WEPS Testing Report

1
WEPS Testing Report
Mike Sporcic National Wind Erosion
Specialist National Technology Support Center
Central July 19, 2007
2
Background

• WEQ has been with us since early 1980s
• State east of the Mississippi River do not have
  any tool to estimate erosion
• We need a faster, robust, consistent and precise
  way to estimate erosion
• One that can tell us something about off site
  issues as well as onsite

3

• NRI data from WEQ Critical Period Method
• Notice Eastern US shows no erosion.
• Red gt8 t/a
• Tan 5-8 t/ac
• Yellow 3-5 t/ac
• Green 1-3 t/ac
• White 0-1 t/ac

4
Testing

• ARS-WERU at Manhattan, KS has worked 16 years on
  a new model.
• A final version (1.0) was given to NRCS for
  testing April 15, 2005.
• Half of the testing was done in 2005 the other
  half in 2007.
• There was almost a years time between early and
  late testing.
• Over 30 State level staff in 10 phases (sessions)
  in 22 CMZs have tested the model.

5
(No Transcript)
6
(No Transcript)
7
Testing Issue-Growth

• Probably the top issue is with the EPIC (growth
  sub-model inside WEPS)
• Some crops do not grow the first time they are
  run in the model
• An estimated 10 percent (my number) of the runs
  will require additional adjustments to the crop
  file to run

8
Testing Issue-Recommendation

• We have requested that Dr. Wagner develop a third
  way in the model to grow a proper yield and the
  residue needed.
• We have suggested that ARS use the accepted crop
  growth curves in RUSLE2 to set the total biomass
  the model should use.
• Insure WEPS will grow all crops by rebuilding or
  correcting current NRCS crop files.
• Long term, have fixed crop growth curves for
  crops that do not grow correctly in the model.

9
Testing Issue-Yield

• Yield is related to growth. The model can produce
  the correct yield about 30 of the time (my
  number)
• The remaining 70 (my number) of the time it
  requires Calibration
• 90 (my number) of the time this generates the
  correct yield and residue needed.
• 10 of the time it fails to estimate the erosion
  or crashes

10
Testing Issue-Recommendation

• Calibration has resolved most of the yield issue
• We think, if localize crop records are developed,
  the model will run most of he time at a given
  field office
• Training will be the key to this issue

11
Testing Issue-Interface of the Model

• ARS has spent a considerable amount of time to
  make the interface easy and fast
• Most testers like the current model very much
• There may be small terminology issue on the
  simulation area definition
• A calibration button on the main toolbar is
  needed.

12
Testing Issue-Recommendation

• NRCS considers the interface as finished
• We feel it is fine for now.
• We will need to train users on how to identify
  correctly the simulation versus the field area.

13
Testing Issue - Calibration

• Testing has shown that most runs need
  calibration. There are to many steps needed to
  make a calibration run.
• If the default yield in the crop record is
  changed in MCREW, we would like the model to run
  in the calibration mode without any other
  changes. As you come out of the editor and hit
  the run button on the interface, the model would
  run the in calibration mode.

14
Testing Issue-Recommendation

• Calibration should automatically run if MCREW
  yield value is edited.

15
Testing Issue - Reports

• The PDF output document currently produced in
  WEPS are great!
• Testers have requested an additional combine
  report where the Run Summary, Mgt output, and
  Crop Summary are combined into one report has
  been requested.

16
Testing Issue-Recommendation

• A big thanks for the work that was done on the
  reports
• ARS should finish the one additional report

17
Testing Issue Windgen Climate Data

• Climate stations are too far apart
• There can be as much as a 10X difference, one
  adjacent side to another.
• Lack of data available
• The west has micro topography that interfere with
  correct wind speeds
• North American Regional Reanalysis (NARR) has
  data that may give us additional stations
• Sample testing has shown that the NARR data needs
  additional work to be useful
• Costs to purchase needed data

18
Testing Issue-Recommendation

• NRCS should budget to purchase the needed data
  for the model
• Continue to work on ways to make the limited data
  work in the western states
• State Agronomists identify land areas to which
  current Windgen stations apply. This should be a
  State GIS map.
• Use North American Regional Reanalysis (NARR) or
  similar Windgen data to create additional (closer
  spacing) at least in the West Region.

