National Center for Earthsurface Dynamics

1
National Center for Earth-surface
Dynamics Our purpose is to catalyze development
of an integrated, predictive science of the
processes shaping the surface of the Earth, in
order to transform management of ecosystems,
resources, and land use.

• Three Integrative Projects
• Desktop Watersheds
• Stream Restoration
• Subsurface Architecture

2
Desktop Watersheds Goal

• To discover and advance the fundamental
  relations needed to predict landscape evolution
  and to model the coupling of ecosystem, landscape
  and land-use dynamics.

3
Desktop Watersheds

• Three Grand Challenges
• Development of a mechanistic understanding of the
  processes driving erosion and deposition that
  shape landscapes
• Development of a mechanistic understanding of the
  linkages between physical and ecological
  processes

Lidar-derived shaded relief map of the Fox Creek
area, ACRR
3) Application of understanding to the prediction
of the linkages between landuse and ecosystem
response to guide land management decisions
We focus these questions on channels and the
watersheds that feed them.
4
Conservation of mass equation
(1)
surface elevation change uplift spatial
gradient in sediment transport rate
Dietrich et al., 2003, AGU Monograph 135
 U f (?) Uplift field (not just vertical
component)   f (?) the notion of
Geomorphic Transport Laws Boundary and
initial conditions (history)
 U f (?) uplift field (not just vertical
component)   f (?) Geomorphic
Transport Laws Boundary and initial
conditions (history)
All of geomorphology in one equation!
5
1) What is the topographic signature of tectonic,
climatic, and other external influences on
watersheds?
2) How can the driving equations of landscape
evolution be scaled-up to coarse-grained
applications at large scales (such as entire
mountain ranges)? 3) Do biotic processes
influence large-scale topographic form?
6
(2)

• Where in the landscape do ecological regimes
  change, what factors cause these changes, and how
  would the locations of these ecological
  boundaries shift under altered climate, landuse
  or biological states?
• How does the physical organization of the
  landscape provide a template for organization of
  the ecologic and channel-scale processes in the
  watershed?

Power and Dietrich, 2002, Ecological Research
7
(3)
Goal Develop (process-based) models using
digital topographic and surface attributes (such
as geology, vegetation, and landuse) to predict
the linkages between landuse and ecosystem
function in order to guide management decisions.
Skid trails, NCED SF EEL lidar data
Debris flow tracks, 1996, Oregon Coast Range
8
Desktop Watersheds application
1. Extrapolate site specific information to
entire watersheds

• 2. Develop hypotheses about expected resource
  properties to guide field work (one approach
  analytical reference state)
• examples of resource properties
• landslide location
• river bed grain size
• abundance of wood in channel
• stream temperature

9

• Becomes a null hypothesis to guide field work
  it is a state against which to detect and measure
  deviations from predictions caused by processes
  not included in the model

10

• Dynamic models of linkages between landuse and
  resource state (the Holy Grail of theoretical and
  applied studies)
• examples
• effects of forest clearing on runoff magnitude
  and channel dynamics
• sediment production and routing through channel
  network

11
BACKGROUND
A term that has been used for over 10 years with
varying enthusiasm is Watershed Analysis Goal
in Watershed Analysis is to establish causal
relationships between and among physical and
biological (and chemical) processes in order to
assess the effects of past and future land use
activity and to guide montioring and restoration
of watersheds and their ecosystems. What
was? What is? What could be? What now? What
happened (after restoration occurs)
Resource assessment
Prescription
Adaptive management
12
Physical and biological processes not linked
13
Elements of a desktop watersheds model directed
at salmon
W. Dietrich and F. Ligon, in prep
14
ALSM
Airplane or helicopter
33khz UF plane
15
First Surface
Bare Earth
Contour Map
16
(No Transcript)
17
(No Transcript)
18
National Center for Airborne Laser Mapping
Dietrich and Perron, in press, Nature
Collin Bode
19
Helicopter lidar survey
20
Large landslide on Ten Mile Creek. Possibly
leading to downstream fill terraces.
21
(No Transcript)
22
channels
roads
Channel head
23
Desktop Watersheds application
1. Extrapolate site specific information to
entire watersheds

• 2. Develop hypotheses about expected resource
  properties to guide field work (one approach
  analytical reference state)
• examples of resource properties
• landslide location
• river bed grain size
• abundance of wood in channel
• stream temperature

24
Power, et al., 1998, Experimental Ecology

• Example digital terrain model as field guide
• Analytical reference state
• Idealized, quantitative statement of an expected
  condition based on simple observable properties
  such as topography and general climate setting
• Key assumption relatively immutable properties
  of a specific watershed, such as its topography,
  drainage area to a point, local slope, aspect,
  and climatic setting can be quantified and used
  in simple mechanistic models to calculate an
  expected condition in the system
• A theoretical condition that can be calculated
  for any watershed, but will differ between
  watersheds depending on the intrinsic properties
  of the watershed
• Becomes a null hypothesis to guide field work
  it is a state against which to detect and measure
  deviations from predictions cuased by processes
  not included in the model

