Artificial Radionuclides Database in the Pacific Ocean: HAM Database

The database “Historical Artificial Radionuclides in the Pacific Ocean and its Marginal Seas”, or HAM database, has been created. The database includes Sr, Cs, and Pu concentration data from the seawater of the Pacific Ocean and its marginal seas with some measurements from the sea surface to the bottom. The data in the HAM database were collected from about 90 literature citations, which include published papers; annual reports by the Hydrographic Department, Maritime Safety Agency, Japan; and unpublished data provided by individuals. The data of concentrations of Sr, Cs, and Pu have been accumulating since 1957—1998. The present HAM database includes 7737 records for Cs concentration data, 3972 records for Sr concentration data, and 2666 records for Pu concentration data. The spatial variation of sampling stations in the HAM database is heterogeneous, namely, more than 80% of the data for each radionuclide is from the Pacific Ocean and the Sea of Japan, while a relatively small portion of data is from the South Pacific. This HAM database will allow us to use these radionuclides as significant chemical tracers for oceanographic study as well as the assessment of environmental affects of anthropogenic radionuclides for these 5 decades. Furthermore, these radionuclides can be used to verify the oceanic general circulation models in the time scale of several decades.


INTRODUCTION
Chemical tracers are useful tools to understand the physical and biogeochemical processes in the ocean. Tritium, CFCs, and others have been used as transient tracers of physical processes, such as mixing and advection, in the ocean. The distributions of chemical tracers have been used for validation of global oceanic circulation models.
Artificial radionuclides ( 90 Sr, 137 Cs, and 239,240 Pu) that have been injected into the ocean by global fallout, released from the nuclear processing plants and others, might be also significant transient tracers.
Although the concentrations of these radionuclides in seawater have been determined for the assessment OBJECTIVES One of the major objectives of the HAM database is to reconstruct the history of artificial radionuclides in the oceans. Since our interest is mainly in the Pacific Ocean and its marginal seas, our effort to collect data sources was concentrated in these sea areas. We expect that the artificial radionuclides are good transient tracers of oceanographic studies in the Pacific Ocean and its marginal seas. 137 Cs, which is a soluble radionuclide, is a good tracer for water mass movement. 239,240 Pu, which is a typical particle reactive radionuclide, could also be used for biogeochemical studies of trace metals, including radionuclides, in the ocean. The HAM database can be applied to validate the results of the 3D simulations using ocean global circulation models.
Some results of the studies using the HAM database are already published. A typical example is a study on the formation of deep waters in the Sea of Japan. It provided new information that the formation of deep waters occurred in the 1990s based on the analysis of the 137 Cs concentration data in the database ( Fig. 2 in [3]). The temporal variation of the 137 Cs concentration in the surface water in the North Pacific during these 4 decades was examined by Hirose and Aoyama (Figs. 1, 2, 3, 6, and 12 in [4]). Aoyama and Hirose also examined the temporal variation of the 137 Cs inventory in the North Pacific, then, increases of the 137 Cs inventory at the lower latitude of 10-20° N in the North Pacific were found in the 1970s and 1980s. It is concluded that surface and subsurface southward transports were considered as the source of this increasing 137 Cs inventory (Figs. 3 and 4 in [5]). In the studies of temporal variation of 137 Cs and 239,240 Pu in the Pacific Ocean using an oceanic general circulation model, the HAM database was also used to verify the results of the model calculation by Tsumune et al. [6,7].

Sources of the Data in this Database
We collected about 90 literature citations to construct the HAM database during these 7 years. As shown in Table 1, the sources of data are as follows:

Handling of Original Data in the HAM Database
All of the data in the publications/reports are treated as follows: a.
Input original data in original unit with data originator name. b.
The values in the original unit are converted to the values in the SI unit, Bq m -3 . c.
Create Throughout the handling of the data, the data originator name, the name of the reference literature, and original values with original unit were kept. When counting errors were reported in the original literature, those were also kept throughout the handling. To identify the original literature, a unique code, of which field name in the database is MRIREFNO, is given for each literature and added to each record. For each profile, a unique number, of which field name is profileID, is also given. A code to identify that the data are in a certain profile, of which field name is PS, is given as P. A code S at the field PS is given for the concentration data of the surface water only without deeper layer sampling. Keeping these original and created data throughout the process allows us to maintain the database more easily. Users of the database can also see the original values and unit together with the values in SI by the database creator.

