Scenarios of geomorphic change

1
Scenarios of geomorphic change in Suisun Bay
1867-1887, and 2030
Neil K. Ganju University of California,
Davis U.S. Geological Survey, California Water
Science Center
Background courtesy of Plymouth Marine Lab
2
Why geomorphic modeling?

• Contaminants in the sediment bed
• Tidal flat and tidal marsh loss
• Sea-level rise

Hornberger et al., 1999
3
Motivation for geomorphic modeling

• Bathymetric change in Suisun Bay, California
  Cappiella et al., 1999
• Affected by transport of hydraulic mining debris
  
• Rapid deposition followed by erosion
• Rare historical data!

4
Historical sediment loads
5
Historical sediment loads
Black linedaily, red line10-y running mean
6
Prior efforts in geomorphic modelingcalibration
to stage, salinity, SSC

• Adequate for tidal-timescale simulations
• Not adequate for decadal-timescale simulations
  small errors grow to confound bathymetric
  prediction
• Need to calibrate to same type of data that you
  are trying to model

25 y simulation performed for SFO runway
expansion model calibrated to stage, salinity,
SSC Courtesy of URS Corporation
7
Feedback between process timescales
8
(No Transcript)
9
Hydrodynamic/sediment transport model

• Regional Ocean Modeling System (ROMS) v. 3.0
• Supported by Rutgers, UCLA, USGS
• Open-source, community sediment transport model
• Solves Reynolds-averaged Navier-Stokes equations
  in separate 2D and 3D modes
  (mode-splitting)
• Too many configuration options to mention

10

• Talk outline
• Background
• Tidal-timescale modeling ETM
• Annual-timescale modeling sediment fluxes
• Decadal-timescale modeling bathymetric change
• Future scenarios geomorphic change

11
Tidal-timescale modeling

• Idealization Delta configuration
• Forcings tides and salt at seaward boundary
• Calibration tidal stage by varying bed roughness
• Validation salinity structure, ETM dynamics

12
Carquinez Strait ETM

• Gravitational circulation (GC) common in
  Carquinez Strait
• Topographic control (bump) halts GC on north side
• Near-bed particles trapped
• Longitudinally fixed ETM formed
• What about lateral variability?

Schoellhamer and Burau, 1998
13
Quantifying displacement four sensor method
14
Lateral ETM dynamics

• Increased tidal energy and
• decreased stratification yield
• Southward position (away from the topographic
  control side)
• Higher vertical position due to greater mixing
• Decreased tidal energy and
• increased stratification yield
• Northward position (towards the topographic
  control side)
• Lower vertical position due to less mixing

15
Mechanism 1 particle trapping

• Gravitational circulation and particle trap
  present on spring tides
• On neap tides, particle trapping strengthened
  more on north side

Spring
Neap
16
Mechanism 2 secondary circulation

• On neap tides, sediment accumulates on north side
• On spring tides, secondary circulation sends
  near-bed sediment south

17

• Talk outline
• Background
• Tidal-timescale modeling ETM
• Annual-timescale modeling sediment fluxes
• Decadal-timescale modeling bathymetric change
• Future scenarios geomorphic change

18
Annual-timescale modeling

• Sediment flux data at two boundaries of Suisun
  Bay (5 y of data)
• Interannual processes determine net sediment
  budget
• Forcing add measured winds to drive simple
  wind-wave model
• Calibration 2 y of data (1997, 2004) by varying
  bed characteristics
• Validation 3 y of data (1998, 2002, 2003)

19
Idealized boundary condition seaward SSC

• Measured data not complete gaps due to
  instrument fouling
• Need a synthetic function for historical runs,
  this is an opportunity to test an idealized
  function
• Combine signals from flow, wind, spring-neap
  cycle, and noise

SSCCAR 69.9Qs-16
20
Wet years 1997 (cal) and 1998 (val)
Dashed model Solid McKee et al., Ganju and
Schoellhamer
21
Less wet years 2004 (cal), 2002-2003 (val)
Dashed model Solid McKee et al., Ganju and
Schoellhamer
22
Yearly comparison Net
23
Residual error (kg/s)
Explanation for 1998?Blame Ganju and
Schoellhamer (2006)!
24

• Talk outline
• Background
• Tidal-timescale modeling ETM
• Annual-timescale modeling fluxes
• Decadal-timescale modeling bathymetric change
• Future scenarios geomorphic change

25
Decadal-timescale modeling

• Bathymetric change for Suisun Bay (1867-1887 grid
  has full coverage)
• Forcing idealized winds
• Forcing wind-wave model that accounts for
  changing bathymetry
• Idealization accelerate bathymetric changes
• Idealization use subset of flow hydrographs to
  represent full set
• Calibration match net bathymetric change in
  shallowest 2 m by varying wave period

26
Idealized boundary condition winds

• Composed of seasonal, weekly, and daily
  frequencies
• When used for 2004 simulation, net fluxes
  unaffected
• Can be modified for possible changes in wind
  regime in future

27
Input reduction morphological hydrograph

• Same concept as morphological tide
• Find limited set of forcing data to represent
  full set
• Necessary in system with significant freshwater
  flow
• Use matching procedure to identify most common
  hydrographs

28
Computational reduction morphological
acceleration

• With ROMS, we can update bed level changes at
  every time step
• Provides feedback to hydrodynamic module
• With morphological acceleration, we speed up the
  feedback
• Erosional and depositional fluxes scaled up
  linearly by MF
• With MF20, can we represent 20 y with 1 y
  simulation?

for ?w gt ?c
29
Hindcasting results qualitative performance

• General features
• Deposition in off-channel bays
• Net erosion in northwest channel
• Erosion in landward main channel
• Explanations for areas
• without agreement
• Grain-size distribution
• Wave model
• Consolidation?
• Benthic processes?