25

• Simple digital terrain models (some examples)
• Location where shallow landslides are most likely
• Channel head location
• Channel morphology from local slope and drainage
  area
• Median grain size of river bed surface throughout
  channel network
• Stream temperature

26
Shallow landslide location
SHALSTAB (Dietrich et al., 2001,AGU, WSA
various) Sinmap (Pack, D. Tarboton others R.
Sidle (various papers) R. Iverson (various
papers)
SHALSTAB a model for estimating likely
locations for shallow landslide based on linking
a steady state subsurface flow model and an
infinite slope stability model
27
Relative potential for shallow landsliding
Based on 2 m airborne laser swath mapping data
Based on 10 m USGS topography
Coos Bay, Oregon
Dietrich, et al., 2001, AGU, WSA
28
1)     
29
Landslide scars
Channel stops
On-channel reservoir
Channel head
30
Channels are estimated to occur where Agt5000
m2, and A/(S2b)gt25 m or Agt100000 m2
A drainage area (m2) S local surface
gradient b cell size (m)
USGS bluelines
Agt5000 m2
Combined thresholds
31
(No Transcript)
32
(No Transcript)
33
gt 20 colluvial
Upper limit of fish habitat
Steelhead?Chinook
 
 
lt .1
sand
Modified from Montgomery and Buffington, 1997
34
Channel slopes for segments average 16 m long for
entire basin and 68m long for slopes less than 8
Napa River Watershed 1000 km2 of 1.5 m or so
digital data Work done to assist TMDL studies
35
Bar-pool
chinook
steelhead
Step-pool
Ephemeral- non-fish habitat
Colluvial -cascade
Channel bed morphology based on local slope
36
Of the 5598 km (3479 mi) of channels estimated to
be in the Napa watershed, 70.1 is steeper than
8 and is unlikely to provide fish habitat. The
remainder (about 1040 mi of channel) is divided
roughly equally into that which is more favorable
to Steelhead and that which is more favorable
habitat for Chinook.
37
Estimation of median grain size of the bed
(threshold channel) Assume median particle size
just starts to move when flow reaches a bankfull
stage which typically has a peak flow recurrence
interval of 1 to 2 years. Assume a threshold
Shields number (K) for initial motion and solve
for grain size Estimate depth based on
empirical depth-drainage area relationship and
use local slope Method tends to over predict
grain size because effects of bed topography
roughness, wood, irregular banks, and other forms
of resistance are not considered.
38
Median grain size of the river bed (mm)
Calculations on slopes less than 8. On steeper
slopes theory fails
39
(No Transcript)
40
Buffington et al., 2004
41
Buffington et al., 2004
42
Solar Irradiation influence of topography
Predicted stream temperature under current
conditions
Douglas Allen, et al, in press
1-D steady state numerical model for hottest day
in summer
43
Predicted stream temperature with mature forest
blocks (shaded area) and riparian forest
restoration
Predicted stream temperature under current
conditions
44
Lack of Quantitative Relationships
?
Habitat Parameter
Sediment Input
45
Napolitano et al., in review
46
Growth in length and mass decreased linearly with
proportion of fine sediment in the bed
Length gain (R2 0.63, P lt 0.0001). Mass gain
similar (R2 0.59, P lt 0.0001).
Subtle et al.
47
Power et al. analysis of network structure of
food webs at the ACRR
48
(No Transcript)
49
South Fork Eel, TMDL analysis
50
A Desktop Watershed Model for Salmon
Spawning success
Summer and winter rearing success
Collaboration with Partner Stillwater Sciences
Dietrich and Ligon, in prep
51

• Dynamic models (some examples)
• Runoff models (numerous models exist)
• Slope dependent mass transport by soil
  (relatively advanced)
• Storm-driven slope stability and prediction of
  landslide size (some startsflux is much harder
  to predict than location)
• Stochastic debris flow runout, entrainment, scour
  and deposition (some starts)
• Storm-driven Horton overland flow erosion (still
  very weak on this!!)
• For all hillslope processes above prediction of
  sediment size distribution and durability of in
  put(anyone done something?)

Dietrich and Ligon, in prep
52
(No Transcript)
53
Dynamic (continued examples) 7) Channel sediment
routing (storm-based, accounting for storage,
sorting, and particle breakdown) (Yantao Cui et
al. have started on this, among others) 8)
Prediction of effects of sediment supply on
biologically significant channel attributes
(surprising lack of field data or theory) 9)
Prediction of network-based ecosystem communities
(some starts). 10) Dynamic fish equation (a long
term goal)
54

• Some needs
• Improved tools to use high resolution topographic
  data
• Field data quantifying the linkages between the
  Geo , the Bio and the POP
• Field data relating changes in river currencies
  (heat, wood, sediment, nutrients, runoff) and
  river ecosystem state
• 4) Field evaluated theory to guide the
  development of dynamic models