Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
G Use of Atmospheric Models in Response to the Chernobyl Disaster Summary of a presentation by James Ellis, LLNLINARACi The Chernobyl nuclear accident occurred on April 26, 1986. The size of this release was unprecedented, releasing millions of curies of radioactive material including iodine-131, cesium- 137, and strontium-90 all of which are potentially harmful to human health. The initial release occurred on a Friday night. By Sunday afternoon, contamination readings were picked up on workers at a nuclear power plant in Sweden. Within a few hours it had been determined that the contamination source was from a nuclear power plant to the south. By Monday, the Russians admitted that a major accident had occurred, and on the same day, LLNL/NARAC was notified by the Department of Energy to begin predicting the consequences. NARAC worked round-the-clock for two weeks, providing assistance in modeling the transport and deposition of the radioactive cloud. NARAC utilized three different model codes for these analyses 2BPUFF, PATRIC, and MATHEW/ADPIC. The 2BPUFF model is a two-dimensional long-range transport and diffusion model used mostly for estimating the Chernobyl accident release amounts of radioactivity. The PATRIC model is a three-dimensional puff and diffusion model that had been specifically designed to treat continental and hemispheric scales. The MATHEW/ADPIC is a combined mass- consistent wind flow model and a particle-in-cell dispersion model that was to calculate consequences over 200 km or less. For the previous 12-year period, real-time radiological dose assessments had been done at scales up to 200 km. For the Chernobyl event, MATHEW/ADPIC had to be rapidly modified to expand its capability to approximately a 2000-km domain. The Air Force Global Weather Center provided meteorological data for this work. Based on samples collected around Europe, the time and strength of the release were estimated, and approximately 40 percent of the total radioactive material was estimated to have been released in the initial blast (the rest came from the ensuing fire over the next five to six days). There was a considerable amount of rain in the area, which led to "hotshots" of wet deposition across Europe. Scientists were not able to model these washout processes, because they did not have the needed meteorological precipitation data or fine-scale forecast model precipitation pro- ~ Ellis emphasized that it is difficult to reconstruct the exact history of LLNL involvement in this event, since those who participated directly in this response effort are no longer at LLNL. 87
88 APPENDIX G ducts. With today's high-resolution mesoscale forecasts, they would have been able to do a much better job of modeling the deposition patterns. NARAC models were used to estimate air concentration and ground deposition of key radioactive elements and the corresponding exposures and potential health effects. It was esti- mated that iodine deposition in the United States was insignificant but that dangerous levels of ground deposition of key radioactive elements had occurred throughout Europe. The greatest risk was not from direct exposure but from exposure through the food chain. In particular, radioactive material was deposited on farmlands and on grass eaten by cattle, forcing many countries to destroy exposed milk and crops. Aircraft measurements of radioactive material at 17,000-30,000-ft altitude above Europe, the Japan Sea, and the West Coast of the United States indicated that radioactivity from the reactor accident had gotten higher in the atmosphere than initially thought to have been possible. Based on NARAC's knowledge of the thermal energy of the blast, it did not understand how the material could have risen so high; after examining the prevailing weather patterns, NARAC surmised that convective activity in the area had driven the radioactive material up to these high altitudes in the atmosphere. The upper-level flow reached the United States (from across the Pacific) by Day 10. LLNL modeling of this event matched well with the readings from aircraft measurements. Even with the limitations of the meteorological data and model prediction capability available in the 1986 time frame, overall agreement between ground-based and aircraft measurements and model estimates was within a factor of 2 or 3. Since that event, there has been a lot of activity in Europe to improve the models used in response to nuclear accidents, culminating in development of the Realtime Online Decision Support System (RODOS) for nuclear emergency management (www.rodos.fzk.de/RodosHomePage). The dispersion models being used in this system and those being used by other national organizations within Europe have been improved from those originally used in the Chernobyl response. One of the strengths of RODOS has been to link these dispersion models to better atmospheric prediction models and to various dose pathway models, including sophisticated watershed models, with the objective of providing tools to the decision-maker for making well-informed decisions.