Observed 1867-1887 change
Modeled 1867-1887 change
30
Hindcasting results quantitative performance

• Sutherland et al. (2004) use Brier Skill Score
  (BSS)
• Phase term, i.e. erosion/deposition in right
  spots (perfect 1)
• Amplitude term, i.e. changes of correct magnitude
  (perfect 0)
• Volume term, i.e. net change over domain (perfect
  0)
• BSS ranges for classifications are proposed

31

• Talk outline
• Background
• Tidal-timescale modeling ETM
• Annual-timescale modeling fluxes
• Decadal-timescale modeling bathymetric change
• Future scenarios geomorphic change

32
Future scenarios modeling

• How will Suisun Bay respond to climate change and
  anthropogenic forcing (land-use)?
• Not trying to predict future state, just a
  scenario of change
• Most important (i.e. quantifiable) changes
  altered freshwater flows, sea-level rise,
  decreased sediment loads from watershed
• Approach use morphological acceleration factor,
  and three morphological hydrographs

33
Future scenarios morphological hydrographs

• Three morphological hydrographs
• Picked three from 1990-2006 period
• Peak flow, total load most important
  characteristics
• MH1 intermediate Q, Qs (1999)
• MH2 low Q, Qs (2001)
• MH3 high Q, Qs (2006)

34
Future scenarios four simulations

• Scenarios (each scenario has 3 MHs and MF20)
• 1 Base-case (B)
• 2 Warming and sea-level rise of 2030 (WS)
• 3 Decreased sediment loads and sea-level rise of
  2030 (DS)
• 4 Warming, decreased sediment loads, and
  sea-level rise of 2030 (WDS)
• Sources for signals
• Warming Knowles and Cayan (2002), changes are
  minor
• Sea-level rise 0.002 m/y over 30 y, 0.06 m to
  seaward tides
• Sediment loads Wright and Schoellhamer (2004)
  decrease extended to 2030

35
Future scenarios approach

• Scenarios of change
• Interested in differences between scenarios, not
  absolute predictions
• Difference between B and WS gives sea-level rise
  effect
• Difference between WDS and WS gives sediment
  supply effect
• Difference between WDS and DS gives warming
  effect
• Time frame
• Simulation of 1990-2010 geomorphic change
• Base-case represents 1990-2010 under present
  conditions
• Scenarios represent 1990-2010 under 2030
  conditions

36
Base-case morphological hydrographs
MH1 intermediate MH2 dry year, more intrusion
from seaward end, more deposition in deep
channels MH3 wet year, more seaward transport,
more deposition in shallowest 2 m
37
Scenario results changes in relative water depth

• Positive values mean deeper water
• Note Bed levels increase in WS, decrease in DS
  and WDS
• WS warming sea-level rise
• DS decreased sediment supply sea-level rise
• WDS warming decreased sediment supply
  sea-level rise

38
Scenario results changes in bed level

• Sea-level rise WS B
• Sea-level rise dominant signal
• Leads to 9 decrease in wave orbital velocity
• Less redistribution
• Warming WDS DS
• Minor changes in redistribution
• Fringe changes due to phasing of flow-induced
  water level and wind-waves (very minor!)
• Decreased sediment supply WDS-WS
• Erosion everywhere except fringes
• Changes in sediment transfer between shoals and
  fringes during wind-wave period

39
Estuarine geomorphic number

• Import forces sediment supply (Qs), volume,
  depth (h)
• Export forces aspect ratio (area/depth), tidal
  prism (Qp), flow (Q)
• Express as dimensionless ratio

40
Estuarine geomorphic number simple simulation

• Initial depth 4 m
• 100 y simulation
• Typical range of values
• Geomorphic change a non-linear function of
  sediment supply, especially under low sediment
  supply conditions

1850?
2030?
1990
1867
Depth
41
Estimates of contaminant loads via erosion
Combined RMP bed sediment sampling results for
Hg Same maps exist for MeHG, PCB, PAH, PBDE
42
Estimates of contaminant loads via erosion
Scenarios of geomorphic change in Suisun
Bay Worst case scenario (last panel) 0.005 m
net change in erosion Difference between
base-case and worst-case Combine these detailed
spatial results with spatial contaminant
conc. Ignores variation of conc. with depth in
bed (ok if top 15 cm is actively mixed?)
X
kgHg/yArea x ?b x
43
Acknowledgments

• David Schoellhamer, Bassam Younis, Paul Teller
• UC-Center for Water Resources
• CALFED
• USGS-Priority Ecosystems Science Program
• USGS-Community Sediment Transport Model
• Entire ROMS community
• Numerous USGS collaborators
• http//ca.water.usgs.gov/mud