Sediment Sources, Baseline Sediment- Transport Rates and the Effectiveness of Restoration Measures for Reducing Loads to Receiving Waters

Transcription

1 Sediment Sources, Baseline Sediment- Transport Rates and the Effectiveness of Restoration Measures for Reducing Loads to Receiving Waters Andrew Simon Cardno, Oxford, MS, USA

2 Not to Blame for the Content of This Presentation Prof. Andrew Simons Dept of Fisheries, Wildlife and Conservation Biology Bell Museum of Natural History, St Paul, MN

3 Sediment Delivery to Receiving Waters: Issues Degradation of water quality and habitat Loss of reservoir capacity with associated issues of water supply and hydropower generation Damage to estuarine systems; Damage to marine fisheries Damage to reef ecosystems

4 Fundamental Issues for Receiving Waters How much is being delivered? Where is it coming from (sources)? Spatial distribution? Important processes? Delivery to the channel? Can we reduce it? Need to know where it s coming from Need to address ALL of the significant delivery processes if management strategies are to be effective

5 Catchment Models Rainfall-runoff processes Sediment detachment by Rain drop impact Overland flow processes USLE, MUSLE, RUSLE for fields and slopes Edge of Field Erosion * sediment delivery factor Crude channel erosion based on (??) Streambanks (??) often are determined by subtracting predicted upland values from the measured total at the outlet. Why do we ignore streambanks? Lack of data? Difficult to simulate within catchment model? Pathway dependency / Agency inertia?

6 Sediment Export: A + B = C A = overland flow sources: uplands and fields B = channels and gullies C = delivery to receiving water body But, does B = C A?? Assume we can rely on calculation of loads ( C ) based on measured values; How robust are the calculations of A loads from uplands and fields (USLE/RUSLE);? Are the sediment-delivery and storage terms reliable? Should we not use a more deterministic approach?

7 Sediment Delivery: Chesapeake Bay, 2011 August 23, 2011 September 13, 2011

8 Kolan River Sediment Delivery: January 2013 Queensland Floods (Great Barrier Reef) Burnett River

9 Hoteo River, NZ; March 2011 Kaipara Harbor, North Island, NZ: Important Maori and Commercial Fishery being impacted by deposition of fine sediment

10 Potential Sediment Sources Upland hillslopes Forests Pastures Timber production / Roads Foot slopes and Fields Forests Pastures Cropping Channels Beds Banks

11 Hillslope Sediment Eroded hillslope sediment can be trapped (stored) on lower slopes and in fields Poor Connectivity Low sedimentdelivery ratio

12 Streambank Sediment Sediment from stream banks is delivered directly to the channel, available for downstream transport. Good Connectivity Perhaps 100% sediment delivery

13 Sediment Sources ~100 years ago upland & agricultural sources dominated December 1906 December 2013 Images provided by Mark and Loma Page, 2014

14 There is a growing body of evidence that channel systems, particularly streambanks, are now the dominant source of sediment in many disturbed systems Oklahoma Tennessee Minnesota Nebraska Kansas Mississippi

15 For Example, in Kansas ~ 631,000 tonnes eroded during 7 months About 42,000 dump trucks Photo and data from Phil Balch

16 Example Contributions of Streambank Erosion Stream Ecoregion Dominant Bed Material Contribution from Banks James Creek, MS Southeastern Plains Sand/Clay 78% Shades Creek, AL Ridge and Valley Gravel 71-82% Goodwin Creek, MS Burnett River, QLD O Connell River. QLD Upper Truckee River, CA Le Sueur River, MN Obion-Forked Deer River, TN Mississippi Valley Loess Plains Sand/Gravel 70% Reef Catchment Sand 44%-73% Reef Catchment Sand/Gravel 64% Sierra Nevada Gravel 47% Western Corn Belt Plains Mississippi Valley Loess Plains Gravel 75% Sand 81%* *Represents percent contribution from channel sources

17 Example: Coral Sea & Great Barrier Reef Burnett River: January 2013

18 How Much Sediment and Where is it Coming From? ~ 8% from streambanks, (Brodie et al., 2003) Is that Correct?? Determine bank-derived sediment loads Empirical and modeling (BSTEM-Dynamic) Compare to SedNet (catchment model) Implications for sediment management?

