Article pubs.acs.org/est
Wet Deposition of Fission-Product Isotopes to North America from the Fukushima Dai-ichi Incident, March 2011 Gregory A. Wetherbee,†,* David A. Gay,‡ Timothy M. Debey,§ Christopher M.B. Lehmann,‡ and Mark A. Nilles∥ †
U.S. Geological Survey (USGS), Branch of Quality Systems, Mail Stop 401, Bldg. 95, Box 25046, Denver Federal Center, Lakewood, Colorado 80225, United States ‡ University of Illinois, Illinois State Water Survey, National Atmospheric Deposition Program, 2204 Griffith Drive, Champaign, Illinois 61820, United States § U.S. Geological Survey (USGS), Energy, Minerals, and Environmental Health, National Reactor Facility, Mail Stop 974, Bldg. 15, Box 25046, Denver Federal Center, Lakewood, CO 80225, United States ∥ U.S. Geological Survey (USGS), Office of Water Quality, Mail Stop 401, Bldg. 95, Box 25046, Denver Federal Center, Lakewood, Colorado 80225, United States S Supporting Information *
ABSTRACT: Using the infrastructure of the National Atmospheric Deposition Program (NADP), numerous measurements of radionuclide wet deposition over North America were made for 167 NADP sites before and after the Fukushima Dai-ichi Nuclear Power Station incident of March 12, 2011. For the period from March 8 through April 5, 2011, wet-only precipitation samples were collected by NADP and analyzed for fission-product isotopes within whole-water and filterable solid samples by the United States Geological Survey using gamma spectrometry. Variable amounts of 131I, 134Cs, or 137Cs were measured at approximately 21% of sampled NADP sites distributed widely across the contiguous United States and Alaska. Calculated 1- to 2-week individual radionuclide deposition fluxes ranged from 0.47 to 5100 Becquerels per square meter during the sampling period. Wet deposition activity was small compared to measured activity already present in U.S. soil. NADP networks responded to this complex disaster, and provided scientifically valid measurements that are comparable and complementary to other networks in North America and Europe.
■
INTRODUCTION On March 11, 2011, a magnitude 9.0 earthquake centered off the Pacific Coast of Japan triggered a tsunami and a (reported) 14 m tidal wave that inundated a large part of northern coastal Japan.1 With the resulting massive flooding, the Fukushima Dai-ichi Nuclear Power Station was left without primary or backup electric power and thus cooling water, leading to catastrophic failure among its six nuclear reactors. Subsequent explosions led to a month-long discharge of radioactive material into the atmosphere, which has since spread around the globe.1 Fission-product isotopes, including iodine-131 (131I, half-life 8.0 days), cesium-134 and −137 (134Cs, half-life 2.1 years; 137Cs, half-life 30.2 years) were reported in many different countries in the Northern Hemisphere, at variable but low levels.1 Due to potential health implications, the International Commission on Radiological Protection (ICRP) recommends continued monitoring of all radioactive fallout from any release incident.2 Various fission products were released as gaseous compounds and associated particulates during the incident, including larger quantities of short-lived 131I and longer lived 134 Cs and 137Cs. These fractions can be removed by wet © 2012 American Chemical Society
depositional processes, which include precipitation incorporating gases and particulates as they fall through the atmosphere (washout, below-cloud scavenging) or entrained directly into the cloud droplets (rainout, in-cloud scavenging).3 Dry deposition can also occur as fallout of particulates and gases to the surface through gravitational settling. Systematic monitoring of atmospheric concentrations and wet deposition of radionuclides has occurred since the origin of above-ground nuclear testing. For example, Lockhart and Patterson evaluated atmospheric measurements of fission products from nuclear weapons testing series at 25 sites along the 80th meridian (N. & S. America) and in the Pacific and identified correlation of low atmospheric concentrations, particularly for 137Cs, during periods of rainfall and higher concentrations during dry periods.4 Received: Revised: Accepted: Published: 2574
September 13, 2011 January 13, 2012 January 26, 2012 February 22, 2012 dx.doi.org/10.1021/es203217u | Environ. Sci. Technol. 2012, 46, 2574−2582
Environmental Science & Technology
Article
(RadNet) operates radiation air monitors continuously in 48 states.26 Multiple-event, bulk 131I, 134Cs, and 137Cs activity data collected March 8 to April 5, 2011 were obtained from RadNet (http://www.epa.gov/radnet/radnet-data/index.html), and arithmetically averaged for each station for comparison to NADP results. NADP contacted the U.S. Department of Homeland Security to outline the NADP response, and offered samples. NADP coordinated with the Canadian Air and Precipitation Monitoring (CAPMoN) network, which shipped samples from Snare Rapids, Northwest Territories and Saturna Island, British Columbia sites to the United States Geological Survey (USGS) for analysis. NADP coordinated with the USGS to measure radionuclide fallout in NADP wet-deposition samples on March 14, 2011. Weekly sample retention began with precipitation collected on March 15, 2011 to establish a baseline prior to radionuclide arrival in the U.S. Archiving of Mercury Deposition Network (MDN) excess samples from sites in California (CA), Washington (WA), and Alaska (AK) began March 18, 2011. Samples were prioritized for analysis consistent with direction of atmospheric transport. Samples from the western U.S. were analyzed first, and samples from the eastern U.S. were analyzed last.
