Soil and Water Assessment: dynamic on a daily basis

1
Soil and Water Assessment dynamic on a daily
basis Jeff Arnold, Jimmy Williams, Raghavan
Srinivasan, Jim Kiniry, Cole Rossi With thanks
to Nancy Sammons and Georgie Mitchell
2
History

• USDA-ARS Support 30 Years of Model
  development
• ? EPICField Scale
• ? ALMANACField Scale
• APEXFarm Scale
• ? SWATWatershed Scale
• University Support Texas A M, many others
• International Conferences and Development
  Collaboration
• Soil Chemistry

3
Arc-Apex Arc-SWAT, VizSWAT/VizAPEX Jimmy
Williams, R. Srinivasan, etc.

• GIS interface
• Spatially distributed
• can set up APEX, SWAT, and/or hybrid
• utilize national databases (weather, soil)
  readily
• Can incorporate up-to-date SWAT model alterations
  (C-Farm, Metals, landscape version)
• Available by summer, 2008

4
General Description

• ? Continuous Time
• Daily Time Step (Sub-hourly)
• One Day Hundreds of Years
• ? Distributed Parameter
• Unlimited Number of Subwatersheds
• ? Comprehensive Process Interactions
• ? Simulate Management

5
(No Transcript)
6
Hydrologic Balance
Evaporation and Transpiration
Precipitation
Surface Runoff
Root Zone
Infiltration/plant uptake/ Soil moisture
redistribution
Lateral Flow
Vadose (unsaturated) Zone
Revap from shallow aquifer
Percolation to shallow aquifer
Shallow (unconfined) Aquifer
Return Flow
Confining Layer
Flow out of watershed
Deep (confined) Aquifer
Recharge to deep aquifer
7
Channel Processes
8
SWAT C is static APEX Daycent several C
pools C-FARM A Simple Model to Evaluate the Soil
Carbon Balance in Cropping Systems (incorporation
into SWAT) Armen Kemanian TAES, TX Stefan
Julich Giessen, Germany Cole Rossi ARS,
TX And of course, Jimmy Jeff
9

• C-FARM continued
• NP driven by SOC CN CP in 110100 fixed
  ratios in general
• Residue inputs vary (wood to manure), therefore
  when it starts decomposing the CNP ratio must be
  driven to that of the OM via N, P immobilization
  or mineralization
• C-FARM checks for both N P availability to
  determine if the C residue decomposing can
  proceed at max speed as set by soil water and
  soil temperature

10
Analytical solutions
C-FARM

• Change in Carbon Storage Inputs - Outputs
• Hénin and Dupuis (1945) dCs/dt hCi kCs
• Cs is the soil organic Carbon (Mg ha-1)
• t is time (year)
• h is the humification constant
• Ci is the carbon input
• k is the apparent soil turnover rate

11
Testing Rothamsted UK continuous wheat
C-FARM

• Treatment 144 kg N ha-1, no residue burn
• 1853 1926 continuous wheat
• 1927 1962 wheat fallow
• 1963 2005 continuous wheat
• Average aboveground carbon input approximately
  2.2 Mg ha-1 year-1

Treatment 0 kg N ha-1, no residue burn 1853
1926 continuous wheat 1927 1962 wheat
fallow 1963 2005 continuous wheat Average
aboveground carbon input approximately 1.2 Mg
ha-1 year-1
12
The core carbon balance equation for each layer
C-FARM

• dCs/dt hCi kCs
• h hc1 (Cs/Cx)n
• k feftkx(Cs/Cx)mCs
• hc depends on soil texture resembling Roth-C
  (MODEL FROM ROTHMASTED JENKINSON)
• Cx depends on soil texture (Hassink and Withmore,
  1997)
• fe soil temperature and water content factor
  (energy balance)
• ft is a function of tillage tool and number of
  operations (NRCS)

13
Nitrogen Cycle
Atmospheric N fixation (lightning arc discharge)
Harvest
NH3
Symbiotic fixation
N2 N2O
fertilizer
fertilizer
runoff
manures, wastes and sludge
ammonia volatilization
denitrification
Soil Organic Matter
mineralization
immobilization
NH4
immobilization
NO3-
NO3-
NO2-
ammonium fixation
anaerobic conditions
clay
nitrification
leaching
14
runoff P (sediment bound)
runoff P (sediment bound)
Manure (surface applied incorporated)
active organic P
active organic P
soluble P
stabile organic P
soluble P
stabile organic P
manure
soil
soil
sediment bound runoff P via lateral flow (tile
flow possible)
leaching
leaching

• New P DiagramWork in Progress

15

• The new SWAT P routines allow for soluble P
  movement throughout the soil profile. Soluble P
  can be distributed with lateral flow, including
  tile flow, in groundwater, surface flow, and in
  percolate throughout the soil profile.

16
P developments

• SWAT monitors 6 different P (same for N) pools in
  the soil 3 organic (active, stable, fresh) 3
  mineral (active, stable, solution)
• Users can 1) define initial concentrations 2)
  utilize SWAT initialization of P pools using PAI
• Have 2 sorption coefficients one for the top 10
  mm and one for the layers below (phoskd,
  phoskdsub)
• P can now be leached through the entire soil
  profile (crack flow, sandy textures)
• P can now be transported through tile drains
• In-stream P change with addition of sediment
  classes

17
Water, Nitrogen and Phosphorus Uptake
18
Pesticide Dynamics
Foliar Application
Surface Application
Degradation
Runoff
Washoff
Infiltration
Degradation
Leaching
19
What can be simulated?

