Development of a Common Wind and Water Erosion Model

1
Development of a Common Wind and Water Erosion
Model

• Dennis C. Flanagan
• Agricultural Engineer
• USDA-Agricultural Research Service
• National Soil Erosion Research Laboratory
• West Lafayette, Indiana, USA

2
Outline

• Brief history of erosion prediction technology
  development in the U.S.
• User needs for a common water and wind erosion
  model.
• Plans for model development.
• Progress to date.

3
History of Erosion Prediction Technology
Development in U.S.

• Universal Soil Loss Equation (USLE) developed
  from 1954 to 1978.
• Revised Universal Soil Loss Equation
  (RUSLE/RUSLE2) from 1987 to 2002.
• Wind Erosion Equation (WEQ) from 1953 to 1965
  and Revised WEQ from 1990 to 1998
• Wind Erosion Prediction System (WEPS) model from
  1985 to 2007
• Water Erosion Prediction Project (WEPP) model
  from 1985 to 2007

4
USLE

• Developed from 1954-1978 to predict long-term
  average annual soil loss on hillslopes.
• Soil conservation experiment station data from
  the 1930s to 1950s was utilized in its
  development (over 10,000 plot-years).
• First publication on USLE was in 1961.
• Implemented in SCS field offices during the
  1960s.
• USLE is an empirical model
• A R K L S C P

5
RUSLE / RUSLE2

• Developed from 1987 to 2002.
• RUSLE was implemented in SCS field offices in
  paper form only in 1993.
• RUSLE2 was implemented in NRCS field offices in
  2002.
• RUSLE/RUSLE2 are empirical models with some
  process-based enhancements.
• Improvements to USLE R, K, C factors.
• Extremely large management rotation databases for
  every state in the U.S.
• Maintained by ARS-Oxford, MS.

6
WEQ / RWEQ

• WEQ was developed from 1953 to 1965 to predict
  soil loss from wind erosion on an average annual
  basis.
• WEQ was adopted by the SCS for predictions of
  soil erosion by wind, and is used mainly in the
  western U.S.
• The difficulty of use of WEQ prompted development
  of several computerized versions by both ARS and
  SCS/NRCS.
• A revised wind erosion equation (RWEQ) was
  developed by ARS from 1990 to 1998, but no
  widespread adoption by NRCS is planned.
• RWEQ is maintained by ARS-Lubbock.

7
WEPS

• Process-based, continuous simulation, wind
  erosion model.
• Developed from 1985-2007 by ARS-Wind Erosion
  Research Unit (WERU) in Manhattan, Kansas.
• Field testing and implementation by NRCS has
  begun in last 2 years.
• Recently WEPS has incorporated WEPP model
  hydrology to decrease run time.

8
WEPS Windows Software
9
WEPP

• Process-based, continuous simulation water
  erosion model.
• Developed from 1985-2007 by ARS, SCS/NRCS, FS,
  BLM, WSU and others.
• WEPP model is maintained by ARS-National Soil
  Erosion Research Laboratory in West Lafayette,
  Indiana.
• Large number of users, both within and outside
  U.S., including Forest Service, BLM,
  universities, consultants.

10
WEPP status

• Current public model version is v2006.5
• V2006.5 contains recent updates to water balance,
  subsurface lateral flow, perennial plant growth
  to better simulate forests on shallow soils above
  bedrock.
• Variety of user interfaces Windows-based,
  Web-based, and GIS-linked.

11
WEPP Windows Interface
12
WEPP Web-based Interfaces
13
Separation of Wind / Water Research

• Initial research studies were focused in areas
  with specific erosion concerns water erosion in
  the eastern and central U.S., and wind erosion in
  the Great Plains.
• ARS programs, experiment stations, research
  units, and funding were separated between water
  and wind erosion locations since the inception of
  the agency in 1953.
• Process-based modeling efforts that began in 1985
  (WEPP and WEPS) were for the most part separate,
  due to the existing institutional framework.

14
This Separation resulted in

• Two separate teams of ARS scientists building
  continuous process-based simulation models.
• Two models that were required to simulate many of
  the same physical processes (soil water balance,
  hydrology, plant growth, residue decomposition,
  soil disturbance by tillage, etc.).
• Separate model interfaces and databases.
• Large potential for different model results (for
  crop growth, runoff, etc.) for same site of
  application, since different science implemented
  in the two different models.