19 Channel Widening ( ) Empirical Analysis Based on Analysis of LiDAR and Air Photos Paradise Dam Simon et al., 2014

20 Unit Bank-Erosion Rates: (m 2 /m) Paradise Dam

21 Empirical Results: (Land lost by bank failures and delivered to the river) About 28 million m 3 (47 million t) (10.2 Mt/y) over the 292 km main stem (4.58 years) About 21 million m 3 or two thirds of these sediments were eroded from banks downstream of Paradise Dam Several reaches had calculated land losses in excess of 1 million m 3. This annual rate represents a short period, characterized by very high flows. Need to simulate a longer period

22 SedNet Long-Term Erosion Rates 1. Long-term erosion rates? (175,000 t/y; 0.175Mt/y) 2. Comparison to other catchment sources? (8%) 3. Sediment export to the Great Barrier Reef? (0.47 Mt/y) 4. Overall management strategy for the Burnett River Catchment? 5. Dredging of the Bundaberg Harbour and the lower River? 6. Are these results unique or typical for the Reef Catchments? Data from Brodie et al., (2003)

23 WATER LEVEL, M 2-D wedge- and cantileverfailures Daily (hourly) time steps Search routine for failures Hydraulic toe erosion Accounts for grain roughness Complex bank geometries Positive and negative pore-water pressures (dynamic groundwater) Confining pressure from flow Layers of different strength Vegetation effects: RipRoot Variable roughness Inputs: g s, c, f, f b, h, u w, k, t c FACTOR OF SAFETY Bank-Stability Model (Dynamic) Ver. 2.3 B Confining pressure Stage Factor of safety Effect of confining pressure Bank failures shear surface Tensiometers (pore pressure) /29/97 01/05/98 01/12/98 01/19/98 01/26/98 02/02/ RIVER STAGE, IN METERS ABOVE SEA LEVEL

24 Average, Annual Bank-Erosion Rates (for different simulation periods) Brodie et al., (2003) BSTEM-derived results for the 4.58 year modeling period within 10% of empirical value; (9.3 Mt/y)

25 SedNet: The Numbers Hillslopes: 1.65 Mt/y Gullies: 0.93 Mt/y Banks: Mt/y (8%) Total: 2.76 Mt/y Scenario 1 Hillslopes: 1.65 Mt/y Gullies: 0.93 Mt/y Banks: 2.0 Mt/y (44%) Total: 4.6 Mt/y Scenario 2 Hillslopes:? Mt/y Gullies:? Mt/y Banks: 2.0 Mt/y (72%) Total: 2.76 Mt/y We have a problem here What if management strategies are based on these results??

26 How Much Sediment Reduction? Targets Level III Ecoregions and Available Data

27 Two-Stage Suspended-Sediment Ratings SUSPENDED-SEDIMENT LOAD, IN TONNES/DAY James Wolf Creek near Looxahoma, MS DISCHARGE IN CUBIC METERS PER SECOND 110,000 T/D 26,000 T/D Q 1.5

28 Median Suspended-Sediment Concentrations Median Suspended-Sediment Concentration at Q in mg/l no data Simon et al (2004)

29 Median Suspended-Sediment Yields Median Suspended-Sediment Yield at Q in T/D/km no data

30 Stage I. Sinuous, Premodified h<h c h c = critical bank height Reference Stages h = direction of bank or bed movement Stage I. Sinuous, Premodified h<h c h Stage II. Constructed Stage III. Degradation h<h c h c = critical bank height h<h c floodplain h = direction of bank or bed movement h Stage IV. Degradation and Widening h>h c terrace h Stage IV. Degradation and Stage II. Constructed Stage III. V. Degradation Aggradation and Widening Widening Stage VI. Quasi Equilibrium h<h c h c = critical bank height h<h h>h c c h>h c h<h c floodplain = direction of bank or terrace terrace terrace bed movement bank h h h h bankfull h slumped material References aggraded material aggraded material Stage IV. Degradation and slumped material tage III. Degradation Widening <h c h>h c Stages I, II V. Aggradation and Widening Stage VI. Quasi Equilibrium primary h>h Stage c I terrace h<h c knickpoint Stage III terrace terrace Stage IV t o p b a bank n Stage V h h precursor plunge k Stage h h VI knickpoint pool bankfull d i r e Stage VI c t i o n o f f l o w slumped material slumped secondary material knickpoint aggraded material oversteepened reach aggraded aggradation material zone aggraded material ning Stage VI. Quasi Equilibrium h<h c Stages I, II terrace primary knickpoint h bank Stage bankfull III Stage IV slumped material

