Management in the Tahoe Basin Soil disturbance & restoration - - PowerPoint PPT Presentation

Erosion Modeling for Land Management in the Tahoe Basin – Soil disturbance & restoration detection thresholds

Mark E. Grismer

Hydrologic Sciences, UC Davis and Integrated Environmental Restoration Services, Tahoe City, CA

Soils Restoration – Local to Watershed Processes - Hypotheses

1.

Improved “soil function” at local-scale (e.g. infiltration, aggregate stability, microbial community structure, soil strength…) leads to reduced sediment fines and nutrient loadings.

2.

Reductions in sediment loadings may be “detectable” within a few years of pre- and post-project monitoring.

3.

Focused discharge and sediment sampling during the daily and seasonal rising limb of the hydrograph provides the nearest approximation to actual daily sediment loading from Tahoe west shore streams.

4.

“Disconnecting” adjoining eroding areas reduces sediment loading disproportionately to area treated.

Related Project Objectives



Compare sediment load-flow relationships developed from estimated and measured data for Ward and Blackwood Creeks to provide some insight into the relative bias or systematic error of previous efforts.



Develop measured TSS, fine-sediment particle (FSP<20 micron) and nutrient (TKN, TN & TP) load- flow relationships for Homewood (HMR) Creek.



Using hourly estimates of mean daily flows and total daily sediment loads, determine which hourly period(s) if sampled alone best represent the daily sediment loading from Ward and HMR Creeks.



Determine if there is a change in HMR Creek watershed sediment yield (kg/ha) per unit flowrate following soils restoration and erosion pathway disconnection work completed in the catchment during summers of 2006-2010.

Process-level Soils Information - Conclusions

 Understanding fundamental soil processes is important

towards restoration or monitoring success, but often such information is lacking.

 Relative levels of aggregation, possible crusting,

repellency, OM %, and microbial community structures in the soil may be linked to runoff particle-size distributions, sediment and nutrient loadings from catchments.

 Similarly, knowledge of these soil processes should

provide insight into the relative merits of various treatments.

 Presumably, plot-scale processes affect those at the

watershed scale…

Soil Restoration – Watershed effects

Sediment Yield Curves – incorporate soil, slope, cover, strength aspects of soil “functionality” into an “effective” erodibility…

• scale to watershed area through surface runoff

routing of different SY areas on daily basis.

• determine changes in watershed sediment & fines

loading after restoration within watershed.

• determine stream monitoring required to measure

minimum soils restoration, or disturbance effects on watershed sediment loading within a prescribed confidence level.

Scaling from plots to basins

Dollar Hill 1D - Sediment vs. Runoff

CS = 3.4525*CR R2 = 0.997 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 0.00 0.40 0.80 1.20 1.60 2.00 2.40 2.80 3.20 3.60 4.00 4.40 Cumulative Runoff (mm) Cumulative Sediment (gm) 10000 20000 30000 40000 50000 60000 70000 80000 90000 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Accumulated Runoff (mm) Accumulated TSS load (kg)

Sedload Predicted

HMR Creek Sediment Loading -Relative SF Predictive Error

y = 37.505Ln(x) - 416.01 R2 = 0.8933 y = -2E-05x + 5.9456 R2 = 0.0031

• 60.0
• 50.0
• 40.0
• 30.0
• 20.0
• 10.0

0.0 10.0 20.0 30.0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000 110000

Actual Sedload (kg) Relative Prediction Error (% of actual)

Even yrs Odd yrs SF=0.156/R-0.7 SF=0.192/R-0.5

Modeling Results Review

 Daily SF analyses enabled more detailed assessment and

allows for evaluation of disturbance or restoration efforts

• n loading changes from the basin.

 SF’s are runoff magnitude dependent; particularly at

small runoff values.

 SF’s are highly variable at low runoff due to sediment

loading hysteresis effects and dominance of channel factors.

 For HMR Creek at 1 mm of runoff, SF = 0.192 suggests

that the RS plot-scale data was ~5 times that needed to represent the basin sediment loading.

 Seasonal or annual sediment loads can be predicted

within 20-30%, rather than orders of magnitude.

 Comparisons of SF functions with adjacent Madden &

Quail basins were similar & suggest possible wider use.

“Proof of Concept” Modeling to Detect Soil Functionality Changes

Example Application – Fuels harvesting/thinning in Madden Cr. Watershed.

 Using existing fire road infra-structure, harvest-

thinning operations from mid-range slope forests assumed to result in soil functional degradation to that equivalent to ski-runs.

 Modeled effects based on daily flows and sed-

loading analyses for period 1994-2004.

