NRCS WEPS UPDATE REPORT

1
NRCS WEPS UPDATE REPORT

• October 23, 2008
• Mike Sporcic
• NRCS, National Wind Erosion Specialist
• Central National Technology Support Center
• Fort Worth, TX

2
Just a Reminder Why NRCS Needs a New Model

• The Wind Erosion Equation is very old technology.
  It started in the 80s.
• NRCS uses a very non-uniform application of the
  WEQ model.
• Some states use the paper version of the model
  (ID).
• Some states use only management period method. (a
  completely different rate of erosion compared to
  the annual method)

3
Why NRCS Needs a New Model

• Some state use only the annual method.
• Some states use both depending on management,
  climate, and the soils.
• Many use the spreadsheet method, my sheet or one
  of there own.
• Some states have modified their I factor values.
• Without regard to the age of the model and
  technical issues, we have no consistency in
  current our predictions!

4
WEQ does not predict erosion well in the Light
Yellow shaded area, Clt10, and the annual method
is used.
5
Wet areas and Organic Soil

• Wet areas shown on the last slide have Climate
  Factors lt10.
• Rainfall areas gt25 inches/yr do not predict
  correctly.
• Organic soils occur in this light yellow area and
  are not modeled well.

6
Clay Issue

• Why did clay soils erode at a greater rate than
  medium textured soils like silt?
• Simple answer is that one of the equations in the
  model was incorrect.
• Dr. Hagen and Dr. Wagner fixed the equation on
  August 8.
• One example we had in TX on a Mercedes Clay soil
  went from 18 t/ac loss to 3 t/ac after the fix.

7
Clay Issue

• With the change, Clay soils eroded equal to or
  more than medium textured soils.
• This solved the issue on the heavy soils.

8
Rock Issue

• As I ran a group of soils picked at random over
  the US I noted large soil loss differences in
  similar soil textures.
• As I checked further I found the low run values
  were soils that had rock in them, from the SSURGO
  data files we use.
• The Mercedes Clay had 12 rock in the database
  and the model used rock to reduce erodiblity on
  the surface.
• This caused the soil loss to go from 18 t/ac to 8
  t/ac (pre-clay adjustment).

9
Rock Issue

• The TX Field Staff told us that there are no rock
  on the surface of the this soil.
• Our national soils staff indicated the rock data
  was just an estimate and could be almost
  anything.
• We asked Larry Wagner if it would be possible to
  have the rock on the surface be a direct entry.
• Larry has agreed to give us a direct entry on the
  main interface to a estimate the surface rock
  cover as we do for RUSLE2.
• This issue will be resolved when the programming
  is complete.

10
Irrigation Issue

• Field Staff out of Lubbock asked why the erosion
  rates seem to go up if irrigation water is
  applied to the field?
• Answer - when water is applied to the surface of
  the soil it lowers surface roughness, both random
  and ridge roughness. It also segregates the
  surface soil and exposes saltation and roller
  soil particles.

11
Irrigation Issue

• The problem was that when the supplemental water
  was added the yields in the run were not adjusted
  to irrigated levels.
• Example running continuous Cotton at Midland TX
  we find 10.5 t/ac soil loss at a calibrated
  dryland yield of 378 lbs lint/ac.
• If we add water and hold the yield the same the
  loss goes up to 11.2 t/ac loss.
• If we add water and calibrate the yield to 1500
  lbs lint/ac we have 3.2 t/ac soil loss.
• This issue is resolved if we adjust yield and
  calibrate.

12
Climate Issue

• For some time now we have been seeking ways to
  minimize the large differences in soil loss
  output when we cross a Cligen or Windgen station
  boundary.

13
Need for consistent station selection

• NRCS needs to have climate data that doe not
  cause large changes on soil loss when crossing a
  boundary from one station to another.
• As we have tested and trained our state staff we
  have noted large changes in soil loss from
  station to adjacent station. This true for
  Cligen and Windgen station data.

14
Lubbock, TX Example of Climate Stations (using 15
year runs) Continuous Cotton
These examples have the new 26,000 plant/ac
cotton crop record used.
15
Greg removed Plainview/Hale, Big Spring,
Childress, Cannon AFB, and Reese will be changed
to Lubbock Intl AP
These climate stations are all within the Lubbock
Windgen station area
16
TX Example of Climate Stations without
Calibration
Cotton, skip row, dryland Climate stations were
the only change. Amarillo FSL was the
soil. Stations are 30-35 mi apart. 26,000
plant/ac
4.5 t/ac 501 lb/ac
8.7 t/ac 333 lbs/ac
7.5 t/ac 390 lbs/ac
5.4 t/ac 329 lbs/ac
5.3 t/ac 418 lbs/ac
9.2 t/ac 392 lbs/ac
17
Planning interpretation without recalibration.

