Suspended Sediment Loads from Bay Area Small Tributaries

1
Suspended Sediment Loads from Bay Area Small
Tributaries
Item 3a

• Mikolaj Lewicki and Lester McKee
• Sources Pathways and Loading Workgroup
• May 14th 2008

2
Project Impetus

• Hg and PCB load calculations in the TMDLs are
  based on estimates of suspended sediment loads
• Krone 1979 (750,000 t)
• Davis et al., 2000 (320,000 t)
• Bay modeling would benefit from accurate
  spatially explicit estimates
• Wetland restoration projects also require
  sediment source information

3
Available Data Sets
Item 3a

• 29 USGS Sediment Gages
• 135 years of record for all the stations

• Does it mean that we have enough information ?
• Is the field data collected in the locations that
  provide the most information
• Is the distribution of sampling stations dense
  enough to include the diversity in land use,
  lithology, or physiography?

4
Suspended Sediment Discharge from the Ungaged
Watersheds
Item 3a

• Potential statistical relationships between a
  watershed characteristic and suspended load
• Drainage Area
• Annual Discharge
• Annual Peak Discharge
• Land Use
• Geology
• Slope
• Other

5
Drainage Area vs. Suspended Load
Item 3a

• There is a weak relationship between drainage
  area and the load. Correlation is better for more
  uniform geomorphic/climatic provinces

6
Which variable is statistically the most
significant for the total annual load
predictions?
Item 3a

• Geology

• Drainage Area

• Annual Discharge

• Land Use

• Watershed average slope ?

• Annual Peak Discharge

• Unit Annual Peak Discharge Correlated with Unit
  Sediment Load

• And the winner is

7
Unit Annual Peak Discharge Correlated with Unit
Sediment Load East Bay
Item 3a
Very Good correlation R2 0.94
8
Unit Annual Peak Discharge Correlated with Unit
Sediment Load North Bay
Item 3a

• Very Good correlation R2 0.92

9
Unit Annual Peak Discharge Correlated with Unit
Sediment Load San Francisco Peninsula
Item 3a

• Correlation could be improved by low
  discharge sampling and better understanding of
  urban watersheds sediment loads

10
Unit Annual Peak Discharge Correlated with Unit
Sediment Load San Francisco Bay
Item 3a
11
Other Statistical Correlations
Item 3a
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for Each of the SF Bay Counties
12
Item 3a
Other Statistical Correlations
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for Each of the SF Bay Counties
13
Other Statistical Correlations 2
Item 3a
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for different types of Geology
(classified based on the rock formations
resistance to erosion)
14
Item 3a
Other Statistical Correlations 2
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for different types of Geology
(classified based on the rock formations
resistance to erosion)
Lower sediment discharge from resistant geology
15
Item 3a
Other Statistical Correlations 2
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for different land use types

• Each watershed was divided into fife generalized
  land use classes (based on NLCD data set)
• Natural
• Agriculture
• Low density urban
• High density urban
• Industrial

16
Other Statistical Correlations 238
Item 3a
Unit Annual Peak Discharge Correlated with Unit
Sediment Load for different land use types
Natural land use type has lower sediment
discharge compared with urban land use type
17
Remaining Work
Item 3a
Correlate Unit Annual Peak Discharge with Unit
Sediment Load for different median watershed
slopes
Correlate Unit Annual Peak Discharge with Unit
Sediment Load for areas with high number of
construction sites
Log Ys 0.059C-0.047 Ys suspended sediment
yield in tons / acre C percentage of land
under construction
18
Item 3a
Remaining Work
Correlate Unit Annual Peak Discharge with Unit
Sediment Load for mature urban watersheds
Croplands 3.6 tone / acre/year
Grass Areas 0.2 tone / acre/year
Natural Areas Sediment yields 0.02 - 0.03 tone
/ acre/year
Urban / Residential Areas 3.9 tone / acre/year
Correlate Unit Annual Peak Discharge with Unit
Sediment Load for watersheds with land slides or
forest fires
Both processes are expected to increase sediment
loads at least 30 times compared to the
background values
19
Questions / Comments
Item 3a