 Due to dependence of sed-loading on flows,

results are considered by incremental flow steps .

Homewood and Madden Creek general land-uses (2008)

Land-use Category Homewood Creek (260.9 ha) Madden Creek (529.5 ha) Area (m2) Fraction of WS (%) Slope (%) Area (m2) Fraction of WS (%) Slope (%)

Dirt Roads 84,497 3.24 49.3 54,135 1.03 49.1 Ski-run Areas- 439,173 16.83 49.6 613,033 11.64 46.8 Forested Areas 2,027,276 77.70 ~43 4,574,505 86.86 ~45 Residential 31,451 1.21 14.0 19,559 0.37 20.3 CICU – Imperv. 4768 0.45 17.9 NA CICU - Pervious 7082 10.6 NA Paved Roads 15,013 0.58 18.5 3792 0.07 15.0 Annual Runoff (mm) & range 70 9.3-193.1 64.5 8.6-181

• Ann. SY (kg/ha/mm) & range

6.14 1.8-11.3 7.88 2.9-13.3 Soils Fractions (Volcanic/Granitic) 0.89 0.11 0.93 0.07

Harvest/Thinned Areas as Fraction of mid-slope forests and basin areas

EP3 Fraction Area (m2) Forest Fraction WS Fraction 5 129935 2.84% 2.47% 10 259869 5.68% 4.93% 15 389804 8.52% 7.40% 20 519738 11.36% 9.87% 25 649673 14.20% 12.34% 35 909542 19.88% 17.27% 45 1169411 25.56% 22.21% 60 1559215 34.08% 29.61%

2000.0 2100.0 2200.0 2300.0 2400.0 2500.0 2600.0 2700.0 2800.0 30.0 32.0 34.0 36.0 38.0 40.0 42.0

Daily Mean Sedload (kg) Daily Mean Discharge (cfs)

Madden Creek - Monitored 1995-97 WYs, Harvested Summer, 1997 and Monitored 1998-99 WYs

Mean Mean+SD 25% harvest 45% Harvest 60% Harvest

Confidence 86.9% 85.6% 83.8% Confidence 94.2% 93.6% 92.7% Confidence 57.5% 56.0% 53.9%

Madden Prelim Harvest Analyses

 Comparison of 11-yr record sed-loading with and w/o

harvesting operations that have a mild, presumably temporary effect on sed-loading rates are difficult to detect across all flowrates with any confidence.

 Detection of changes in sed-loading perhaps more likely

at mid-range flowrates, but depends on previous years effects on channel conditions.

 Need to consider both inter-annual event effects as well

as shorter time scale processes to get a better handle on measuring sed-load changes with sufficient confidence.

“Proof of Concept” Modeling to Detect Soil Functionality Changes

Restored Road & Ski-run area Fractions (%) Area (m2) WS Fraction (%) 50 42,249 1.62 50 10 86,166 3.30 50 20 130,083 4.99 50 30 174,000 6.67 50 40 217,918 8.35 50 50 261,835 10.0 Example Application – Soils restoration in Homewood (HMR) Creek Watershed.

HMR Creek Restoration Confidence Levels of Detection for 1994-2004

Baseline Flow Confidence Levels of Detection Restoration fractions of Road/Ski-run areas N (cfs) 50/50% 50/40% 50/30% 50/20% 50/10% 50/0% 21 28.4 98.9% 97.7% 95.5% 91.9% 86.3% 78.5% 37 22.0 99.8% 99.3% 98.2% 95.8% 91.2% 85.7% 52 18.5 99.9% 99.9% 99.6% 98.5% 95.6% 89.0% 31 15.2 97.6% 97.6% 96.1% 93.9% 90.9% 75.0% 61 12.5 98.5% 97.4% 95.6% 92.7% 88.7% 74.8% 50 9.92 95.1% 92.6% 89.2% 84.7% 79.2% 70.5%

Comparison Periods

Baseline – No restoration 50%, 50% Restoration 50%, 40% Restoration 50%, 30% Restoration 50%, 20% Restoration N Mean Q (cfs) Mean Sed (kg/d) Std. Dev. (kg/d) N Mean Sed (kg/d CL (%) Mean Sed (kg/d CL (%) Mean Sed (kg/d CL (%) Mean Sed (kg/d CL (%)

