It was easy to assure her that even a much more serious release of radioactive material would be no threat at that kind of distance and especially in the southern hemisphere. However, in the vicinity of central and northern Japan the situation is quite different, and as you might expect meteorology and weather prediction likely play a significant role in guiding specific precautionary and reactive measures to be taken during and after the event.
In a past radiological release from a nuclear fuel production facility in Japan, the Japanese Nuclear and Industrial Safety Agency (NISA) worked together with American scientists in operating independent weather forecast and radiological transport & dispersion models in order to predict and perform post-analysis on those plumes. While we don't have any confirmation that similar coordination is underway at this point, it does seem likely that both NISA and the U.S. National Atmospheric Release Advisory Center (NARAC) which is operated by Lawrence Livermore National Laboratory (LLNL) are running models for a range of scenarios to aid in planning and response.
In the case of NARAC, the center utilizes a series of meteorological modeling systems, and then marries the analyzed and forecast weather data with source models that attempt to account for the type and rate of release and whether it is neutrally buoyant, for example, or involves the heat of fires or explosions that would change the character of the initial dispersion. The models available range from relatively fast and simple PC-based rapid response systems (for example, the Hotspot model) to much more sophisticated and high-resolution systems that require more time to run and also a more detailed and carefully considered set of inputs for initialization. Other agencies of the U.S. government also have models for specialized circumstances including nuclear material releases or explosions, or chemical weapons releases, for example the Hazard prediction and Assessment Capability (HPAC) system used by the Defense Threat Reduction Agency.
To give a rough feel for the kinds of information being utilized or produced by some of these models and a sense of how the meteorology of eastern Japan might be affecting the transport of materials released from the areas in question, I ran some trajectory and plume projections using the READY site at NOAA's Air Resources Laboratory. These are NOT specific to radiological releases or meant to indicate specific hazard areas, since it does not include source strength information or any kind of treatment of material deposition, nuclear decay and other processes that are included in NARAC, HPAC and so on. The first two images are 24-hour trajectories showing the movement of parcels of air starting at altitudes of 10, 1000 and 5000 meters above the ground from Sendai airport, Japan beginning early yesterday morning, and a again beginning early today. Notably, the low level trajectory, as is often the case, is more variable in initial direction, while higher altitudes tend more quickly to the prevailing west to east direction of flow. The next two images are the same, except starting from Narita airport farther south in order to bracket the region of concern in between.
The fifth image is an example of a post-processed particle plume that assumes a ten-minute release late last week and shows how it might have progressed after a twelve-hour period, in this case principally moving offshore and out to sea. The next image is an integrated time-of arrival map showing when that truncated plume would arrive at various distances offshore. I also included a link to an animated plume forecast posted online by the German magazine Spiegel. This gives a sense of how a more continuous plume might be affected by passing weather systems and diurnal variations in low-level wind flow. Note that I was unable to find any description of what model is used in this animation, and also note that the contours do not appear to have any specific meaning in terms of radiological dose or concentration of radioactive materials. Instead, they are simply using a generic source rate and showing relative reductions in concentration due to turbulence and diffusion. Still, it gives a rough sense of how the movement of emitted materials may have progressed over the past few days and into the next few. Again, these images and animations are rather generic - the NISA and NARAC simulations would be much higher resolution and detailed, and much more tailored to the specifics of the ongoing incident.
Finally, I've included a few links for those interested in more background on the models and data sets used in the NARAC or those who might like to experiment with running trajectory or dispersion models through the Air Resources Lab site, and a link for updates on the ongoing situation in Japan from the International Atomic Energy Agency.