Abstract

1
Paper Number OS26A-07
A High Resolution Study of Particle Export Using
Thorium-234 in the N. Central Pacific and NW
Pacific as Part of the VERTIGO ProjectSteve
Pike1, John Andrews1, Tom Trull2 and Ken
Buesseler11Woods Hole Oceanographic
Institution, MS 25, Falmouth, MA 02543 United
States ( corresponding author)2University of
Tasmania, Institute of Antarctic and Southern
Ocean Studies, Hobart, Tas 7001 Australia
Abstract As part of the VERTIGO project (VERtical
Transport In the Global Ocean) we used
Thorium-234 (234Th) as a natural proxy for
particle export. As developed in recent years,
application of a small volume sampling method
allowed for relatively high resolution total
234Th sampling in both space and time, during 3
week occupations of two contrasting flux sites
off Hawaii (ALOHA) and in the NW Pacific (K2).
The higher vertical resolution allows us to
constrain not only particle export out of the
euphotic zone, but export and remineralization
processes below. We have made extensive use in
the VERTIGO project of different sediment traps
as well as large volume pumping systems for size
fractionated filtration, and thus can compare for
export calculations of C, N, bSi and PIC, the
ratios of these elements to particulate 234Th for
a wide variety of sinking and suspended
materials. Results to date indicate a rather
large difference in 234Th distributions at ALOHA
and K2. At ALOHA, the 234Th238U disequilibrium
is small, and hence 234Th fluxes on particles are
low, on the order of 300-500 dpm m-2 d-1.
Variability in space and time within a 200 km2
area is also small and hence a 1-D steady-state
model is a good approximation of the flux
conditions. The predicted fluxes are consistent
within errors with simultaneously measured 234Th
trap fluxes. As developed in prior studies, we
can apply the ratio of X/Th on particles to
convert from Th to other elemental fluxes, such
as C, N, bSi and PIC. At ALOHA C/Th on size
fractionated particles decreased with depth, and
increased with size for 1-10, 10-53 and gt53
micron pore sized filters. Interestingly, the
10-53 µm fraction was closest in C/Th ratio to
the sinking material caught in traps. At K2, the
234Th238U disequilibrium was much larger, with
total activities as low as 1 dpm L-1 within the
mixed layer. By sampling with up to 20-24 point
vertical resolution in the upper 300m, we can see
a regular "excess" 234Th feature at about
100-120m at the base of the subsurface
Chlorophyll maximum, and in many cases a small
disequilibrium between 100-250m. This deeper
feature may be an indication of repackaging of
suspended material into sinking particles. Fluxes
calculated from a 1-D SS model thus increase from
about 1700 to 2200 dpm m-2 d-1 between 100-300m,
consistent with sediment traps during our first
deployment. Trap fluxes drop off during the
cruise to values closer to 500-700 dpm m-2 d-1,
and the average 234Th disequilibrium drops
slightly, but there is significant spatial and
temporal variability at K2. Final chemical yields
from K2 will be needed to finalize this 234Th
data set, but early indications are that the
increase in 234Th over time will lead to a lower
predicted 234Th export using a non steady-state
approach to modeling the changing 234Th
activities over the 3 week observation period.
Water Column Activity Time/Space Variation
POC/Th Ratios in Traps and Pumps
In VERTIGO, we measured total 234Th activities on
more than 19 profiles at ALOHA and 25 profiles at
K2. This high resolution data set was made
possible by development of a new 4L method with
rapid processing and ship board analyses of 234Th
via beta counting (followed by 230Th yield
recovery corrections in the lab). Shown here are
time series 234Th profiles for a smaller subset
of central stations, which represent particle
source conditions during the experiments.
Thorium-234 activities are significantly lower at
K2, due to higher particle flux conditions which
remove the particle reactive natural radionuclide
234Th (t1/2 24.1 days) relative to its longer
lived and conservative parent, 238U (dotted line).
A common application of 234Th is its use as a
tracer of upper ocean carbon fluxes. In this
case, the 234Th derived from the water column
distributions (see Flux panels), is simply scaled
to the C/Th ratio of sinking particles to
estimate particulate carbon export. Shown here
is the C/Th ratio on sinking particles collected
by two types of sediment traps employed during
VERTIGO (see poster by Andrews et al.), and for
ALOHA, a comparison to size fractionated
particles collected using an in situ large volume
pumping system.
Measured and Calculated Fluxes
From the water column distribution of total
234Th, one can calculate 234Th export fluxes on
sinking particles. We have used the most simple
1-D steady state model to illustrate the range of
fluxes predicted vs. depth for 234Th, and compare
these fluxes to the average trap activities. At
both sites, the range in predicted fluxes is
significant, reflecting temporal and spatial
variability in particle export not caught in the
traps. Remember too that the 234Th flux is
quantified by the difference in total 234Th and
238U, and as this difference decreases, errors
increase in our ability to quantify 234Th export.
This variability may also reflect temporal and
physical impacts on the 234Th activity budget
that are not quantified in this simple 1-D SS
flux model. Overall, the fluxes measured by our
traps are consistent with the range of fluxes
measured at each site, and the overall decease in
flux between deployments early and late at K2.
This decrease with depth is evident during both
low and high flux conditions, and the absolute
value of this ratio is similar at two sites with
widely differing particle characteristics. When
compared to the filtered particles, an
interesting pattern at ALOHA (K2 data not yet
processed) is that the 1-10 and 10-53 µm size
classes are more similar to the sinking material
caught in the traps than the gt53 µm fraction.
This may be due to the few and rare zooplankton
caught on the larger screens, that are known to
carry significantly higher C/Th than detrital
aggregates, pellets and other sinking debris.
Such comparisons vs. size and sinking are
critical to a more accurate use of 234Th as proxy
for the flux of C and other elements in the ocean.
Thorium-234 data are shown as a color contour
time series plot for the central VERTIGO stations
and compared to evolving temperature, salinity
and fluorescence fields for ALOHA and K2. At
ALOHA, where the 234Th deficit is smaller, there
is a general trend towards lower 234Th, i.e. high
particle fluxes, during our VERTIGO study of
particle flux and remineralization in 2004. At
K2, overall activities are much lower and there
is an increase in 234Th over time during VERTIGO,
consistent with the large decrease in total
sediment trap fluxes for 234Th and other elements
seen during our occupation of this site (see Flux
panels here and Manganini et al. poster).
Carbon flux 234Th flux ? C/234Thsinking part.
Schematic of the 234Th flux approach. Three
scenarios are shown for differing conditions of
234Th238U disequilibria, 234Th flux, sinking
particle C/234Th and the impact on calculated C
flux. The magnitude of 234Th flux is
proportional to the 234Th238U activity ratio
(here lt1 in surface waters, where 234Th solid
line lt238U dotted line). In panel a, the 234Th
flux of 1000 and a sinking particle C/234Th ratio
of 1/4 results in a calculated POC flux of 250.
Panel b shows the impact of a doubling of the
C/234Th ratio for the same 234Th flux (C flux
doubles). Panel c shows how a 50 reduction in
Th flux for the same C/234Th ratio as in b
results in a decrease in C flux by 50. Units
are not needed in these examples, but are
commonly dpm m-2 d-1 for 234Th flux, mmol dpm-1
for C/234Th, and C flux in mmol m-2 d-1. One dpm
1/60th Bq.
Also important in quantifying flux variability
are not just changes with time, but spatial
variations in 234Th activity. Shown here are a
series of spatial maps of 234Th for all stations
averaged at selected depths. While overall 234Th
activities at ALOHA are higher than K2, there are
significant spatial differences in 234Th,
suggesting low, but variable fluxes. K2 shows
more coherence in these 234Th activity maps, but
lower 234Th overall, and a temporal trend that
hidden in these spatial maps (see Flux panels).
With continued improvements in 234Th methods, it
is now possible to obtain very highly resolved
vertical profiles of this particle flux tracer.
Two examples are shown here with 20 point
vertical resolution for total 234Th from K2, with
comparisons to CTD based sensors which indicate
layering of large particles (scatter), biomass
(flu), small particles (transmission), and
significant stratification of density and low
oxygen at relatively shallow depths. Most of the
234Th removal takes place within the mixed layer
down through the base of the fluorescence
maximum. Immediately below the mixed layer and
chlorophyll maximum, there is a 234Th excess
peak, which is indicative of shallow
remineralization.