Monitored 1995-96, restoration '96, monitored '97-98

29 18.6 780 93.8 15 716 97.8 725 95.9 734

• 92. 8 743 88.1

Monitored 1995-96, restoration '96, monitored '97-99

29 18.6 780 93.8 23 713 99.3 721 98.4 730 96.7 739 93.4

Monitored 1995-96, restoration '96, monitored '97-99

18 15.4 712 153.2 9 681 76.7 689 70.3 697 63.1 706 55.4

Monitored 1995-96, restoration '96, monitored '97-98

8 12.1 814 69.6 29 653 99.9 661 99.9 669 99.9 677 99.9

Monitored 1995-96, restoration '96, monitored '97-99

8 12.1 814 69.6 46 642 99.9 650 99.9 658 99.9 666 99.9

Monitored 1995-96, restoration '96, monitored '97-98

14 9.87 614 92.1 13 545 91.3 551 88.9 558 86.2 565 82.9

Monitored 1995-96, restoration '96, monitored '97-99

14 9.87 614 92.1 15 537 95.2 543 93.6 550 91.5 557 88.9

Can we improve on monitoring of Hillslope Restoration changes ? Basics – Sedloading & Streamflow

 Extensive datasets (1999-2001) from Blackwood and

Ward Canyons (Andy Stubblefield PhD) and HMR Creek (2009-2011) on Tahoe west shore.

 Data collected at 15-min intervals enables analyses at

multiple time steps (e.g. 1, 4, 12 hrs).

 Considerable hysteresis between TSS concentrations

(mg/L) & flow (cfs) in daily and seasonal hydrographs.

 Diurnal daily flow peaks increase with increasing

temperature (typically from April-June).

 TSS-loads increase as greater surface areas are exposed

and channel flow velocities increase (non-linearly).

 Seasonal overlay at play as channels “scoured” by rain-

• n-snow, or other large flow events.

Processes – Sedloading & Streamflow

 At small time scales (~1 hr), flow and sed-load

peaks occur simultaneously and recession limb sed-loads only a fraction of rising limb values.

 Diurnal hydrograph rising limb event durations

consistently ~6 hrs and progressively increase average Q and sed-load through April-May until major event occurs that “cleans” channel.

 Increasing sample averaging >~6 hr decreases

sed-load variability, but then includes recession limb hysteresis problems.

Hysteresis in Streamflow & Sediment load at HMR Creek, April-May 2010

0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 49 142 234 326 418 510 602 695 787 879 971 1063 1155 1248 1340

Hours after Midnight 4/30/10 Average Hourly Flowrate

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0 120.0

Average Hourly TSS Flow (L/sec) TSS (mg/L)

Linear regression relationships for turbidity probes in Homewood Creek (2009-2011)

Relationship

n Slope Intercept R2

TSS (mg/L) vs Turbidity (ntu)

57 1.802

• 0.093

0.975

FSP (mg/L) vs Turbidity (ntu)

36 0.491 0.178 0.855

TKN (ppb) vs Turbidity (ntu)

34 3.878 282.3 0.038

TN (ppb) vs Turbidity (ntu)

34 18.62 103.8 0.811

TP (ppb) vs Turbidity (ntu)

34 2.187 22.32 0.887

Total annual and spring-summer sediment load from Homewood Creek in 2010 WY as measured and estimated by different sampling periods

2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000 26000 28000 30000 32000 34000 Load (kg) 9-10 am 10-11 am 11-12 am 12-1 PM 1-2 PM 2-3 PM 3-4 PM 4-5 PM 5-6 PM 6-7 PM

Daily Sampling period Total Sediment load (kg)

all data April - Sept

Comparison of estimated (1997-2002) and measured (April-May 2001) daily Sediment Load- Flow relationships for Blackwood Creek

y = 35.049x4.4106 R2 = 0.6656

1.0 10.0 100.0 1000.0 10000.0 100000.0 0.10 1.00 10.00 Mean Daily Flow (m3/s) Daily Sediment Load (kg) Measured Estimated

Comparison of estimated (1997-2002) and measured (April-May 1999-2000) daily Sediment Load-Flow relationships for Ward Creek

y = 416.88x1.3732 R2 = 0.5153

y = 6.9835x4.2281 R2 = 0.7948

100.0 1000.0 10000.0 100000.0 1.00 10.00

Mean daily Flowrate (m3/s) Total Daily Sediment Load (kg)

Q<4.0 Q>4.0 Est<2.0 Est>2.0

Daily measured (2009-2011) sediment (TSS) Load-Flow relationships for HMR Creek

y = 0.0084x

1.7004

R

2 = 0.9372

y = 0.0206x

1.4032

R

2 = 0.4837

y = 0.0009x

2.096

R

2 = 0.7535

0.01 0.10 1.00 10.00 100.00 1000.00 10000.00 1.0 10.0 100.0 1000.0

Mean Daily Flowrate (L/s) Total Daily Sediment Load (kg)