15
In 2004

• The Natural Resources Conservation Service
  re-evaluated its need for erosion prediction
  technology from ARS.
• High priority long-term need of NRCS was
  development of a common wind and water erosion
  process model, to work with a single interface
  and database and give consistent results for
  plant growth, water balance, crop yield, etc.

16
From March 1, 2004 Letter from NRCS

• For the long term, NRCS proposes to collaborate
  with ARS to build a single process based model to
  make erosion prediction calculations. NRCS
  proposes that this model be capable of making
  rainfall induced rill and interrill erosion
  computations, as well as computations for wind
  erosion together or independently of one another.
  This model would naturally incorporate the
  technologies currently in WEPS, the Water Erosion
  Prediction Project (WEPP), and those found in the
  Water Erosion Prediction Project - Simulation of
  Production and Utilization of Rangelands
  (WEPP-SPUR). Unlike the current models, the
  model proposed by NRCS would operate as a single
  decision support tool, and use common databases.
  
• - Larry Clark, NRCS Deputy Chief Science
  Technology

17
2004 ARS NSERL Erosion Prediction Program
Redirection

• Modification of NSERL erosion prediction CRIS
  research project
• Stop new development work on existing WEPP model
  science and interface code
• Minimize resources towards current WEPP model
  code and user support
• Focus majority of resources towards development
  of new wind and water erosion model.
• Top short-term priority - Incorporation of WEPP
  hillslope erosion science within the Object
  Modeling System (OMS) being developed by
  ARS-GPSRU in Fort Collins, CO.

18
2004 Project Objectives

• Short-term
• Incorporate the WEPP hillslope erosion code
  within OMS.
• Evaluate the feasibility of using OMS as the
  platform for the full combined wind and water
  erosion model.
• Develop a complete project plan for development
  of the new model.
• Long-term
• Develop a fully functional continuous simulation
  wind and water erosion process model for field
  application by 2011.

19
New CRIS project

• Entitled Common Modular Wind and Water Erosion
  Modeling for Conservation Planning
• Recently approved (12/2006) through OSQR
• 2006-2011 Develop a common wind and water soil
  erosion model for use by NRCS field offices.
• Utilize water erosion components from WEPP model,
  and wind erosion components from WEPS model.
• Develop necessary interfaces and databases for
  the new modeling system.

20
New CRIS Project Objectives

• Integrate the WEPP and WEPS model erosion
  technologies through the use of the Object
  Modeling System (OMS) into a single wind/water
  erosion prediction system using common databases
  and interfaces at the plot and field scale.
• Incorporate, test and verify new erosion science
  or related components, such as winter processes,
  tillage erosion, ephemeral gully erosion,
  irrigation erosion and rangeland erosion, into
  the integrated erosion prediction system.
• Cooperate with all ARS scientists and NRCS staff
  involved with the CEAP effort to extract relevant
  modules from existing models and integrate them
  into the OMS for development of regional water
  and air quality models at the plot, field, and
  watershed scales.

21
New CRIS Project Milestones

• 12 months
• Development of wind detachment component in OMS
• Testing/validation of hydrologic water erosion
  prototype
• Develop user requirements for system with major
  user agencies
• Addition of dynamic water erosion calculations
• 24 months
• Development of detailed software design document.
• Unified Plant Growth Model incorporated into OMS.
• Prototype OMS wind-water model with most needed
  components.
• Validation of single event wind erosion
  predictions.
• Addition of tillage erosion modules into OMS.

22
New CRIS Project Milestones

• 36 months
• Development of ARS interface for
  testing/validation
• Development of core combined model databases
• New winter components added into OMS
• Addition of tillage erosion simulation into
  wind-water model
• 48 months
• Development of NRCS interfaces for model testing
  and training.
• Validation studies on wind, water and tillage
  erosion predictions.
• Addition of rangeland components
• 60 months
• Integrated field-scale erosion prediction system
  initially for cropland applications and
  prediction of wind, water or tillage erosion
  delivered to NRCS.
• Model technical and user documentation written.
• Creation of irrigation erosion modules.
• Testing of rangeland and irrigation erosion
  modules
• Prototype regional field-to-watershed model for
  CEAP, integrating appropriate modules from
  wind-water system.

23
Two Development Paths

• First Path Extract individual components from
  WEPP and WEPS and other relevant models. Develop
  modules within OMS from these components, then
  build new model within OMS. (as written in Plan)
• Second Path Utilize WEPS model code as the
  basic framework and add WEPP model water balance,
  runoff, water erosion components.

