Calibration of CroPMan with Texas High Plains Limited Irrigation CottonSorghum Cropping System Data

1
Calibration of CroPMan with Texas High Plains
Limited Irrigation Cotton-Sorghum Cropping System
Data

• A.M. Cranmer1, J.P. Bordovsky1
• W.L. Harman2, E.M. Steglich2, and J.R. Williams2
• 1Texas Agricultural Experiment Station Halfway,
  Texas
• 2Blackland Research and Extension Center
• Texas Agricultural Experiment Station Temple, TX
• Texas Agricultural Experiment Station, Halfway,
  Texas

2
Introduction

• The Southern High Plains is in a transition from
  irrigated acres to dryland (rain-fed) production.
  
• Seasonal rainfall averages between 12 to 14
  inches per year however, this is composed of
  extremes making dryland production unstable.
• Optimum economic allocation of limited water
  resources depends on the crop(s) being produced,
  the cropping system, irrigation capacity, pumping
  cost and commodity price.

3
Objectives

• Evaluate a management tool that is currently
  being used to assist producers so they may
  determine and optimize their irrigation water
  value.
• CroPMan was calibrated in an attempt to provide
  an irrigation management tool for the Texas High
  Plains, particularly in areas of low irrigation
  capacity.

4
CroPMan

• CroPMan is a production-risk management model
  developed by scientists at the Blackland Research
  and Extension Center in Temple, TX and is the
  windows based application of the USDA-ARSs
  Environmental/Policy Integrated Climate (EPIC)
  model.
• CroPMan is used to assess real-time situations
  in the field, estimating crop and
  soil-water-fertility status, and to project
  implications of additional irrigation and
  fertilization (i.e., timing and amount) on crop
  productivity (Gerik, 2005).

5
Methodology
Texas High Plains

• Study site Halfway, Texas
• Growing Season 2001, 2002, 2005
• Treatments 4 Crop Sequences x 3 Irrigation
  Capacities
• Design Randomized Complete Block
• Number of replicates 4
• Total experimental units 48 Plots

6
Weather Problems Halfway, Texas, 2003 and 2004

• 2003 Hail and Wind Storms Compromised Field
  Experiments
• 2004 Record Rainfall Amounts Impacted Field
  Experiments

7
Treatment Layout Halfway, Texas, 2001, 2002, 2005
Crop Sequences CCC cotton cotton
cotton CCS cotton cotton sorghum CSC
cotton sorghum cotton SCC sorghum cotton
cotton Irrigation Capacities 2.5 Gallons
per Minute 1.25 Gallons per Minute 0.0 Gallons
per Minute (Dryland)
8
Treatment Layout Halfway, Texas, 2001, 2002, 2005
9
CroPMan Inputs

• Tillage
• Cultivation, Planting, Harvest
• Irrigation
• Application, Timing, Amounts
• Weather
• Temps, Radiation, Wind, Humidity, Precipitation
• Cropping System
• Crop, Rotation
• Fertility
• Type, Amount

10
CroPMan Views
Weather Analyzer
11
CroPMan Views
Options Menu
12
CroPMan Views
Cropping Systems Menu
13
CroPMan Views
Management File
14
CroPMan Views
Management File
15
CroPMan Parameters

• Biomass to Energy Ratio (BER)
• Affects high yields
• Harvest Index (HI)
• Affects high yields
• Lower Limit of the Harvest Index (WYSF)
• Affects low yields

16
CroPMan Views
Biomass to Energy Ratio
17
CroPMan Views
Harvest Index
18
CroPMan Views
Lower Limit of Harvest Index
19
CroPMan Views
Create Run Menu
20
CroPMan Views
Output Screen
21
CroPMan Simulations
22
CroPMan Simulations
23
CroPMan Simulations
24
Comparison of simulated to actual grain sorghum
yield for cropping years 2001, 2002 and 2005.
25
Comparison of simulated to actual lint yield for
cropping years 2001, 2002 and 2005 before parm
adjustment.
26
Comparison of simulated to actual lint yield for
cropping years 2001, 2002 and 2005 after changes
made to the parm controlling the volatilization
fraction.
27
Ratio of best fit simulated to actual lint yield
by year. Ratios in 2003 and 2004 are well above
one, indicating modeled results were well above
actual yields.
28
Simulated vs. estimated actual lint yields using
best fit parameters for the 2006 crop year.
29
Summary

• The simulation before parm adjustment resulted in
  a regression beta coefficient near 1.0 and a
  coefficient of determination of 0.77.
• Simulated cotton yields following sorghum were
  often much lower than simulated or actual
  continuous cotton yields.
• The simulation after parm adjustment improved
  validation and resulted in a regression beta
  coefficient closer to 1.0 and a coefficient of
  determination of 0.86.

30
Summary cont.

• Simulated cotton yields following sorghum were
  improved for a much better fit after the parm
  adjustment.
• Data scatter from the ratios of actual to
  simulated yield by year were within 15 in 2001,
  but increased up to 40 in 2002 and 2005.
• At this level of accuracy CroPMan may be suitable
  for regional applications but not for field
  specific irrigation decisions.

31
Summary cont.

• However with the results of the recent parm
  adjustment and increase in model accuracy after
  the adjustment, additional changes may lead to a
  validated model that could be very useful as a
  decision making tool for the Texas High Plains.

32
Acknowledgements
Jim Bordovsky Dr. Wyatt Harman Evelyn
Steglich Dr. Jimmy Williams Joe Mustian Doug
Nesmith Larry Francis Omar Amawi
TAES Halfway, TX TAES Temple, TX TAES
Temple, TX TAES Temple, TX TAES Halfway,
TX TAES Halfway, TX TAES Temple, TX TAMU
College Station
33
Questions ?