Geographical Distribution
The total numbers of the data in the HAM database for the three major radionuclides in each sea area are summarized in Table 2. In the database, sea area code is given in each data. The geographical distributions for each decade from the 1950s to the 1990s are shown in Figs. 1 to 5, respectively. Since our effort was concentrated in the Pacific Ocean and its marginal seas, more than 80% of the data for each radionuclide are for the Pacific Ocean and the Sea of Japan in the HAM database as shown in Table 2. In the 1950s, a few 137 Cs and 90 Sr concentration data at the western North Pacific near Japan were available as shown in Fig. 1. The 239,240 Pu concentration data did not exist in this decade in HAM database. The number of 137 Cs concentration data increased abruptly in the North Pacific in the 1960s, especially in 1966-1968 as shown in Fig. 6. The station locations of 137 Cs concentration data covered well over the Pacific Ocean, while 90 Sr concentration data and 239,240 Pu concentration data showed still less and sparsely, respectively, in the 1960s (Fig. 2). In the 1970s (Fig. 3), the number of 137 Cs concentration data decreased, however, the station locations of 137 Cs concentration data covered well over the Pacific Ocean. The station locations of 90 Sr concentration data and 239,240 Pu concentration data covered the North Pacific Ocean. The geographical distributions of 137 Cs concentration data and 90 Sr concentration data in the 1980s were concentrated at the western North Pacific, while that of 239,240 Pu concentration data covered over the North Pacific (Fig. 4). No data area at the central and the eastern North Pacific for the three radionuclides was seen in the 1990s (Fig. 5).

Temporal Variation of the Numbers for the Three Radionuclides
The temporal variation of the numbers of the data for the three radionuclides in each year is shown in Fig.  6. 137 Cs and 90 Sr concentration data were available since 1957, while 239,240 Pu data were available since 1964. A maximum number of 1268 for the 137 Cs concentration data was in 1967, while the numbers of 90 Sr and 239,240 Pu concentration data were 83 and 15 in the same year, respectively. This tendency to have so much 137 Cs concentration data and less 90 Sr and plutonium concentration data was observed during the 1960s. The numbers of the data for the three radionuclides became the same due to the GEOSECS project, in which the analysis for the three radionuclides was done simultaneously in the early 1970s. The numbers of the data for each radionuclide were around 100 per year throughout the 1980s and the 1990s.

The Number of the Data for the Surface Water and the Profiles
The number of the data for surface water and profiles in each of the oceans and marginal seas are summarized in Table 3 for each radionuclide. For the 137 Cs concentration data, 55.7% of the total amount of 7737 data is for surface data, then the remaining 44.3% is for the profiles. The numbers of the profile for the 137 Cs is about 430. For the 90 Sr concentration data, 42.7% of the total amount of 3972 is for the surface and 57.3% is for the profiles. The numbers of the profile for the 90 Sr is about 320. For the 239,240 Pu concentration data, 23.6% of the total amount of 2666 is for the surface and 76.4% is for the profile. The numbers of the profile for the 239,240 Pu is about 240.

Counting Error of the Data
The number of the data associated with counting error for the 137 Cs concentration data is 7164, which is 92.6% of the total amount of 7737 in this database. For the 90 Sr concentration data and the 239,240 Pu concentration data, 98.0% of the total amount of 3972 and 96.0% of the total amount of 2666 in this database are reported with counting error.   For the 137 Cs concentration data, the magnitude of the counting error varied within 30% in general as shown in Fig. 7(a). Of the total amount of 7164, 208 (2.9%) showed larger counting error exceeding 100%. Those larger counting errors were seen for the data in deeper layers [24,25,29,30,36,39,47,48,75,78,79,81,83,85,91] where the 137 Cs concentrations were generally very low. For the 90 Sr concentration data, the magnitude of the counting error varied within 30% in general as shown in Fig.  7(b). Of the total amount of 3892, 205 (5.3%) show larger counting error, which is exceeding 100%. Those larger counting errors were seen for the data in deeper layers [29,30,39,45,47,48,73,74,77,78,79,84,85] as well as the 137 Cs concentration data. For the 239,240 Pu concentration data, the magnitude of the counting error varied between 6 and 30% in general as shown in Fig. 7(c). Of the total amount of 2562, 167 (6.5%) show larger counting error, which is exceeding 100%. Those larger counting errors were seen both in shallower layers and deeper layers [47,48,54,60] .

DATABASE DISTRIBUTION AND CITATION OF THIS DATABASE
This HAM database is freely available on request. When results or illustrations obtained using the HAM database are presented in papers, their authors should cite this paper together with a version of the HAM database used.
The description in this paper on the HAM database is based on the version U as of 29 Dec. 2001.

CONCLUSIONS
The database regarding "Historical Artificial Radionuclides in the Pacific Ocean and its Marginal Seas", HAM database, has been created. The data in the HAM database have been collected from about 90 literature citations, which include published papers/reports; annual reports of the Hydrographic Department, Maritime Safety Agency, Japan; and unpublished data provided by individuals. The data of the concentrations of 90 Sr, 137 Cs, and 239,240 Pu have been accumulating since 1957-1998. The present database includes 7737 records for 137 Cs concentration data, 3972 records for 90 Sr concentration data, and 2666 records for 239,240 Pu concentration data. The spatial variation of sampling stations in the HAM database is heterogeneous, namely, more than 80% of the data for each radionuclide is from the Pacific Ocean and the Sea of Japan, while a relatively small amount of data is from the South Pacific. This HAM database will allow us to use these radionuclides as significant chemical tracers for oceanographic studies as well as the assessment of environmental affects of anthropogenic radionuclides for these 5 decades. Furthermore, these radionuclides can be used to verify the oceanic general circulation models in the time scale of several decades.