24
Object Modeling System (OMS)
25
What is the Object Modeling System?

• An object-oriented toolset to build, run, and
  deploy simulation models
• An object-oriented framework for the management
  of reusable simulation component libraries
• A collaboration infrastructure for common model
  development

26
Object Modeling System

• Modeling framework to support the model
  development/application lifecycle
• OMS Facilitates
• Code reuse and sharing
• Capture of legacy knowledge
• Collaborative development
• Database access
• Verification/validation
• QA/QC
• Maintenance and change management

27
Components in OMS

• Scientific component
• Infrastructure component
• Utility component

• Components are software units that are
  context-independent both in the conceptual and
  technical domain
• Well adopted methodology for software reuse

Internal hidden behavior
Component
Input
Output
Well known interfaces
Component
28
Scientific Component

• Represents a basic processing unit
• Conceptual purpose
• Runoff computation
• Soil erosion computation

Component
Input
Output

• Components are tagged by implementing
  interfaces
• Native, Runable, Stateful, Visualizable
• Customization by implementing these interfaces
• Minimum Runable

Component
29
Step 1 Create Components
Input Data
Create Components
30
Step 2 Create Model
Component Library
New Model - Training
31
Step 3 Build Model From Components
Attributes
Component Connectivity (Hookups)
Model Building Structures
32
Step 4 Run Assembled Model
33
Modeling Projects
Component Editor
Output Analysis
OMS also has output graphic and parameter editing
capabilities
Component Library
Parameter Editor
Assembled Model
34
OMS Workflow Summary
Component Builder
Model Builder
Output Analysis
Analyze
Execute
Publish
Integrate
Component Library
Model Runtime
Data Analysis
Model Application
Component Integration
Component Library Management
Component Development
35
(No Transcript)
36
First Development Path

• Extract individual components from WEPP and WEPS
  and other relevant models. Develop modules
  within OMS from these components, then build new
  model within OMS.
• Advantages modular approach best for long-term
  agency code maintainability, can access and use
  existing components in OMS library, NRCS desires
  new model development in OMS, multiple spatial
  representations for wind and water may be easier.
• Disadvantages - OMS system not fully developed
  and easy to use, incorporation of legacy models
  in OMS can be difficult and time-consuming,
  agencies continued support of OMS is uncertain.

37
Path 1 - Progress to Date

• WEPP hillslope water erosion code extracted and
  stand-alone program created (2004).
• WEPP hillslope surface hydrology (infiltration,
  runoff) extracted and stand-alone program created
  (2005).
• Stand-alone hydrology and erosion code converted
  to components in OMS (2005).
• Single storm and continuous hydrology/erosion
  model created in OMS (2005).

38
Approach

• Initially in 2004-2005
• Convert hillslope erosion component from WEPP
  into a standalone Fortran program.
• Test and verify standalone program against
  original WEPP v2004.7 model
• Incorporate standalone program into OMS, test and
  verify.

39

• September-October 2004
• Extracted relevant hillslope erosion code from
  WEPP v2004.7 for single storm.
• Removed all common blocks and moved only
  necessary variables into argument lists.
• Created input files to just conduct single storm
  water erosion calculations.
• Tested standalone for range of inputs slope
  lengths, gradients, and shapes and compared to
  WEPP v2004.7 output.
• This resulted in corrections to the standalone
  code and ultimately a verified single storm
  program that operated for a single spatial plane.

40
Results of Final 10/2004 Standalone Verification
Tests
41

• January-February 2005
• Made code active to handle multiple spatial
  planes.
• Modified input files to provide information
  necessary for multiple planes.
• Tested standalone for range of inputs 1, 2, 4,
  10 overland flow elements and compared to WEPP
  v2004.7 output.
• This resulted in corrections to the standalone
  code and ultimately a verified single storm
  program that operated for a multiple spatial
  planes. This contained spatial looping similar
  to WEPP in the standalone MAIN program.

42
Standalone Fortran erosion code 2/2005

• MAIN program and 30 subroutines under it. MAIN
  contained spatial plane (iplane) looping similar
  to WEPP.
• Reads from a single input file and creates 2
  output files, almost identical to current WEPP
  outputs.
• Will compile and run with standard F-77 to F-95
  compilers.

