In late 1989, the Republic of the Marshall Islands commissioned an ind
ependent radiological monitoring program to determine the degree of de
position and the geographical extent of atomic weapons test fallout ov
er its islands. The sampling and measurement of Cs-137 in vertical soi
l profiles has been an important part of that monitoring program. Over
200 profiles were collected in 5-cm increments to a depth of 30 cm. T
he sampling sites included locations at all 29 of the nation's atolls
and five separate reef islands. The nation is unique in that a rainfal
l gradient exists from south to north, the northern islands receiving
approximately one-third the annual precipitation of the southern islan
ds. Annual rainfall and soil particle size distribution was studied to
determine the inter-relationships with cesium permeation and areal in
ventory (Bq/m(2)). Relaxation lengths of cesium in the soil column wer
e calculated from measurements of Cs-137 activity in profile increment
s. Using a common model for a depth-dependent concentration gradient (
C(d)= C-o e(-ad)), the inverse of the rate of change of concentration
(i.e. 1/alpha) has been defined as the relaxation length of the profil
e. In this study, the relaxation length was calculated from the concen
tration data of the first three sample increments (i.e. 0-5, 5-10, 10-
15 cm) and from all six increments to a total depth of 30 cm. Relaxati
on lengths with r(2) values greater than or equal to 0.90 were used fo
r data analysis. Typical values for relaxation lengths in the drier no
rthern atolls are 7-12 cm, though our observations included values bet
ween -800 and 3500 cm. Marshall Islands soils are primarily calcium ca
rbonate, and with the absence of clay material, Cs-137 permeation was
expected to increase with increasing rainfall and increasing particle
size. Our findings showed that Cs-137 permeation was highly variable w
ith particle size, though relaxation length was significantly correlat
ed with annual rainfall. Areal inventory was determined to be signific
antly correlated with the 0-5 cm Cs-137 soil concentration. The 0-30 c
m fitted relaxation length was also determined to give a good estimate
of the areal inventory of Cs-137 in the soil to a depth of 30 cm rega
rdless of the rate of change in concentration with depth.