31 Reference Suspended-Sediment Yields SEDIMENT YIELD AT Q 1.5, IN T/D/KM Minimum 1st Quartile Median 3rd Quartile Maximum Stable General "reference" 0.31 T/D/km 2 SOUTHEASTERN PLAINS SEDIMENT YIELD AT Q 1.5 IN T/d/km Minimum 1st Quartile Median 3rd Quartile Maximum Stable Reference ARIZONA/NEW MEXICO PLATEAU 1.3 T/d/km 2

32 Reference Suspended-Sediment Concentrations SEDIMENT CONCENTRATION AT Q 1.5, IN mg/l Minimum 1st Quartile Median 3rd Quartile Maximum Stable General "reference" 47.9 mg/l SOUTHEASTERN PLAINS SEDIMENT CONCENTRATION AT Q 1.5 IN mg/l mg/l Minimum 1st Quartile Median 3rd Quartile Maximum Stable Reference ARIZONA/NEW MEXICO PLATEAU

33 How Much Reduction? PERCENT OCCURRENCE I II III IV V VI STAGE OF CHANNEL EVOLUTION Both watersheds contain unstable channel systems characterized by incised channels and unstable banks ER 27: Ft Cobb and Little Washita Channels

34 ER27 Results Suspended-sediment yields from Cobb Creek and Little Washita River are among the highest in the ecoregion (1-2 orders of magnitude greater than the median value for stable sites) at Q 1.5 for mean-annual yield

35 Simon and Klimetz, 2008 How Much Reduction?

36 Duration of High Concentrations and Benthic Populations TOTAL # BENTHIC MACRO-INVERTEBRATES Data from 8 sites in Ecoregion 74 Y = 96585x R 2 = y = x R 2 = ANNUAL DURATION OF SUSPENDED SEDIMENT ABOVE 1000 mg/l (MIN)

37 Metrics for Stable and Unstable Sites Frequency Duration PERCENTAGE OF TIME EQUALLED OR EXCEEDED INTERIOR PLATEAU Ecoregion 71 Stable sites Unstable sites AVERAGE NUMBER OF CONSECUTIVE DAYS EQUALLED OR EXCEEDED Stable sites Unstable sites INTERIOR PLATEAU Ecoregion 71 PERCENTAGE OF TIME EQUALLED OR EXCEEDED CONCENTRATION, IN MILLIGRAMS PER LITER Stable Sites Unstable Sites BLUE RIDGE Ecoregion 66 AVERAGE NUMBER OF CONSECUTIVE DAYS EQUALLED OR EXCEEDED CONCENTRATION, IN MILLIGRAMS PER LITER BLUE RIDGE Ecoregion 66 Stable Sites Unstable Sites CONCENTRATION, IN MILLIGRAMS PER LITER CONCENTRATION, IN MILLIGRAMS PER LITER

38 Reference Frequency of Exceedance (%) Lowland coastal areas Piedmont Highly erodible loess area Plains Need to link with functional traits of aquatic organisms

39 Effectiveness of Mitigation Strategies Bank stability is increased by. Vegetation mechanical, hydraulic and hydrologic effects drainage toe protection / flow reduction

40 Effectiveness of Toe Protection 6000 STREAMBANK EROSION IN CUBIC METERS BW 1.94 E BW 1.94 TP 1 BW 2.39 E BW 2.39 TP Hydraulic toe erosion Geotechnical erosion Lake Tahoe Basin UT 4.51 E No veg UT 4.51 TP 4 UT 13.1 E UT 13.1 TP SITE UT 4.51 E UT 4.51 TP 10 UT 8.45 E UT 8.45 TP WA 2.48 E WA 2.48 TP 5 WA 3.60 E WA 3.60 TP 1 Simon et al (2009) Toe protection = 86% (average) Top-bank vegetation = 53% (root reinforcement) Bed-slope reduction (meandering) = 42-54%

41 ELEVATION, IN METERS ELEVATION, IN METERS Example: Modeled Mitigation Results Start Profile No Action With rock to to 1.5m With ELJ, average per m & per structure With rock to to 1.5m, with vegetation No imposed changes in geometry STATION, IN METERS 20 2:1 Start Profile Hoteo River, NZ: Site 27A :1 Start Profile With reconfigured bank, 2:1 With reconfigured bank, 1.5:1 With reconfigured bank, 1.5:1, with vegetation Flattening of banks STATION, IN METERS