Published measurements and modeling results of both wet and dry 131I, 134Cs, and 137Cs deposition rates after the Chernobyl Nuclear Power Station (present day Ukraine) incident of April, 1986 include: the incident area itself,5 Belarus soils,6 Ireland/Northern Ireland,7 Sweden,8 the Greenland Ice Sheet,9 New York City area,10 and over multiple years in Germany,11 Greece,12 northern Croatia,13 northern England,14 the United Kingdom,15 and the United States,16 among others. In the UK, the 137Cs species were mostly in the particulate phase, and removal was predominately wet deposition. However, 131I removal was important for both wet and dry deposition to grass.15 Other studies suggested over 60% of total deposition was as wet deposition.8,10 Estimates of Fukushima Dai-ichi fission-product releases and subsequent measurements of wet deposition have been recently published. Chino et al. estimated daily 131I and 137Cs release rates (emissions) over the first month.17 For example, the maximum emission rate was 1016 Becquerels per hour (Bq/h) total activity for 6 h during daylight on March 15, 2011. High emission rates (>1014 Bq/h) continued through March 30. Takemura et al. determined that atmospheric low pressure concurrent with large-scale updrafts northeast of Japan effectively moved contaminated surface air to nearly 5 km in altitude on March 14 and 15.18 A strong jet stream over Japan transported released materials across the Pacific in 3 to 4 days. Arrival over North America began in Washington State on March 17, and in wet deposition in California on March 18, confirmed independently by Leon et al.,19 the U.S. Environmental Protection Agency (USEPA),20 and Norman et al.21 Additionally, Priyadarshi et al., reported evidence of Sulfur-35 (35S) in sulfate aerosol in southern California (March 20− 28).22 They suggest leakage of neutrons around one of the reactors converted salt water chlorine (35Cl) into radioactive 35 S through a multistage decay, followed by transport across the Pacific as 35SO42‑. Activities of 131I, 134Cs, and 137Cs in air were estimated to have diluted by a factor of 105 to 108 during trans-Pacific transport.18,19 Masson et al. used trajectory analysis and observations of 131I, 134 Cs, and 137Cs from 150 European sampling locations to show initial detection in Iceland on March 19 (7 days post release), and peak 131I activity at 5.5 millibecquerels per cubic meter (mBq/m3), which occurred on March 28−30, 2011.23 The first detections in Germany were reported on March 21, with activities of 0.03−0.5 Bq 131I/L in rainwater.24 NADP Response. The National Atmospheric Deposition Program (NADP) has continuously monitored concentrations and rates of wet deposition in North America since 1978 (http://nadp.isws.illinois.edu). During the Fukushima incident, the NADP/National Trends Network (NTN) consisted of 244 sites and the Mercury Deposition Network (MDN) consisted of 104 sites across North America. NADP has continent-scale networks for monitoring precipitation quality and chemical deposition. The cooperative nature and funding of the NADP allowed for rapid adaptation to monitor wet deposition of particulate-bound and water-soluble fission products, which added to the body of information on the incident, as recommended previously by Fallon-Lambert and Bowersox.25 Eight U.S. federal agency partners, Environment Canada, numerous states and many other organizations cooperated to use NADP’s existing networks for this purpose. The U.S. Environmental Protection Agencey (USEPA) National Air and Radiation Environmental Laboratory
■
MATERIALS AND METHODS Field Collection and Laboratory Processing of Samples. Per NADP protocols, NTN samples are decanted from collector buckets to 1 L Nalgene bottles in the field and shipped without preservation to the NADP Central Analytical Laboratory (CAL) located in Champaign, IL (http://nadp.isws. illinois.edu/lib/manuals/opman.pdf). As part of routine sample preparation, the CAL filters approximately 250 mL of each NTN sample through a 47 mm diameter, 0.45 μm pore-size, polyethersulfone filter. Filters from NTN samples for weeks ending March 15, March 22, and March 29, 2011 were collected in labeled plastic Petri dishes and shipped to Denver, CO for gamma spectrometry analysis. Fission products were assumed to adsorb onto aerosol particles. Therefore, filters were analyzed before the water samples. The remaining wholewater NTN samples were shipped to the USGS Branch of Quality Systems, in Denver, CO in their original “as-received” 1 L Nalgene field bottles. USGS added Ultrex nitric acid (JT Baker) to each field bottle to approximately 0.5% acid by volume. Samples were shaken, placed in a warm water bath for one hour to desorb fission products, and then allowed to leach for 24 h at room temperature. Two-week composite NTN samples were made for each site in proportion to the measured precipitation depths for each week, volume permitting. When insufficient volume was available, samples were analyzed without compositing. Per NADP protocols, MDN samples are prepreserved with 20 mL 1% (volume:volume) hydrochloric acid for mercury analysis. Bromine monochloride (1%) is subsequently added in the laboratory prior to mercury analysis. Aliquots were obtained for mercury analysis, and remaining MDN samples were split into clean borosilicate glass bottles for gamma spectrometry analysis. One-week MDN samples were analyzed; 2-week composited samples were not prepared. In response to the incident, additional polyethylene bottles were shipped to site AK00 at Dutch Harbor, AK, and 18WA at Seattle, WA and installed in the auxiliary sample collection train. Samples collected in the polyethylene bottles were not prepreserved, nor 2575
dx.doi.org/10.1021/es203217u | Environ. Sci. Technol. 2012, 46, 2574−2582
Environmental Science & Technology
Article
Figure 1. Atmospheric back trajectories from 35 NADP sites; 5000 m end-height altitude, March 22 to April 5, 2011.