Q<100 Q>100

Daily measured (2009-2011) total phosphorous (TP) Load-Flow relationships for HMR Creek

TP = 1.6073x1.0931 R2 = 0.9883 Q>100 TP = 0.5246x1.2939 R2 = 0.8896 Q<100 TP = 1.9099x1.0264 R2 = 0.997 1.0 10.0 100.0 1000.0 10000.0 1.0 10.0 100.0 1000.0

Mean Daily Flow (Lps) Daily TP Load (g)

2010 WY 2011 WY 10/24/10 storm

Summary of optimal hourly sampling period at West-shore creeks based on the different statistical methods

Analysis Method Creek Data Period Optimal sampling Associated Figs or Tables

HMR 2010 WY 1 PM

• Fig. 4

Annual Load HMR 2011 WY 3 PM

• Fig. 5

Blackwood 4-5/2001 5-6 PM Table 5 Ward 4-6/1999-00 3-4 PM Table 5 RMSE HMR 2009-11 4-5 PM Table 5 Ward 4-6/1999-00 1-2 PM Table 6 HMR 2010 WY 3-4 PM Table 6 T-test HMR 2011 WY 4-5 PM Table 6 Ward 4-6/1999-00 3 PM Fig.12 HMR 2010 WY noon Fig.13 Regressions HMR 2011 WY 2-3 PM Fig.14

Hydrograph rising limb sediment yields at HMR, Blackwood and Ward Creeks during spring snowmelt periods

HMR = 4.30x - 0.915 R2 = 0.932 Ward = 0.0073x3.1714 R2 = 0.8144 HMR = 4.4422x2.8127 R2 = 0.9539 Blkwd = 0.0007x5.9707 R2 = 0.9494

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5

Discharge (m3/s) Rising Limb Sediment Yield (kg/ha)

Ward '99 Ward '00 Ward '01 Blkwd '01 HMR '11

Summary of soils restoration work in the HMR Creek watershed (WS)

Summer- Year Type Area (m2) Roaded area Fraction (%) Ski-run area Fraction (%) Net WS Fraction (%)

2006 Road 2234 2.6

• 0.09

2007 Road 7483 8.9

• 0.37

2008 Road 4515 5.3

• 0.55

Road 4145 4.9

• 0.70

2009 Ski-run 3143

• 0.7

0.82 2010 Road 5603 6.6 1.04 Totals 27,123 28.4 0.7 1.04

Hydrograph rising limb sediment yields at HMR Cr. during spring snowmelt periods in 2010 & 2011

y = 4.30x - 0.915 R

2 = 0.932

y = 9.34x - 1.04 R

2 = 0.9646

y = 5.17x - 1.11 R

2 = 0.931

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20

Rising Limb Flowrate (m3/sec) Sediment Yield (kg/ha)

5/15 - 6/15/11 6/21 - 6/22/11 5/3 - 6/6/10 Adjusted 6'11

Where do we go from here ?

 Detection of soil restoration or disturbance effects are

difficult to measure at the watershed scale as affected areas are often small overall – nothing new there…!

 Original estimated load-flow relationships may slightly

• ver-estimate actual daily loads from Ward & Blackwood

creeks.

 Continuous flow/TSS monitoring through late spring

snowmelt period can assess changes in TSS, FSP and nutrient loadings following “treatments” within watershed.

 Measurement of daily hydrograph rising limb sediment

loads during the seasonal rising limb hydrograph may enable quantitative assessment of load reductions within specified confidence levels.

Related Papers



Grismer, M.E., C. Shnurrenberger, R. Arst and M.P. Hogan. 2009. Integrated Monitoring and Assessment of Soil Restoration Treatments in the Lake Tahoe Basin. Environ. Monitoring & Assessment. 150:365-383.



Grismer, M.E., Drake, K.M. and M.P. Hogan. 2010. Adaptive Management and Effective Implementation of Sediment TMDLs in the Lake Tahoe Basin. Watershed Science Bulletin. Fall (Oct.), pp.42-48.



Grismer, M.E. 2012. Erosion Modeling for Land Management in the Tahoe Basin, USA: scaling from plots to small forest catchments. Hydrological Sciences J. 57(5):1-20.



Grismer, M.E. 2012. Detecting Soil Disturbance/Restoration effects on Stream Sediment Loading in the Tahoe Basin – Modeling Predictions. Hydrological Processes. Submitted.



Grismer, M.E. 2012. Soil Disturbance/Restoration effects on Stream Sediment Loading in the Tahoe Basin – Detection Monitoring. Hydrological Processes. Submitted.