319(h) project meeting with Science Panel, June 8, 2011

Attending: Deanne DiPietro, Alex Young, Becca Lawton, Laurel Collins, Betty Andrews, Gretchen Hays (by phone), Caitlin Cornwall (recorder).

Presentation by Becca Lawton: Summary of the Sonoma Creek TMDL process

  • 303d listing driven by declining fish numbers.
  • Limiting factors analysis for steelhead: barriers, predation, low flows, sediment in gravels, loss of shelter, pool filling, increased suspended loads
  • Sediment source analysis: Natural sources about half from channels, human caused about ¾ from channels. 68,000 tons/year from channels, 64% of that human-caused.
  • TMDL plan: Required: sediment from road and stream crossings, etc. Recommended: habitat enhancement, reduce sediment discharge from channels.
  • Question now: how and where do we focus actions to reduce sediment loss from channels? Pilot study looked in upper watershed only, Glen Ellen and above. 319h project will improve methods from pilot study and expand to whole watershed.

Presentation by Alex Young: Pilot project

  • Used data that was available at the time.
  • Coarse filters: sediment production by subwatershed from sediment source analysis, locations of observed salmonid presence, instream habitat quality from DFG habitat survey protocol (impervious suface in subwatershed, canopy cover, shelter, frequency of primary pools, % ag by subwatershed, all ranked using NOAA Fisheries criteria).
  • Reach-level filters: long-term sediment supply by reach over last 150 years, willing landowners, overland erosion, road sources.
  • Overlaid all these layers, then selected 13 sites for field inspection.
  • Room for improvement:
  • much more sediment-related reach scale data, now all in GIS, than we’ve done so far.
  • We have more salmonid presence information to integrate
  • Making use of morphologic units

Presentation by Laurel Collins: Sediment data we have, and ideas for fieldwork

[missed first few minutes]

  • sediment trap analysis at bottom of watershed. Yield from sediment trap is only slightly smaller than yield from sediment source analysis, so there is only a small amount of storage.
  • Created morphologic units based on alluvium (on valley floor) versus geology (on slopes) and topography
  • Created a densified drainage network including ditches and storm drain systems (which system drains into Schell, not Sonoma)
  • Chose representative sites covering the range of stream orders in each morphologic unit. In field, estimated bank retreat and bed incision to estimate volumes of sediment lost. Also estimated % fines.
  • Results are in terms of the entire 150 years since European settlement.
  • Plot of drainage length versus sediment supply for each morphologic unit, shows some quite good regressions. Drainage length is better correlate than drainage area.
  • Plot of sediment supply and other variables against distance up Sonoma Creek, shows some possible causes of highs and lows.
  • Ideas for work under this project
  • We also have a hand-drawn map showing categories of incision depth along streams—not yet digitized.
  • Normalize sediment supply for certain types of channels, so we can see which locations are highly divergent from the norm.
  • Somehow estimate the elements of Rosgen stability classes (floodprone width, etc). Use estimate of top of bank, use FEMA 100-year width, use riparian width?
  • Assess impact of 2005/6 flood at more than the 3 sites already assessed.
  • Bankfull channel geometry, funded by EPA
  • Field-checking predictions of sediment supply
  • Assess whether high long-term supply sites are the same as the high short-term supply sites.
  • Map nickpoints, revetments (so we know we don’t need to work there)

Discussion

  1. What changes should be made to the draft project framework?
  2. What can we do with the data we have?
  3. What should our field effort look like?

Discussion about project goals

Project outputs

  1. top 10 sites for cost-effectively reducing fine sediment delivery, as high in the watershed as possible, and improving aquatic, riparian, and terrestrial habitat quality. Looking for sweet spots where there is high sediment supply but low entrenchment.
  2. model for others to benefit from
  3. a model for future use in Sonoma Valley, that is adaptable to future data and the effects of implementation actions

Process for site selection: input from Science Panel and field personnel  field work  GIS analysis  feasibility/fundability/bang-per-buck analysis

Strategies

  • Increase sediment deposition on benches
  • Increase width and cover of rip veg (in Napa, says Gretchen, sites with lots of veg, especially alder, changed the least during 05/06 flood)
  • Raise bed level?
  • Widen channels, even where they are entrenched, to give sediment somewhere to deposit
  • Accommodate a wide range of implementation actions, from classical restoration to flooding.

Model inputs that need to be built

  • riparian (deep-rooted woody or non-woody perennial) veg width or potential width
  • riparian maturity, since shallow roots are ineffective
  • availability of streamside land to widen riparian forest and/or lay back bank
  • plantability. E.g. can’t expose Huichica formation or plant on it
  • long profile or approximation. Highly recommended, but not funded. Possible short-cuts: riffle crests, PWA data, Laurel has something?, bridge as-built work done by Laurel and Arthur. Betty has historic FEMA profile data of patchy quality. Check out DFIRM data on GoogleEarth. Make profile complexity a screening step just before final selection.
  • Entrenchment. Estimate using simple inputs.
  • Risk of acute erosion from floods, especially considering climate change. Even where bankfull conditions look fine, unstable higher banks would go in a flood.
  • Normalized view of the current 150-year sediment volume data. Normalize by %fines, or by bankfull cross-sectional area?
  • Incision severity or risk. Incision is a bigger, more fundamental, problem than lateral bank erosion. Map where channel has eroded to bedrock (soft vs hard), plus all grade controls. Build our own data-driven classes for stability and channel type, instead of trying to fit into Rosgen’s.
  • Use recharge map to locate sites where overbank flooding would be desirable.
  • Likely spawning sites. Use grain size data from hab survey?

Data to be collected in the field

  • Regional curves [didn’t get what field data this needs]
  • Hard and soft bedrock
  • Nickpoints
  • Bank erosion hotspots
  • Revetments
  • A few more sites to assess 05/06 flood impact
  • Sites where sediment could be stored

Where to focus fieldwork

  • Tribs with fish (use ALL our info), morphologic units with high supply
  • Get input from field staff at SEC, RCD, SCWA, SDC, SLT, OSD, Bouverie)

Action items:

  1. Gretchen question for Laurel: in less-altered parts of the watershed, is drainage area a better correlate with sediment supply than drainage length?
  2. Gretchen: share Julie Beagle (spelling?) thesis on incision rates and timing of incision episodes, from Carneros Creek watershed
  3. Would be fun to ask community to tell us of erosion sites they know of.
  4. Follow up with Betty: how does what you heard today intersect with scoping study?
  5. Urge RCD to finish report from PWA on Schellville.