• Calculating 234Th flux from 234Th activities
• d234Th/dt (238U - 234Th) l - PTh V
• where l decay rate
• PTh 234Th export flux
• V sum of advection diffusion
• low 234Th high flux
• ThgtU indicates remineralization
• need to consider non-steady state and physical
  transport

A conceptual view of the impact of various
biogeochemical processes on C/234Th ratios and
particle sizes. As thorium associates
principally with surface sorption sites and
organic carbon is dominated by pools internal to
cells, one might expect C/234Th ratios to
increase as particle size increases, with the
volume to surface area (VSA) ratios of spheres
representing the upper limit for the relationship
(all other cell/particle shapes have lower VSA
trends with size). Particle sizes in real marine
systems tend to increase as a result of complex
biological processes, however, including
aggregation of small, neutrally buoyant cells
into larger sinking particles and the generation
of fecal material. Rapid aggregation of small
particles alone without loss of mass would
probably yield no change in VSA ratios and hence
no change in C/234Th, while consumption of
particles by zooplankton would result in
preferential assimilation losses of carbon and
hence a decrease in C/234Th ratios in larger
fecal pellets. Processes that affect the Th side
of the ratio (Th speciation), are not likely to
be linked to particle size in a general way.
These include increases in dissolved and
particulate Th-binding ligands or sorption sites,
which would increase or decrease C/234Th ratios,
respectively.
Interestingly, in both profiles, there is a
sub-mixed layer deficit between 100-150 and
250-300m. 234Th trap fluxes increase between
150, 300 and 500m, consistent with a process at
depth that enhances particle export (perhaps
zooplankton driven- a topic of ongoing
investigation in VERTIGO). Both features would
be missed with traditional 234Th sampling. This
opens up the possibility of new applications of
234Th towards understanding particle sources and
sinks in association with physical and biological
stratified systems in the Twilight Zone.