Linking Sediment Transport in the Hudson from the Tidal River to the Estuary

Transcription

1 Linking Sediment Transport in the Hudson from the Tidal River to the Estuary Or, what happened to all the mud from Irene? David Ralston, Rocky Geyer, John Warner, Gary Wall Hudson River Foundation seminar October 2015 Landsat 5, August 31,2011; credit: USGS/NASA Earth Observatory

2 Poughkeepsie 120 km Newburgh Bay 95 km Salinity intrusion at low Q r Irene: Aug head of tides at Troy 240 km Haverstraw Bay (ETM) High Q r GW Bridge (ETM) Battery 60 km 0 km 18 km Atlantic Ocean August 31, 2011

3 Tropical Storms Irene and Lee Irene (Aug 28) + Lee (Sep 7) rainfall totals (Aug 25-Sep8) [ Mohawk Albany Poughkeepsie NYC

4 Mohawk R. (30% of watershed area) Catskill Creek (3%) Esopus Creek (4%) Upper Hudson 0.2 Mton (40%) 1.4 Mton 0.7 Mton 0.03 Mton Rondout Creek (10%) 0.2 Mton Tributary sediment loads Total sediment input from Irene and Lee: 2.7 Mton Long-term annual avg.: 0.5 Mton Mohawk Cohoes (near head of tides) Suspended sediment ~ 2.2 g/l August 29, 2011 (credit: G. Wall)

5 Estuaries efficiently trap sediment High sediment concentrations and deposition rates, efficient trapping estuarine circulation stratified unstratified seaward flow Ocean River ETM Fluvial sediment Marine sediment Does the Hudson look like this cartoon?

6 Lower Hudson ETM: near GW Bridge high deposition rates after spring freshet June 1998 June cm Woodruff et al. 2001

7 Lower Hudson ETM: near GW Bridge high deposition rates after spring freshet 12 cm 9 days May/June 2001 Traykovski et al Atlantic Ocean

8 Lower Hudson ETM: near GW Bridge high deposition rates after spring freshet high sediment concentrations (>1 g/l) BUT, at intermediate salinities, not at salinity limit Geyer et al. 2001

9 Upper Hudson ETM: Haverstraw Bay High deposition rates; frequent dredging Often near limit of salinity intrusion Nitsche et al., 2010

10 Upper Hudson ETM: Haverstraw Bay High sediment concentrations at flood tide salinity fronts, Ebb concentrations reduced by stratification Suspended sediment concentration Salinity Observations, Fall 2009 Ralston et al., 2012

11 Poughkeepsie 120 km Salinity intrusion at low Q r tidal fresh tidal fresh head of tides at Troy 240 km What happens to sediment in the tidal river? Much less studied, but important in major river systems Amazon (600 km), Ganges- Brahamaputra, Changjiang ( km), Mekong (>190 km); also Thames (160 km), CT (100 km) Can sequester an estimated 1/3 of sediment discharge [Milliman and Farnsworth 2011] High Q r Newburgh Bay Haverstraw Bay (ETM) GW Bridge (ETM) Battery 95 km 60 km 0 km 18 km Tidal and river forcing both contribute to transport, with different time scales and amplitudes Tidal velocities usually dominate, control river geometry, vary at spring/neap cycle River discharge varies at event, seasonal, and interannual time scales; magnitude at times similar to tides, but unidirectional River discharge controls sediment supply Atlantic Ocean

12 Irene/Lee observations Gaged flow & SSC (91% of watershed) ADCP w/ SSC calibration at Poughkeepsie ` Mix of turbidity, salinity, and water level sensors; some calibration during Irene/Lee.