Activities were recorded in picocuries per liter (pCi/L) then converted to Becquerels per liter (Bq/L) (eq 1).
were they acidified by USGS. All MDN samples were analyzed by gamma spectrometry with no preprocessing or compositing. Sample Analysis. Samples were analyzed by U.S. Geological Survey in Denver, CO using gamma spectrometry with high purity germanium (HPGe) detector systems. Detector efficiency calibrations were conducted for the sample-todetector geometries used in this study. Background radiation is reduced with a lead shield surrounding the samples and detectors. Two HPGe detectors, (1) Canberra with 14% relative efficiency and (2) Ortec 40% relative efficiency, were calibrated using National Institute of Standards and Testing-traceable Europium-152 (152Eu) check sources for validation. Calibration was checked several times a day to ensure accuracy within ±1000 electron volts of measured peak energies. For the water samples, two standards with 1 μCi 152Eu in (1) 500 mL and (2) 1 L standard Marinelli beakers were used. Gamma spectrometry analysis was accomplished using Canberra Industries “Genie 2000” ISO 11929 compliant software. Analytical details are given in the Supporting Information and by Wetherbee et al.27 Weekly filter samples were analyzed in their original Petri dish containers. Filters were also composited by stacking them on the detector by site or grouping them by region in plastic bags. Finally, all of the filters were composited as one sample and counted for 24 h. All water samples were put into Marinelli beakers for analysis.
activity(Bq/L) = A pCi / L 0.037
(1)
where: ApCi/L= activity of radioactive isotope, in picocuries per Liter (pCi/L), and 0.037 = conversion factor from picocuries to Becquerels. Measured 131I activities were corrected for decay to the latest date and time that the collectors were open during precipitation per NADP records. The decay-corrected activities are 2.2−3.2 times higher than the measured activities. RadNet also decaycorrects its 131I values to the time of sample collection.28 Fission-product deposition in Becquerels per square meter (Bq/m2) was calculated by eq 2. deposition(Bq/m 2) = A pCi / L Ppt mm0.037
(2)
where: Pptmm = precipitation depth (mm) from NADP rain gage. (Note: 1 mm depth over 1 square meter = 1 L.) Only one 134Cs activity value was quantified by the automated peak analyzer. USGS manually reanalyzed the gamma spectra and quantified 134Cs peak areas for 23 additional samples/sites. Procedures for manual 134Cs reanalysis are described in the Supporting Information and by Wetherbee et al.27 Quality Assurance. In addition to environmental samples, four blank filter samples and acidified deionized water samples were analyzed, all with no fission products detected. NTN samples were analyzed for the week March 8 −15, 2011, before 2576
dx.doi.org/10.1021/es203217u | Environ. Sci. Technol. 2012, 46, 2574−2582
Environmental Science & Technology
Article
Figure 2. NADP, RadNet, and UC Berkeley 131I activities (Bq/L), NADP-estimated 131I deposition values (Bq/m2), and interpolated precipitation depths (mm) obtained from National Weather Service Cooperative Observer Program (COOP) stations for the period March 15 to April 5, 2011.30.
