SF Bay Hg Coordination Meeting Feb 2009 Hg in SF Bay Cores SF Bay Hg Coordination Meeting Feb 2009

Background USGS (1990) coring Why Recore? 90+ sites screened only 2 depositional sites fully characterized Subsurface max @ 2 sites Pollutant reservoir? Why Recore? 2 of 90 sites probably not representative Old cores unusable for chem analyses

Core- What Is It Good For? (absolutely something) Bay pollutant inventory Erosional time bombs? Model validation Conceptual &/or mechanistic Model development Empirical, mechanistic, hybrid

Distribution of Sites (Bay) Representative inventory, sedimentation 3 sites Central Bay, 2 sites each other segments Preference to RMP repeat stations

Distribution of Sites (Wetland) Loading history Depositional zones 1 site each segment Pt Edith Martinez Wildcat Richmond Damon Sl. Oakland Greco Island Coyote Creek Alviso Marina

Conceptual Model Sedimentation (from isotopes, bathymetric history) Similar in segment (shared water, sediment) But mesoscale differences (trib/shore proximity, etc) Pollutant (Hg) distribution function of Sedimentation history Local land use/ loading

Dating: Bathymetric History (USGS Bruce Jaffe) Sum bathymetric changes between surveys + deposition – erosion Some sites depositional & erosional different periods

Dating: Isotopes (USC Hammond) Cs in A-bomb Pb decay Cs & Pb similar max ~1960 Pb decay half life 22 yrs Decay/ mixing dilution can look similar Cs & Pb similar likely mixing dilution

Results Sites within segments similar (from bathymetry and radiodates) Suisun, San Pablo eroding Central, South neutral/eroding Lower South accreting Wetland Hg indicates loading history Subsurface max in wetlands everywhere Layer often near surface (1950s?)

Lower South / South Bay 1960 = 30-60cm bay, 80cm wetland Lower South / South Bay 1960 = 30-60cm bay, 80cm wetland 12-15cm bay, 30cm wetland 1960 1960 1960 1960

Central Bay 1960 = 5-20cm bay, ?? wetland

San Pablo & Suisun Bay 1960= 2-5cm bay, ??wetland 2-80cm?! bay, ??wetland

Results Bay core Hg often ~uniform, or complex RMP segment avg ~ core top section Weak/no subsurface max in Bay sites Mixing, erosion (or ~constant loads)? All could give ~uniform profile Some sites not well predicted (esp SU002) Bathymetry, Cs, Hg mismatch Multiple deposition & erosion w/ seds from different watersheds

Implications? Few Hg ticking time bombs in Bay So far, so good (2 of [90 + 11]) Largely WYSIWYG (surface ~ middle) Wetland cores capture historical pulses Historical Hg loads mostly eroded, dispersed in Bay

Next Steps Finish wetland radiodating More normalization? (high TOC in wetlands) Model accretion vs mixing for isotopes (USC) Understand data discrepancies Analytical variation, spatial/temporal differences Implications for other work More coring, modeling Partial report 2009 Q1, Pulse of Estuary All done 2009 Q3?

SF Bay Hg Coordiantion Meeting Feb 2009 (RMP Annual Meeting Oct 2008) SF Bay MeHg Mass Budget SF Bay Hg Coordiantion Meeting Feb 2009 (RMP Annual Meeting Oct 2008)

WWMMBD? What Would the MeHg Mass Budget Do? Track MeHg specifically MeHg <1% of totHg, poor MeHg:totHg correlation Synthesize- do Bay data make sense given… Loading, production, degradation, sed-water exchange, etc. Quantitative dimension to MeHg conceptual model ID key factors for MeHg fate Refined models feasibility/needs E.g. temporal & spatial detail What it won’t/can’t do Turn water to wine (not J) Identify “hot” spot impacts (1 box) Predict long term (no Hg linkage) Synth knowledge. This essentially serves as a quantitative conceptual model of MeHg processes, helping us identify key factors By developing an understanding how the model responds compared to the actual system, we can determine the feasibility… WWMMBD?

MeHg 1 Box Model Many simple assumptions (from PCB 1 box) One water compartment One sediment compartment (10cm mixed layer) Daily time step Annually uniform loads (no seasonality) Uniform mixing Equilibrium partitioning Long term steady state Simplifications worked for PCBs, PBDEs Can it work for MeHg?

Does It Work? - Reasonably Base case = average initial concentrations (from RMP monitoring) loading/process parameter values At steady state Sediment mass ~ Water mass lower Fast! <1 month

Base Case Run Mass (inventory) vs daily flux/degrade/produce Water Mass External load>, Net sediment to water exchange = Degradation>, GG outflow, >> bio-uptake,volatilization Total (Water+Sediment) Production ~balances degradation >> all other processes * Flux box within ~2x: ~.014 kg/d (Choe et al, N Bay) Mass in water (kg) 0.38 kg Ext load 0.024 kg/d Sed to water 0.0064 Outflow 0.023 Water Degrade 0.0075 Fish uptake .0001 Volatilization <0.0001 Mass in Sediment 31.0 Methylate 1.82 Sed Degrade 1.80 Sed to Water Burial 0.0074 If you’ll notice, the daily meth and degradation rates are only about 1 order of magnitude < than the inventory, which suggests That the system will respond very quickly changes in these parameters.

