RMP Coring Plans 2010 and Onward

RMP CFWG Meeting June 2009

Core- What Is It Good For?

• Bay pollutant inventory– erosional time bombs?

• Model validation– Conceptual &/or mechanistic

• Model development– Empirical, mechanistic, hybrid– Can recalibrate, but better up front

Approach RMP/CEP 2006

1. Select RMP S&T (random) and continuous (undiked) wetland sites

2. Vibracore in Bay, Livingstone (piston) core in wetlands/watershed

3. Freeze long core sections in field

4. Saw core into 2.5cm sections

5. Analyze sections at ~15 year intervalsUsing prelim. radiodating, literature estimates

RMP/CEP 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

• +1 watershed site– Alviso Marina

Lessons RMP/CEP 2006

1. Waiting for preliminary radiodating slowed study

2. 17 cores x 10 sections/core = 170 samples

3. Created backlog at RMP laboratories (3+ years equivalent of S&T samples)

4. Total cost $300k+, hard to reduce, especially analytical costs

(1 site = 10 samples x X analytes)

Conceptual Model

• Sedimentation (from isotopes, bathymetric history)– Similar in segment (shared water, sediment)– But mesoscale differences (trib/shore

proximity, etc)

• Pollutant distribution function of– Sedimentation history– Local land use/ loading

Bay Hg Results

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

LSB001LSB002

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SB001SB002

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

CB001CB002CB006

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SPB001SPB002

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SUB001SUB002

< 1960

< 1960

< 1960< 1960 < 1960 < 1960

< 1960< 1960

Conceptual Model Fits

• Radiodating fits bathymetric history– North Bay erosive (137Cs, 210Pb near surface)– Central, South Bay ~neutral, or erosive– Lower South Bay depositional

• Contamination fits sediment history– Top core sections ~ RMP surface sediments– Lower contamination in deepest sections

• pre industrial background

– Contaminants elevated in industrial period• Metals ~uniform downcore, PCBs higher nearer surface

Wetland Hg Results

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

LSB AlvisoLSB Coyote

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SB Greco

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

CB Damon

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SPB Wildcat

CONC (mg/Kg)

0 5

DE

PT

H (

cm)

0

100

SUB Edith

< 1960

< 1960

< 1960< 1960< 1960

Wetland Results

• Radiodating fits sea level rise– All areas net depositional (2-3mm/year)– Lower South Bay subsiding, higher deposition

• Contamination fits sediment history– Top core sections ~ RMP surface sediments– Usually lower contamination in deepest sections

• pre industrial background

– Contaminants elevated in industrial period• Sharper/higher peaks than in Bay cores

• Watershed (Alviso) site ambiguous– Rapid deposition, but where is Hg peak?

Need Cores?

• Better than before, but enough?– N = 11 from RMP/CEP + 2-5 from USGS

• N depends on which analytes

– Maybe OK for Baywide scale but not enough for segment specific modeling (N=2)

• Time frame needed– Before models need new data

• Resolution needed

Option 1: Repeat 2006 Effort

1. 10-11 RMP S&T (random) sites, 5 wetland, all Bay segments

2. Gravity/hammer core in Bay, Livingstone (piston) core in wetland

3. Freeze/saw subsampling 2.5cm sections4. Analyze up to 10 sections at ~10cm

(skipping) intervals (150-165 samples)PCBs, PBDEs, metals, TOC, grainsize

5. Total cost $350k+ ($300k in 2006)

Option 2: Incremental Efforts

1. 2 RMP S&T (random) sites (optionally +1 wetland?) in one Bay segment per year

2. Gravity/hammer core in Bay, Livingstone (piston) core in wetland

3. Freeze/saw subsampling into 2.5cm sections

4. Analyze up to 10 sections at ~10cm (skipping) intervals

PCBs, PBDEs, metals, TOC, grainsize

5. Cost ~$50k for Bay cores, ~$75 w/ wetland

Budget 2010, Two Bay Cores

Sample Collection & Processing $5,500

Laboratory Analysis $34,250

Radiodating 6600

PCBs, PBDEs 18750

Metals, Hg 6400

TOC, grainsize 2500

Proj Mgmnt, QC & Reporting $8,000

Total $47,750

+/- Incremental Approach

• Few sites, more often (e.g. 2 per segment, yearly)+ Better workload for labs (20+ vs 100+ samples at

once)

+ Costs better spread for RMP

+ Can get info on segments w/ greatest data needs first- Long time to get full Baywide set