43

• March-April 2005
• Spatial looping in MAIN program and all
  subroutines removed, so that code could be better
  utilized within OMS with other models, as well as
  with potential spatial representation needed for
  wind erosion.
• Existing standalone Hydrology component (based
  largely on WEPP) from Ascough was converted into
  a format ready for OMS inclusion.
• Pass file creation with information generated by
  Hydrology standalone and needed for Erosion
  standalone calculations was added to Hydrology
  code.
• This work resulted in standalone Hydrology and
  Erosion code that would function in tandem to do
  infiltration/runoff calculations (Green-Ampt),
  runoff hydrograph and peak rate calculation
  (kinematic wave), and hillslope interrill/rill
  erosion calculations for a single storm/single
  plane.

44
Standalone Fortran hydrology code 4/2005

• MAIN program and 19 subroutines under it.
• Reads from a single input file and creates 2
  output files hydrology output identical to WEPP,
  and a hydrology-to-erosion pass file (runoff
  depth, peak rate, intensities, durations).

45

• June 2005
• Individual Hydrology and Erosion Models were
  created within OMS, tested and verified against
  the standalone programs.
• A linked hydrology and erosion model for a single
  storm and spatial plane was created in OMS.
• The linked OMS model was expanded to successfully
  perform spatial (multiple planes) and temporal
  (multiple storm days) looping.

46
OMS individual Hydrology and Erosion Models
6/2005

• To build models in OMS, the functionality in the
  standalone Fortran MAIN programs had to be
  duplicated.
• All processing logic in existing MAIN had to be
  either moved to one of the existing subroutines,
  or a new component created.

47
OMS with 6/2005 Temporal/Spatial Erosion Model
Erosion Model
Output window showing model screen outputs
48
Erosion Model named erroder here
Conditional Daily Time step goes through loop
for number of days read in from climate file.
Conditional to do infiltration, runoff and
erosion calculations, only if there is rainfall
on the day
Conditional to do Erosion Calculations only if
there is outflow from or inflow to plane.
49
Path 1 - Progress to Date (cont.)

• Continuous water balance model (based on RZWQM)
  constructed in OMS, and hydrology and water
  erosion modules linked into this (2006).
• Wind detachment stand-alone code from WEPS
  converted into an OMS module (2006).
• Prototype wind and water combined model
  constructed in OMS, linking wind detachment
  module with water balance model (2006).

50
10/2006 Wind Water Model Prototype in OMS
51
Wind Water Model Prototype in OMS
Water Balance Initializations
Daily loop
Potential Evapotranspiration Calculations
24 Hour loop
Infiltration Calculations
Soil Water Redistribution
Erosion by Water Components
Erosion by Wind Component
52
OMS Model water erosion outputs
53
OMS Model wind erosion outputs
54
Future OMS Work

• Add more components (plant growth, parameter
  estimation, etc.) to library and link these with
  existing prototype.
• Properly set up spatial looping to represent both
  a gridded wind detachment region and a water
  erosion hillslope profile.
• Test OMS models against original WEPP/WEPS
  models, and compare to field experiment data.

55
Second Development Path

• Utilize the WEPS model code as the basic
  framework and add WEPP model water balance,
  runoff, water erosion components.
• Advantages WEPS is already being implemented
  and tested by NRCS, WEPS databases and interface
  already available, WEPS has recently incorporated
  WEPP water balance/hydrology code.
• Disadvantages - WEPS code can only simulate
  single accounting region does not provide
  spatial representation currently in WEPP or in
  water erosion model in OMS, does not provide
  modules for OMS code repository, does not help
  with agency goals of more maintainable and
  reusable model components.

56
Path 2 - Progress to Date

• WEPP Water Balance incorporated into WEPS model
  code (2005-2006).
• WEPP kinematic wave computations for prediction
  of peak runoff rate, and also prediction of
  effective rainfall intensity and associated
  durations added to WEPS code (2006)
• WEPP hillslope (interrill/rill) erosion code
  added to modified WEPS model (2007), and is
  currently being tested and parameters linked to
  WEPS values where possible.

57
Path 2 Future Work

• Complete linkage of WEPP parameters to WEPS
  values.
• Where not possible or feasible to utilize
  existing WEPS information, need to add WEPP
  components to generate necessary information
  (e.g. water sediment particle size, water
  erodibility parameterization/updating)
• Test and verify combined WEPS/WEPP code against
  individual models.
• Modify WEPS interface/database to provide
  additional data necessary for WEPP.
• Possibly develop a new combined interface.

58
Summary

• Current development efforts are towards creation
  of a combined wind and water model based upon
  WEPP and WEPS science, for ultimate use in NRCS
  field offices.
• Two development paths for a common model are
  being pursued at present one building a modular
  combined model within OMS, the other utilizing
  WEPS as the basic framework.

59
Questions??