• Cropping systems with full crop rotation and farm
  management
• Planting / Beginning of growing season
• irrigation
• fertilisation
• pesticide application
• tillage operation
• harvest and kill (above-ground biomass is
  removed)
• kill/end of growing season (above-ground
  biomass becomes litter)
• grazing (biomass removal and manure input)
• auto irrigation
• auto fertilisation
• street sweeping
• release/impound (for rice)

20
Incorporation of Landscape Processes in a
Watershed/River Basin Scale Model
Jeff Arnold, Peter Allen, Martin Volk, Jimmy
Williams, Dave Bosch, and Cole Rossi
Grassland Soil and Water Research Laboratory
21
Landscape Modeling Approach
22
Landscape Units
23
Landscape Routing

• Landscape Positions
• (Flood Plain, Hillslopes, Divide)
• Riparian Zones

Divide HRUs
24
Landscape Delineation Slope
Position Method
Valley Downhill Flow Accumulation Ridge
Uphill Flow Accumulation
25
Y-2 Weir
26
Lateral Soil Flow
Measured 106 mm/yr Simulated 102 mm/yr
27
Soil Moisture
Ridgetop
Hillslope
Valley
28
Future Developments
Kinematic Wave Eqn for
Overland and Channel Routing between
Landscape Units Sediment and Nutrient Routing
Across Landscape Testing at Gibbs Farm
Watershed in Tifton, GA with
Riparian Zones Testing on Larger Watersheds with
Defined Flood Plains GIS Interface and
Documentation
29
 

• FOREST GROWTH in SWAT
• Grow from seedlings to maturity
• Litter layer and wetlands Ruth McKeown
• Link with ALMANAC (Agric. Land Mgt Alternatives
  with Numerical Assessement Criteria-Kiniry et
  al., 1992) plant growth simulates plant
  competition Jim Kiniry and Doug McDonald
• SWAT2009
• Code development for Hg, Mn, Cu, As, Al and Fe
  (EPA HAWQS)-in progress
• CFARM C,N,P, roots inclusion into SWAT (ASABE,
  Chile)
• Inclusion of Hooghoudts Kirkham Tile Eqns
  (Drainmod)-Daniel Moriasi

 
30
Forest Growth in SWAT (b/f change)

• Plant growth in SWAT assumes a uniform, single
  plant community
• Contains the basic parameters from ALMANAC to
  simulate plant growth (ALMANAC utilizes water
  balance, nutrient balances, and soil erosion from
  EPIC)
• The HRU (unique combination of soilland
  useslope) allows for a delineation of forest
  stands
• Can grow trees from seedlings to maturity (no
  competition) based on PHU
• Watershed-scale can capture forest mgt as a fxn
  of site characteristics

31
Forest Growth in SWAT Improved

• Can simulate complex successional changes in
  forest ecosystems (crops, crop trees, forbs,
  grasses, and woody species) including
  disturbances regrowth, mixed canopes,
  understory growth, and control of undesirable
  plants/weed infestations over long periods of
  time (inlcluding initial stages after disturbance
  and forest immaturity)
• ALMANAC has the ability to accurately simulate
  competition for light, nutrients and water for
  several plant species based on interaction of
  light partitioning eqns with long-term leaf area
  development eqns algorithm that limits
  potential leaf area based on species height and
  leaf area index in previous years

32
ALMANAC SWAT cont

• Algorithms include impact on water quality and
  quantity daily time-step
• Inclusion in SWAT can capture important forest
  dynamics without excessive simulation time on the
  complexities of forest growth (based on RUE, LAI,
  HI, PAR-photosynthetic active radiation)
• Can capture tree thinning, tree productivity,
  competition intensity
• Uses sigmoid eqns to describe long-term height
  and leaf area growth using year as dependent
  variable vs. PHU

33
 

New in SWAT2005
AUTOCALIBRATION UNCERTAINTY ANALYSIS AVSWAT
Interface Ann van Griensven and Mauro
DiLuzio Documentation is not complete Testing
is ongoing
 
 
 
34

35
Current Uses

• CEAP National Assessement Initial effort will
  include for cropland/CRP adding rangeland CEAP
• CEAP WAS
• EPA HAWQS
• Mekong River Basin
• TMDLs
• SWIM hydrotopes

36
(No Transcript)
37
Mekong River Basin and Mekong River profile from
headwaters to mouth (from MRC Overview of the
Hydrology of the Mekong basin, Vientiane, 2005
report).
38
Always improving

• Impact of soil structure on root distribution
• (SWAT assumes homogeneously distributed)
• More accurate values for root and leaf resistance
• to water flow based in measurements
• (Leaf resistance should vary with water
• stress-cavitation in xylem)
• Minimize complexity of model increase detail
  while
• maintaining user-friendly program
• Channel deposition and scour
• Data and more data re soil moisture,
• sediment transport from terrestrial environment
• to stream transport, landscape position-dependent
  data
• .More collaboration