Item_4_2005-09-SedSampling_Strategy

4.26 CEP/RMP Sediment Core Plan
Draft & Comments
CFWG Sept 2005
General Objectives (for CEP & RMP)
• Build understanding of ecosystem characteristics
and processes…
• To provide sufficient (better) basis for deciding
among possible management alternatives
Specific Objectives (for this Study)
• Estimate future loads from eroding buried
contaminants
• Estimate historic loadings of contaminants
(especially recent decades)
• Characterize contamination with depth to assess
current status and likely future changes
• Provide data for parameterization and evaluation
of the multi-box model
If ~20 Cores Budgeted:
• Random sampling
– Pros: aim to be representative of studied system
– Cons: muddy signal, may need many samples
before important factors ID’d
• Deterministic sampling
– Pros: build process with relatively few samples
– Cons: selected samples often not characteristic of
larger system (e.g. USGS depositional cores)
Sediment Core Sampling Strategy
• Hybrid Approach
– Some samples specifically to understand loading
history (deposition only zones)
– Remainder of sites to begin representative
characterization of Bay sediments
Initial Effort Timeline
• One cruise 1 year (not 2 years)
– Relatively few stations spaced widely apart
– Long lived isotopes, pollutants, ~1 yr storage not too
bad if stored properly
– Not enough samples (e.g. 10 in yr 1) for info to
change stratification midstream
Sampling Effort Distribution
• Use general Bay segmentation scheme
– Suisun, San Pablo, South, Lower South Bay
• Areas with bathymetric change history mapped
• Different sediment and pollutant loading quantities and sources
• Cores from each segment
– 1 depositional (wetland or deep Bay?)
– 1 erosional
– 2 more “characteristic” of that segment
• Depositional, neutral, or erosional
Modeling & Other Needs
• Hybrid approach seeks some of each
– Assumes near 0 possibility that any model can do without
verification of loading histories or process outcomes
• ? Is how much of each ?
– ~20 samples not a large number
– Numbers likely not definitive of loading or representativeness
– Collect excess samples of each type?
• Will we be able to get around to analyses soon enough?
Example Stratification
+
0
-
Distribution of Sites
Segment
Suisun Bay
San Pablo Bay
South Bay
Lower South Bay
Guadalupe
Total
Erosional Neutral Depositional Total
3
1
4
3
1
4
2
1
1
4
1
3
4
1
9
1
7
17
Future loads from erosion
• Unmixed buried sediments become mixed
• Pollutant profiles at depth < = > mixed layer?
– Introduced and averaged into mixed layer
– Widespread characterization preferred
• Rate of erosion
– Extrapolate recent history (Jaffe rates)
– Predicted from model (calibrated to Jaffe?)
Future Loads w/ no More Cores
• Assume a range of profiles from existing data?
– Pollutant profiles at depth < or = or > mixed layer
– How much higher or lower is reasonable?
– Trend is likely depth > or = surface
• USGS cores depth > surface
• Many others suggest depth ~ surface
– Guessing time scale of change much harder
Historic Contaminant Loadings
• Depositional cores reflect past conditions
• Potential limitations:
– Ambiguous chronology: multiple tracers?
– Natural or manmade disturbances: choose sites
carefully?
– Not representative of Bay: what does it reflect?
• Deep bay or wetland?
Historic Contaminant Loadings
• Deep bay cores
– Pros:likely less hot spot influence, more representative of
segment conditions
– Cons: integrate more processes = more degrees of freedom
(mixing w/ other seds, long response time), location history
uncertain
• Wetland cores
– Pros:respond more quickly to changes, reflective of local
sources (hotspot, effluent or tributary), better mapped, history
– Cons: more reflective of local sources, slow accretion
Alternatives Historic Loadings
• Assume Breivik close enough?
• Revise Breivik with upper and lower bounds?
– How low/high is reasonable?
• Monitor current loads with time series
– Currently @ Guadalupe river, Mallard Island
• Representative enough?
– Project trends backward for hindcast?
Contaminant Profiles w/ Depth
• System status not just in surface 5cm sediments
• Dominant sedimentation regimes in each
segment sampled ( - 0 + sedimentation)
• Spatially variable but representative in long term
– (must start somewhere)
• Alternatives (?)
– Assume concentration < = or > surface sediments
• Even larger uncertainties
Data for Model
(Multi box or other)
• Past & future loads, current system status are
needed to model
– (input parameters, initial or final conditions)
• Loads and status data needed independent of
model
– Relative priority somewhat model dependent
– Which is more important?
Comments on Plan to Date
•
•
•
•
•
•
•
Too expensive (much of CEP budget)
Underfunded for objectives
Too little of budget for interpretation/reporting
Any geologic cores for sediment transport perspective?
Why not Cs137 or mix of other tracers?
Can cores be collected and stored for later analysis?
Why wetland cores (not in model)?
Comments on Plan to Date
• Expect high % of useless cores
• Use screening analyses
• Need more sections per core for temporal trends
(fewer cores more sections)
Regional Board Wish List
• Some Central Bay sites
• Trade OC pesticides for other organics in some
samples?
– Selenium (for a background baseline?)
– PCDD/Fs
– Background PCB levels