• Other details can be decided depending on program needs

Back to Basic Questions

• Do we need more cores?– Probably, especially if we plan sub-segment scale

models

• All at once, or a few at a time?– A few at a time is easier for many logistical and

budget reasons

• When, where, how many?– Best early/before models finished, random/

representative of areas modeled, # samples depending on model Q’s

Nitty Gritty Details

• Example design to follow if desired

Siting Approach 2010+

• Use new random RMP S&T sites+ Continue to build larger scale spatial data

± May get stations w/ similar characteristics

- Low odds for special sites of interest (hot spots, watershed deltas, dep/erosional

Sampling Approach 2010+

• Unpowered (push/gravity/hammer) cores+ Low equipment needs/cost- Cores >1m deep may be difficult

Sectioning Approach 2010+

• Field freezing, saw sectioning+ Easy sectioning post sampling (solid core) + No hold time issues (within ~1 year)- Freezing causes core distortion- Clean (e.g. Teflon) saw difficult/impossible,

but the devil we know

• (Alternatively) Field or lab extrusion+ No freezing needed (~ambient or cooler

chilling)± Cleanliness unknown, depends on

execution (staff, location, equipment)- Need extruding equipment/attachments

Analysis Approach 2010+

• Select set interval subsamples+ Know which sections to send to labs a priori+ Results from radiodating/chem labs sooner- Less flexibility on section spacing (longer or shorter

cores, fast/slow deposition areas)± (alternatively wait for radiodating, irregular spacing)

• Skip sections+ No need to composite 1 sample = 1 section, - May miss narrow peaks (how likely?)± (alternatively composite sections)

Analysis Approach 2010+

• Analyte selection- highest needs for particle associated persistent pollutants of concern– PCBs– Hg– PBDEs (top 5 sections each core)– ???

• Add geologic /anthropogenic factors– ICP-MS trace elements– TOC, grainsize

Approach 2010+ Summary

1. 2 RMP S&T (random) sites (optionally +1 wetland?) in one Bay segment per year

2. Gravity/hammer core in Bay, Livingstone (piston) core in wetland

3. Freeze/saw subsampling into 2.5cm sections

4. Analyze up to 10 sections at ~10cm (skipping) intervals

PCBs, PBDEs, metals, TOC, grainsize

5. Cost ~$50k for Bay cores, ~$75 w/ wetland

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

– max ~1960

• Pb decay – half life 22 yrs– Decay/ mixing dilution

can look similar

• If Cs & Pb similar– likely mixing dilution

LSB001: Fast accumulation

~150cm to 1960 ~60cm to 1960

LSB002: Fast accumulation

~130cm to 1960 ~30cm to 1960

LSB Wetland Deposition

~80cm to 1960

Hg Analyses

• ICP-MS HF extract (CCSF) vs CVAFS aqua regia (MLML)CCSF vs. MLML Hg Concentration

CCSF

-0.5 0.0 0.5 1.0 1.5 2.0

MLM

L

-0.5

0.0

0.5

1.0

1.5

2.0

CCSF vs MLML Regression PlotOne:One

Lower South Bay 1960 = 30-60cm bay, 80cm C.Creek

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

Conaway 2004 vs Current

5nmol/g ~ 1mg/kg

Lower South Bay PCBs

• PCB in bay cores max subsurface– LSB001 max @40cm

(60cm =1960)– LSB002 max @30cm

(30cm = 1960)

SB001: Continuous Erosion?

~0cm to 1960 ~15cm to 1960

Core ID: SB001, X: 564867.30345800000, Y: 4163027.619000000001858 depth: -1221898 depth: -1231931 depth: -1571956 depth: -1241983 depth: -1462005 depth: -1602006 depth: -161Reconstructed horizons: 0

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 ~uniform in wetland to 55cm =

1930s? (1960s Cs penetration to 30cm)

South Bay1960 = 12-15cm bay, 30cm wetland

Se Normalized to Percent Fines (mg/Kg)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

Greco Island SB001S SB002S RMP 5 Year Average

Hg Normalized to Percent Fines (mg/Kg)

0.0 0.2 0.4 0.6 0.8

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

Greco Island SB001S SB002S RMP 5 Year Average

Cu Normalized to Percent Fines (mg/Kg)

0 20 40 60 80 100

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

Greco Island SB001S SB002S RMP 5 Year Average

• PCB in bay max near surface– SB001 (continuous

erosion) at top ~5cm – SB002 (no change since

1950s) ~10cm

South Bay1960 = 12-15cm bay, 30cm wetland

CB001: No Change ~1940s

~0cm to 1960 ~5cm to 1960

CB002: Erosion to ~1920s

~0cm to 1960 ~20cm to 1960

CB006: Continuous Erosion

~0cm to 1960 ~12cm to 1960

Core ID: CB006A, X: 566290.21976900000, Y: 4174242.035890000001858 depth: -1061898 depth: -1341931 depth: -1321956 depth: -1831983 depth: -220Reconstructed horizons: 0