composited precipitation samples to evaluate the likely source region. Back trajectories were generated for each day and site with more than 0.25 mm of precipitation, with days defined in Greenwich Mean Time for consistency. The starting hour was defined by the recorded maximum hourly precipitation intensity during the day at the respective site. Figure 1 displays 5000 m end-height altitude trajectories, which indicate the air parcels likely originated near Japan or mixed with parcels with trajectories back to Japan. This overall pattern of the trajectories was also confirmed in more complete transport analyses by Masson et al.,23 Leon et al.,19 and Takemura et al.18
arrival of radioactive material for selected sites (CO90, IL63, NY10, VT01, and WA98) with no fission products detected. Samples were obtained from sites with co-located precipitation collectors at MA01/01MA, CA50/50CA, and CO98/ CO89. No fission products were detected for either site at MA01/01MA and CA50/50CA. Conversely, 137Cs was detected at both CO98 (0.81 Bq/L) and CO89 (0.41 Bq/L). 134 Cs activities were similar at CO98 (0.27 Bq/L) and CO89 (0.33 Bq/L). The CO98 sample was obtained from the single week ending March 29, 2011 because the collector lost power during the previous week. The CO89 sample was a true 2-week composite representing 37.8 mm of additional precipitation during the week ending March 22, 2011. Two samples (CO90 and CA99) were reanalyzed by Eberline Analytical Corporation, Richmond, CA. Reanalysis confirmed independent positive quantification within a factor of 2−4 for 137Cs for both samples (CO90: 0.44 and 0.11 Bq/L, CA99: 0.13 and 0.28 Bq/L) and for 134Cs for the CO90 sample (0.52 and 0.19 Bq/L). 131I decayed below detection by the time Eberline received the samples. Mean absolute differences between laboratories were ±0.33 Bq/L for 134Cs and ±0.24 Bq/L for 137Cs. Air Parcel Back Trajectory Analysis. Confirmatory airparcel back trajectories were estimated to evaluate the likely radiation source region using the National Oceanic and Atmospheric Administration’s Hybrid Single Particle Integrated Trajectory (HYSPLIT) Model.29 One-degree latitude-longitude Global Data Assimilation System (GDAS1) meteorological data were obtained from the NOAA archives. Model boundaries were set at altitudes of 10 m and 10 000 m. Back trajectories were run for 120 h for each of 52 weekly periods where one or more fission products were detected in
■
RESULTS AND DISCUSSION Filter Samples. None of the 280 filter samples contain detectable fission products regardless of compositing method. Samples were rerun in regional groups, and then with all filters together, and no activity was detected. Therefore, it seemed likely that fission products were not primarily adsorbed to insoluble particles as originally expected. However, fission products could have been associated with insoluble particles that passed through the filters. Whole-Water Samples. 131I, 134Cs, and 137Cs were detected in the whole-water wet-deposition samples collected at 35 of 167 NADP sites. Table S-1 (Supporting Information) lists the numbers of water samples analyzed and associated detections of fission products by isotope. The minimum detectable activity (MDA) for these samples is estimated to range from 0.02 pCi to 0.18 pCi, representing 40- and 6 h count times, respectively. Longer count times were used for samples with lower activities. Spatial distributions of 131I, 134Cs, and 137Cs activities and wet-deposition values for NADP, 2577
dx.doi.org/10.1021/es203217u | Environ. Sci. Technol. 2012, 46, 2574−2582
Environmental Science & Technology
Article
Figure 3. NADP, RadNet, and UC Berkeley 134Cs activities (Bq/L), NADP-estimated 134Cs deposition values (Bq/m2), and interpolated precipitation depths (mm) obtained from National Weather Service Cooperative Observer Program (COOP) stations for the period March 15 to April 5, 2011.30.
Figure 4. NADP, RadNet, and UC Berkeley 137Cs activities (Bq/L), NADP-estimated 137Cs deposition values (Bq/m2), and interpolated precipitation depths (mm) obtained from National Weather Service Cooperative Observer Program (COOP) stations for the period March 15 to April 5, 2011.30.
2578
dx.doi.org/10.1021/es203217u | Environ. Sci. Technol. 2012, 46, 2574−2582
Environmental Science & Technology
Article
Table 1. NADP Results for 131I, 134Cs, and 137Cs Activities in Wet Deposition, Estimated Deposition (Flux) Values, and Other Measurements Associated with the Fukushima Dai-Ichi and Chernobyl Incidentsa deposition (Bq/m2)
activity (Bq/L) locations Fukushima Incident North America San Francisco area, CA North America (RadNet) Germany Chernobyl Incident New York City Area Greece
United Kingdom Northern England
Sweden Ireland/N. Ireland U.S., Oregon New Jersey New York City Idaho a
dates
131
3/15 to 4/5/2011 3/18 to 3/26/2011 3/15 to 4/5/2011
1.1−40 1.9−16 0.1−14
0.01−2.0 0.06−0.35 0.05−1.5
3/20 to 5/3/2011
0.03−0.43