Parameter Sensitivity (Mass/Input) at ±3x default input Input Parameter Sediment Water Sediment meth rate 99.3% 66.0% Sediment demeth rate -96.5% -64.1% SSC -0.8% 49.3% External load 0.6% 31.1% Long term net outflow -0.5% -23.5% Tidal flushing ratio -0.4% -18.6% Water column Kd 0.4% -21.8% Particle settling rate -0.3% 16.0% Sediment burial rate -0.2% -11.0% Water demeth rate -9.7% Ocean MeHg conc 0.1% 3.2% Sed/water transfer V 0.0%

WDMMBD? What Did the MeHg Mass Budget Do? Show Bay data reasonable Base case ~T0 state- near “right” Baywide, but non-unique solution (e.g. offsetting errors?) Feasibility/needs of refined model(s) 1 box driven by steady state/equilibrium Basis for more detailed (time/space) model? Much higher data needs Key factors affecting mid term MeHg fate External loads small/medium effect on Baywide scale Very sensitive to de/meth rates Even with this very crude 1 box model we are able to see that .. Ext loads have little effect, and the system is very sensitive to meth/demeth rates, So naturally that leads us to questions of what management options we might explore given that…

Management Strategy – Dr. Evil Acquire $1 Million Option A- Control meth: Sterilize Bay w/ thermonuclear device Option B- Control Demeth: Equip sharks w/ frickin UV lasers If we left the strategy to Dr. Evil, we might end up with solutions like these

Management Strategy -RMP Option C- RMP Mercury Strategy: Where biota affected (food web entry) ID disproportionate (high leverage) pathways ID intervention opportunities THEN act (e.g.holding ponds, aeration, dredging, nutrient reductions, sed amendments, etc) Monitor & model management effectiveness “adaptive management” (likely cost > $1 million) But fortunately cooler heads prevail

Acknowledgements Too many to list… “If I have seen further it is by standing on ye shoulders of Giants” – Sir Isaac Newton HGH? (not Hg) researcher Although like any model it can be incrementally improved as we learn more

LSB Metals Downcore concentrations noisy Coyote Creek Hg max > Alviso! Coyote Hg max @ 1960s depth (80cm) Coyote Cu max @ 40cm = 1980s? ~max Cu discharge late 1970s (Palo Alto) ~surface sediment Cu USGS long term data

SB001: Continuous Erosion? Core ID: SB001, X: 564867.30345800000, Y: 4163027.61900000000 1858 depth: -122 1898 depth: -123 1931 depth: -157 1956 depth: -124 1983 depth: -146 2005 depth: -160 2006 depth: -161 Reconstructed horizons: 0 ~0cm to 1960 ~15cm to 1960

SB002: No Change ~1950s ~0cm to 1960 ~12cm to 1960

SB Wetland Deposition ~30cm to 1960

South Bay Metals Downcore concentrations noisy Cu max @ Greco Island similar to Coyote, but into 1960s zone. Greco Hg max ~constant in wetland to 55cm (1960s Cs penetration to 30cm) = 1930s?

CB001: No Change ~1940s ~0cm to 1960 ~5cm to 1960

CB002: Erosion to ~1920s ~0cm to 1960 ~20cm to 1960

CB006: Continuous Erosion Core ID: CB006A, X: 566290.21976900000, Y: 4174242.03589000000 1858 depth: -106 1898 depth: -134 1931 depth: -132 1956 depth: -183 1983 depth: -220 Reconstructed horizons: 0 ~0cm to 1960 ~12cm to 1960

Central Bay Metals Bay downcore concentrations smaller range than in SB/LSB No dating for wetland cores yet ~20cm surbsurface max for Hg, Se, Cu in wetland, Similarly high conc for Se, Cu @ surface, 60cm

SPB001: Erosion to ~1920s ~0cm to 1960 ~5cm to 1960

SPB002: Erosion to ~1880s ~0cm to 1960 ~2cm to 1960

San Pablo Metals ~20cm surbsurface max for Hg, Se, Cu in wetland No dating for wetland cores yet No secondary metal peaks Deeper concentrations fairly constant

SU001: Erosion to ~1910s ~0cm to 1960 ~2cm to 1960

SU002: Erosion to ~1890s ~0cm to 1960 ~80cm to 1960?!!

Suisun Metals Hg highly variable @ Pt Edith and SU002 No dating for wetland cores yet SU002 max concentrations in top section Hg, Se, Cu, subsurface spikes as well