13 Model d Hydrodynamics (ROMS) and sediment transport (CSTMS) Previously used in the estuary (Ralston et al. 2012) Bed initial condition from surveys (Nitsche et al. 2007) sand mud 100 Discharge and sediment load (USGS, measured or rating curves) for 10 tributaries 50 0 km

14 Model evaluation Water level Salinity Observed Model (surf) Model (bot) Suspended sediment

15 Discharge + sediment flux: observed Irene Lee Poughkeepsie

16 Discharge + sediment flux: observed, modeled Irene Lee Poughkeepsie

17 Where did the new sediment go? Irene end Sediment mass distributions through time Irene start model results

18 Where did the new sediment go? Sediment mass distributions through time Before Lee After Lee model results

19 Where did the new sediment go? Sediment mass distributions through time Fresh, tidal river 1 mo. after Lee Deposition in upper ETM model results

20 Observed accumulation over past ~70 yr. from Frank Nitsche (LDEO) New sediment deposition 1 month after Lee Newburgh Bay model results

21 Suspended sediment from watershed 2 psu 5 psu Advective length scale ~ U river t storm ~ (0.4 m/s)(2 d) ~ 70 km Flood pulses too short to move sediment through tidal river. model results

22 Suspended sediment from bed upper ETM 2 psu 5 psu lower ETM Remobilized bed sediment due to increased stress and reduced stratification dominates SSC signal in the estuary. model results

23 model results Erosion and deposition of bed sediment 1 month after Lee

24 Why is new sediment stuck in the tidal river? Transport in river scales linearly ~ Q r 1 Sediment flux at Poughkeepsie Watershed sediment loading highly non-linear ~ Q r 3 Q s = Q r C s C s ~ Q r 2 (Woodruff 1999) Q s = Q r C s C s ~ U t 2 * * a massive oversimplification - River provides mean flow ( m/s), but tides dominate resupsension (0.5-1m/s) - Enhanced velocity, resuspension, seaward flux during event is brief, much less than input

25 Sediment flux tide Spring tides, low discharge river input Catskill Poughkeepsie Haverstraw model results

26 Transport time scale: T ~ M s /Q s An engineering approach Mass of new sediment in river (M s ) Flux out (Q s )

27 Transport time scale: T ~ M s /Q s An engineering approach Ignores changes in erodibility, settling velocity; sequestering model uncertainty increases w/ time Mass of new sediment in river (M s ) Flux out (Q s ) Mass of new sediment above Poughkeepsie Time scale based on flux at Poughkeepsie model results

28 Effect of discharge? Irene/Lee (2011) 95 th percentile Spring 2014 Winter/spring th percentile observations

29 Spring 2014 typical freshet Salinity at Haverstraw Sediment at Schodack, Tivoli, Newburgh, and Haverstraw observations

30 Spring 2014 typical freshet Mass of new sediment above Poughkeepsie Time scale based on flux at Poughkeepsie model results

31 Winter/spring 2015 below average discharge Salinity at Haverstraw Sediment at Schodack, Tivoli, Newburgh, and Haverstraw observations

32 Winter/spring 2015 below average discharge Mass of new sediment above Poughkeepsie Time scale based on flux at Poughkeepsie model results

33 Transport during event depends on salinity intrusion Storm duration << estuarine response time scale 2 psu 5 psu model results

34 Sediment transport capacity in the estuary Simpler, 1-d model of estuary run for 100 years Channel Cross-section total different model results

35 Sediment transport capacity in the estuary Simpler, 1-d model of estuary run for 100 years NEAP more salt, more trapping less seaward flux Channel Cross-section total SPRING less salt, weaker trapping more seaward flux different model results

36 Summary Sediment load from extreme events greatly exceeds near-term transport capacity Watershed input scales as Q r3, transport as Q r, New sediment trapped in tidal river (~80%) In estuary, bed remobilization due to reduced stratification, increased stress dominates Potential to reintroduce sequestered contaminants Depends on spring/neap phasing, event duration Sediment residence times may be much longer, more variable than thought Previously, focus on annual freshet cycle Slower export of terrigenous material to the coastal ocean? Physics hard to constrain, use geochemistry

37

38 Salinity front sediment trapping: Haverstraw Bay Salinity + bathy bottom stress + sal. SSC + velocity erosion/deposition + sal. Slack before flood Net over 1 tidal cycle model results, Fall 2009

39 Salinity front sediment trapping: GW Bridge salinity + bathy stress + sal. SSC + velocity erosion/ deposition + sal. Slack before flood model results, Fall 2009 Net over 1 tidal cycle