Central Bay Metals

• Bay downcore concentrations smaller range than in SB/LSB

• No dating for wetland cores yet– ~20cm subsurface max for Hg, Se, Cu in

wetland, – Similarly high conc for Se, Cu @ surface,

60cm

Central Bay1960 = 5-20cm bay, ?? wetland

Se Normalized to Percent Fines (mg/Kg)

0.0 0.5 1.0 1.5 2.0 2.5

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

CB - CB001SCB - CB002SCB - CB006SACB - Damon SloughRMP 5 yr average

Hg Normalized to Percent Fines (mg/Kg)

0 2 4 6 8

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

CB - CB001SCB - CB002SCB - CB006SACB - Damon SloughRMP 5 yr average

Cu Normalized to Percent Fines (mg/Kg)

0 20 40 60 80 100 120 140 160 180

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

CB - CB001SCB - CB002SCB - CB006SACB - Damon SloughRMP 5 yr average

• PCB in bay max near surface– CB sites no change or

eroding

PCBs Normalized to Percent Fines (mg/Kg)

0.01 0.1 1 10 100

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

CB - CB001SCB - CB002SCB - CB006SACB - Damon SloughRMP 5 yr average

Central Bay PCBs1960 = 5-20cm bay, ?? wetland

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

San Pablo Bay1960= 2-5cm bay, ?? wetland

• PCB in bay max near surface– SPB sites eroding, no

reservoir

PCBs Normalized to Percent Fines (mg/Kg)

0.01 0.1 1 10

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

SPB - SPB001SSPB - SPB002SSPB - Wildcat MarshRMP 5 yr Average

San Pablo Bay PCBs1960= 2-5cm bay, ?? wetland

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

Suisun Bay1960 = 2-80cm?! bay, ?? wetland

Se Normalized to Percent Fines (mg/Kg)

0 2 4 6 8 10D

epth

(cm

)

0

20

40

60

80

100

120

140

160

SU - Point EdithSU - SU001SASU - SU002SARMP 5 yr Average

Hg Normalized to Percent Fines (mg/Kg)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

SU - Point EdithSU - SU001SASU - SU002SARMP 5 yr Average

Cu Normalized to Percent Fines (mg/Kg)

0 200 400 600 800 1000 1200

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

SU - Point EdithSU - SU001SASU - SU002SARMP 5 yr Average

• PCB signal mixed– Bay sites erosional

according to bathymetry– SU002 mixed history

• (PCBs > RMP surface seds at 100cm)

Suisun Bay PCBs1960 = 2-80cm?! bay, ?? wetland

Hg Normalized to Percent Fines (mg/Kg)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Dep

th (

cm)

0

20

40

60

80

100

120

140

160

SU - Point EdithSU - SU001SASU - SU002SARMP 5 yr Average

Results

• Sedimentation in segment ~similar (bathymetry, isotopes) – Suisun, San Pablo eroding– Central, South neutral/eroding– Lower South accreting

• Erosional areas Cs penetration = mixing?– Pollutant penetration by same mechanism

Results

• Sites within segments similar (from bathymetry and radiodates)– Suisun, San Pablo eroding– Central, South neutral/eroding– Lower South accreting

• Wetland cores indicate loading history– Subsurface max in wetlands everywhere– Layer often near surface (1950s?)

Summary

• Wetland pollutants track loading history– Metals match SB loading history– PCBs max ~1970s

• Bay pollutants more complex– Some sites well mixed, or eroded– Some sites not well predicted (esp SU002)

• Bathymetry, Cs, Hg mismatch • Multiple deposition & erosion w/ seds from different

watersheds

Implications?

• Few ticking time bombs in Bay– So far, so good (2 of [90 + 11])– Largely WYSIWYG

• Surface ~ deeper for metals• Surface >~ deeper for PCBs

• Wetland cores capture historical pulses

• Historical loads mostly eroded, or already dispersed in Bay

Next Steps

• Get rest of data (radiodating, organics)• More analytes (dioxins?)• More normalization? (TOC in wetlands)• Modeling accretion vs mixing for radioisotopes

(USC)• Understand discrepancies among data

– Complex histories

• Develop strategy for future coring• Partial reporting (Bay core metals) for Pulse

– Full reporting 2009 Q3?