42 Cost Effectiveness of Mitigation Scenarios: (HO-27A: 240 m) PERCENT REDUCTION 100% $400,000 90% 80% 70% 60% 50% 40% 30% 20% 10% Percent reduction in retreat Percent reduction in eroded volume Total Cost $350,000 $300,000 $250,000 $200,000 $150,000 $100,000 $50,000 TOTAL COST IN NZ $ 0% No Action With reconfigured bank, 1.5:1 With reconfigured bank, 2:1 With reconfigured bank, 1.5:1, with vegetation With reconfigured bank, 1.5:1, and rock toe to 1.5m With rock to to 1.5m With rock weirs With ELJ, average per m & per structure With rock to to 1.5m, with vegetation $- Ultimate decision by Stakeholders to be based on: Cost and available resources Health, safety and habitat benefits Need to hold the line? Priority of Load reduction, Land Loss vs. Retreat Aesthetics

43 Generic Bank- Stability Modeling to Test Erosion- Control Effectiveness: West Tennessee Used BSTEM-Dynamic; 155 sites; Daily flows for ; 75 th, 50 th and 25 th percentile values for boundary resistance by sub-basin

44 Bank-Stability Modeling to Test Mitigation Effectiveness in West Tennessee Average Erosion Reductions* Riparian planting: 16% (5-year old trees) Bank-toe protection: 83% Bank-toe & bank-face protection: 88% * For median-strength conditions (higher for 25 th percentile-strength conditions) Only 5% improvement for protecting the entire bank face!! Most effective Least effective

45 Flow Control and Bank Erosion: Mitta Mitta River, Victoria, Australia Dartmouth Dam completed in 1979 has significantly altered the flow regime of the Mitta Mitta River; During dry times, bulk-water transfers are undertaken to supplement storage in Lake Hume; These can be at bankfull levels (~10,000 ML/d) and last for months it s a pipe; Example bulk-water transfers

46 Erosion Thresholds as Determined by BSTEM Site 9L 13R 15R 19R 22R 27R 36L Threshold (ML/d) 9,504 12,960 11,232 5,184 5,184 6,912 5,702 Sites with lowest thresholds have the lowest critical shear-stress values at the bank toe; 1.1, 0.8, and 0.1 Pa (equivalent to resistance of sand); Higher erosion rates are associated with the lowest thresholds No erosion thresholds < 5,184 ML/d, indicating that the current optimal release level of 5,000 ML/d is valid for minimizing erosion.

47 Testing of Alternative Flow Scenarios How best to transfer 920,000 ML over 7 months S1: Moderate-Constant - 5,000 ML/d, 167-day duration S2: Maximum Rate of Fall - Variable, six 7,500 ML/d peaks, recession 20 mm/h, 30- day duration S3: Slower and Smaller - Variable, six 6,400 ML/d peaks, recession 5 mm/h, 26 day duration S4: Worst Case-Maximum Flow - one 10,000 ML/d peak for 86 days

48 Erosion for Alternative Flows (7 months) By site From lowest to highest erosion rates (average values, in m 3 /m) Moderate-Constant Flow: 0.016**; Slower and Smaller: 0.037; Maximum RoF: 0.086; Worst Case/Maximum Flow Rate: By flow scenario **Would not be used without some flow variability

49 Use the Right Tools for the Problem (to Reduce Risk and Uncertainty in Design) 1. Integrate flow and sediment model with bankprocesses model, and ultimately 2. Integrate Channel Model with upland model

50 Some Channel-Model Capabilities Process BSTEM HEC- RAS SRH- 2D HEC+ BSTEM SRH- 2D+BSTEM Shear in meanders Bank-toe erosion Mass-failures Bed erosion Sediment transport Vegetation effects Hard engineering Channel evolution Rapid Assessments

51 HEC-RAS/BSTEM: Interaction of Three Processes Currently contains a simplified version of BSTEM: 1-layer No root reinforcement 3. Bank failure Failure 2. Toe scour Scour 1. Incision modified from Gibson et al., 2015

52 Example HEC-RAS/BSTEM Results From Gibson et al., 2015

53 Summary and Conclusions Gravity and the physics of erosion and sediment transport are the same wherever we are, allowing us to quantify force and resistance mechanisms and predict loadings. Sediment eroded from catchments adversely impact receiving water bodies Sources include uplands, fields AND channels, particularly streambanks Determining sources and magnitudes of sediment erosion and delivery are critical to providing resilient sediment management Appropriate field and analytic techniques exist to quantify catchment and channel erosion, sediment delivery, AND potential reductions. They are no more expensive or time consuming than current practices.