• The system applied in Hale, Lubbock, and Crosby
  Co. meets our soil resource quality criteria.
• However applied in remaining counties it does not
  meet the T of the soil and would require
  additional treatment.

18
TX Example of Climate Stations with Calibration
Cotton, skip row, dryland Climate stations were
the only change. Amarillo FSL was the
soil. Stations are 30-35 mi apart. 26000
plants/ac
5.4 t/ac 329 lb/ac
8.6 t/ac 333 lbs/ac
8.7 t/ac 333 lbs/ac
5.4 t/ac 329 lbs/ac
6.3 t/ac 334 lbs/ac
15.5 t/ac 309 lbs/ac
There is a 10.1 ton/ac difference going from
Lubbock to Lynn Co.
19
Planning interpretation with re-calibration.

• The system applied in Hale and Lubbock, meets
  our soil resource quality criteria.
• However applied in remaining counties it does not
  meet the T of the soil and would require
  additional treatment.
• Lynn Co. (Tahoka station) has 3 times more
  erosion than Lubbock Co.

20
Reasons for the Difference

• We discovered the reason for the difference was
  not the change in the data. We found out that 15
  years was not long enough to run the model.
• If I run the model for 30 years most of the
  differences disappear and the soil loss is about
  8 ton/ac and requires some treatment in all of
  the counties.
• Larry is testing how climate is affecting the
  model now.

21
WI Cligen and Windgen Example
22
Mossinee/Central WI Windgen
5.1 t/ac
1.6 t/ac
Corn, grain-Peas, green-Bean, green
rotation. Adam Co run on Volk/Camp Doulas
Windgen Station. Change in Adams Co is Cligen
Stations All run were Calibrated.
1.7 t/ac
1.4 t/ac
Volk/Camp Doulas Windgen
WI Example The sharp change here is a Windgen
Station change
1.5 t/ac
1.4 t/ac
4.7 t/ac
3.1 t/ac
MadsonDane Regional Windgen
23
Draft Windgen Map-MN
24
MN Cligen and Windgen Example
Suggest moving the Windgen boundary to not have a
split in the county.
Marshall Co.
Roseau Muni Windgen
Grand Forks Intl Windgen
34.3 tons/ac SBar-Sunf
9.7 tons/ac SBar-Sunf
2.9 tons/ac Beans-SW
6.7 tons/ac Beans-SW
Thief River Windgen
25
Ideas so far (Cligen)

• The West will use the Cligen station map built
  earlier this year. The model will pick the
  correct station if you click the spot in the
  county of interest.
• Central has maybe two options
• 1. Use the equal distance polygon map I have
  shown and select the station that the field falls
  in. This can be from a map or GIS in the model.
  This would have some counties with up to 5
  polygons in county.

26
Ideas so far (Cligen)

• 2. Assign a station to each county using a map
  or GIS. At lease 1000 counties nation wide do
  not have a station inside the county. Dave and I
  could work to use RockClime to create a station
  in the centroid of the missing counties. This
  would take a lot of time.

27
Windgen Stations

• Greg Johnson has developed and we have digitized
  a Windgen map for the Central region.
• Greg has determined that we do not have enough
  station data to make a complete map for the
  Western region.

28
Windgen for the West
Greg made an attempt to draw the map, the black
lines. We need a map that covers the red-line
area.
29
Climate, General

• NRCS can not put the model in the field until we
  have a way to address the boundary issues with
  the current data stations.
• Windgen and Cligen stations are too far apart to
  have a smooth soil loss transition.
• Greg is testing the use of the North American
  Regional Reanalysis data from NOAA. They have 20
  mile grid data, and would give us a smoother
  transition.

30
Climate, General

• I Think Greg can fill us in on some testing he
  and Susan ONeill are doing.

31
New Summary Report
The new report has a very nice printout.
32
Soil Loss Runs for ID

• We still need to find a way to predict spring
  time erosion in the Potato-WW runs from ID. John
  Tatarko has some samples to yet report on and I
  have created a potato planter operation that
  gives us a residue and roughness needed.
• Some work has been done, but more is needed
  because we do not show erosion in the spring when
  planting potatoes.
• Presently we show about 20 t/ac on the Wynne set
  of runs and about 5 t/ac on the Fort Hall set of
  runs.

33
More on the Residue Issue

• Dave Lightle and I asked the question How does
  WEPS treat buried roots. Answer was it used the
  surfaced roots as residue.
• RUSLE 2 does not surface roots.
• Tom Golhke noticed that if you burn of the
  surface residue of a barley crop down to 8
  surface cover, you can have 40 plus residue at
  seeding with a chisel, cultivation and a packing.
  This is to high!