19
Testing Issue Cligen Data

• Some Cligen data does not represent the cropland
  near it
• The FS database used is incomplete. There are
  more data stations with good data that is not in
  the model
• Model needs an editor that will allow good data
  stations to be added

20
Testing Issue-Recommendation

• NRCS will identify where current stations do not
  reflect the conditions of the cropland and where
  they think good data is available. We will tie
  the stations to areas on the maps to be used.
• Develop additional Cligen data using WEPP Cligen
  protocol. NRCS-NWCC, Greg Johnson, Mike Sporcic
  and Dave Lightle should develop or identify an
  editor to allow NRCS to add stations when good
  data is available

21
Testing Issue Importing Single Year Management
Files

• WEPS is capable of importing both multiyear and
  single year management files from RUSLE2.
• RUSLE and WEPS count cropping years differently
• Some adjustments on dates and operations are
  required

22
Testing Issue-Recommendation

• Short tem solution, WEPS SKEL file import
  redesign to create both single year mgt files for
  use as mono cultures or as components of crop
  rotations
• Redesign MCREW to read files the same as RUSLE

23
Testing Issue Reasonable Answers

• WEPS produces reasonable answers when compared to
  WEQ
• AZ and HI do not produce high enough erosion
  rates compared to what we expect
• Wind eroding land east of the Mississippi will
  have erosion estimates for the first time

24
Testing Issue-Recommendation

• As NRCS goes through implementation of Version
  1.0, assistance to the data management staff,
  wind erosion specialist, and state and field
  office users will be needed
• This will involve some model work, and mostly
  crop database correction.

25
Testing Issue Missing Soil Data

• This was an early NRCS issue. Many soil surveys
  were missing data when testing started.

26
Testing Issue-Recommendation

• NRCS has corrected this issue.
• ARS modified the code to keep the model from
  running with default data when NRCS soils data
  was missing.
• NRCS field staff must notify the State Soil
  Scientist if any map units show errors.

27
Testing Issue Long Run Times

• Long run time were an issue early in the testing
• ARS changed coding to include WEPP hydrology
• MCREW editing still a slight issue

28
Testing Issue-Recommendation

• Overall speed is greatly improved
• An evasive memory leak still could use some
  correcting
• Speed is generally acceptable now

29
Testing Issue Model Boots Slowly

• On my machine with a fresh computer reboot, it
  took 70 seconds to boot and 30 seconds to see the
  WEPS splash.
• Subsequent reboots of WEPS are much faster.

30
Testing Issue-Recommendation

• This issue is fine now. We may need to tell
  everyone to just wait a minute

31
Testing Issue Growing Two Crop

• Growing two crops at the same time is not
  possible for the model
• NRCS has systems that use a cover crop at the
  same time as the main crop is growing
• Crop databases do not have enough crop files to
  simulate the necessary rotations

32
Testing Issue-Recommendation

• NRCS will need to instruct staff to identify the
  missing crop situations so that the additional
  files can be build
• ARS and NRCS will need to build any new files to
  model the two crop system.

33
Testing Issue Organic Soils

• 5 states have wind erosion problems that can not
  be simulated currently
• The model needs ability to predict erosion on
  organic soils
• Samples have been sent to the ARS Lab in Manhattan

34
Testing Issue-Recommendation

• Continue to characterize the samples and develop
  a method to predict wind erosion on organic soil
• Let NRCS know if they can help ARS in anyway.

35
Testing Issue - Databases

• Some crop and operations are not in the model
• RUSLE2 management files have not been converted.
• Some files do not run correctly

36
Testing Issue-Recommendation

• NRCS convert the RUSLE management files
• ARS continue to assist NRCS diagnose crops files
  that do not run correct

37
Testing Issue Add SCI to Model

• In the ID testing session (Nov 07) the group
  requested the Soil Conditioning Index (SCI) and
  the Energy Calculator be added to the model
• This would save running RUSLE to get the SCI when
  wind erosion is the only type of erosion present
• ARS estimates it will take 25,000 dollars to
  write the code needed

38
Testing Issue-Recommendation

• NRCS secure the funding needed to add this
  function to the model
• ARS continue to plan to add this function to the
  model

39
Summary
40
Summary cont.
41
Summary cont.
42
Questions?