34
Root Surfacing

• Larry has given me a version of the model to test
  where we do not surface roots as residue.
• Toms run had 17 surface cover and 331 lbs/ac on
  the Spring Barley burn run at seeding time.
  After the Apr 25 burn there was 145 lbs/ac
  surface residue. Soil loss was 5 t/ac
• With the root surfacing turned off, the after
  burn residue was 228 lbs/ac and at seeding the it
  was 4 and 71 lbs/ac. Soil loss was 11.6 t/ac.
  This is much better.

35
Root Surfacing, cont

• Leigh Cranmers skip row cotton at Lubbock with
  roots has a soil loss of 7.5 t/ac at a 30 yr run
  and 4 cover, 83 lbs/ac residue at planting time
  with root surfacing.
• Without the roots, the planting time residue is
  1 at 37 lbs/ac and the soil loss is 8.2 t/ac.
• Several other runs show similar changes.
• This change seems to have address Toms request
  from the Western runs that he made.

36
A Last Comment on Climate

• Currently we found out that the differences in
  the TX Cligen example were due to not running the
  model long enough.
• With four different seed numbers the range of
  soil loss was 5 t/ac to 9.5 in Lubbock. At
  Tahoka the rang was 6 t/ac to 7.5 t/ac
• With the model run for 100 years, the average for
  Lubbock is 8.4 t/yr and 7.9 t/ac for Tahoka.

37
A Last Comment on Climate

• The last major item to deal with is the number of
  years needed to run the model.
• In the case of the TX climate we need more years
  to reach a good estimate.
• I have retested the troublesome runs in TX and
  found that 30 years seems to be fine.
• Larry is working on some testing to find the
  correct length of years needed to run the model.
  He will comment on the work underway.

38
Added 15 years, 4.5 t/ac jump down.
Added 15 year, 2.4 t/ac jump up.
At 15 years the diff. is 10.8. At 30 years the
diff is 2.3. 30 yrs may be long enough to run at
this site.
39
Just Some Muck Pictures from MI
40
Just Some Muck Pictures from MI
41
Just Some Muck Pictures from MI
42
Just Some Muck Pictures from MI
43
Just Some Muck Pictures from MI
Sand is very evident in many of the organic soils
44
Just Some Muck Pictures from MI
This is not the moon! It is a fresh disking of
dry muck. We sank in 4 inches or so. This soil
had very little sand in the surface and was
starting to move with less than 6 mph of wind.
45
Short Term Muck Fix

• We plan to make one or more Organic Soil files to
  use in WEPS to represent all organic soils. This
  will cover Sapric, Histic, Folist, Fibric, and
  Histic Intergrades.
• The surface 15 inches will be a Loamy Fine Sand
  or (from a field sample FL), and the remaining 45
  inches will represent an organic soil that we
  have lab data for. This will help the hydraulic
  model run like a muck.
• Soil data from FL was sandy enough to use without
  adding the LFS to the top 15 inches. It is
  actually a LS.

46
Soil Data for Short Term fix

• Larry West and Cathy Seybold sent some (averaged)
  muck soil data for Michigan and Florida.
• An .IFC record was constructed for both the sets
  of data.
• Some soil parameters in WEPS have range limits.
  If the data was outside the limit, a number
  closes to the limit was used.
• A third soil was evaluated from D.L. Mokmas
  unpublished muck data. It was very similar to
  the MI .IFC soil.

47
(No Transcript)
48
Findings of Organic Soils Runs

• The MI soil shows lower soil loss. 0.1 t/ac in
  Lapeer Co., MI to 19.0 t/ac in Perquimans Co. NC.
• The MI soil may be to low for runs in MI. I need
  to make more runs with vegetables to be sure.
• The FL soil shows higher soil loss. 0.7 t/ac in
  Lapeer Co. MI to 68 t/ac in Perquimans Co. NC.
• The FL soil may be to high in some cases.

49
Soil Loss by Depth Reflecting Bulk Density
Changes.
1 t/ac book muck 5 t/ac CL
50
Short Term Decision

• We will be visiting the FL area near Palm Beach,
  FL to see how MI compares with FL.
• We will make a decision on the short term at the
  end of that field review.
• This idea appears to be a good short term
  solution that follows our current National
  Agronomy Manual.

51
Long Term Muck Fix

• We plan to support three research efforts.
• Support Teb Zobecks effort to bring a portable
  wind tunnel to MI. Need dollars.
• Support John Tatarkos effort to study the
  surface conditions through a tillage season. He
  will need capable staff time and dollars.
• Jerry Grigars will bring some large samples of
  common MI mucks to the WERU lab for wind tunnel
  studies. We need some staff time and dollars.
• Findings will be added to Version 2 of the model.

52
WEPS Summary

• We need to solve the Climate and Wind Data
  issues.
• Need to address the run length in years and the
  Run Time concerns.
• We need to ensure the model is accurate and gives
  consistent answers.
• There additional items that have been requested
  that relate to ease of use that need to be done.
• We need resolutions within months.

53
Questions
Sometimes you just feel like youre stuck!