bcc29eb3-4607-11de-9a91-00219b106224

eng

Mark Gervais

Department of Soil Science, University of Manitoba

(204) 474-8666

Room 362 Ellis Bldg

Winnipeg

MB

R3T 2N2

Canada

Paul_Bullock@umanitoba.ca

author

2009-05-21

ISO 19115

ISO19115:2003/Cor 1 2006

PAMII Model Wheat Study Output

2008-01-01

Publication

Input data for PAMII include the following: 1. daily maximum and minimum air temperature 2. daily precipitation 3. seeding date 4. twice daily upper air temperature, wind speed, dew point depression and geopotential height for the standard levels of 1000 mb, 850 mb, 700 mb and 500 mb 5. roughness length 6. terrain drag coefficient 7. soil moisture contents at field capacity, permanent wilting point and saturation 8. soil moisture content for the top zone and total soil profile on the seeding date 9. soil saturation matric potential 10. the slope of the soil water retention curve (semi-log). PAMII simulates soil moisture content in three soil layers: the top zone (upper 10 cm), the root zone (starts at 5 cm and grows to maximum of 120 cm) and the sub zone (soil below the bottom of the root zone and 120 cm depth). The model simulates evaporation and transpiration separately. The evaporation term is based solely on the top-zones water content. The evaporation flux is restricted by soil resistance term which restricts the movement of water from the top-zone to the soils surface. The transpiration flux is restricted by a bulk canopy resistance term which reflects the physiological resistance of the entire crop. Canopy resistance is calculated as a function of fractional leaf area and the fraction of soil water in the root zone that is plant available. A stability-adjusted aerodynamic resistance term which reflects atmospheric ability to transport water vapor is used to restrict both the evaporation and transpiration flux. PAMII calculates atmospheric resistance by simulating the vertical profile of the planetary boundary layer. Model output includes the following: 1. Root Zone water content (both as mm and % of available water) 2. Top Zone water content (both as mm and % of available water) 3. Actual evapotranspiration (mm) 4. Relative Humidity (%) 5. Biometeorological Time 6. Leaf Area 7. Potential evapotranspiration (mm) 8. Total soil moisture content

To assess the accuracy of the second-generation Prairie Agrometeorological Model (PAMII) as an agrometeorological model for spring wheat on the Canadian Prairies, a study was conducted to validate the model using field measurements gathered from a previous experiment. Two locations in Manitoba and three locations in Saskatchewan had been used to encompass various soil and climatic conditions throughout the Prairies (Winnipeg and Carman, Manitoba and Melfort, Regina and Swift Current, Saskatchewan). Monitoring occurred within the growing season from 2003 through 2006. Each study site consisted of a weather station adjacent to a set of spring wheat research plots. Soil water content was measured on a weekly basis in Manitoba and a biweekly basis in Saskatchewan using a neutron probe to a depth of 120 cm below the surface. To account for spatial variation in soil water content a network of access tubes were installed at each site within several different plots of wheat. Soil characteristics such as texture and water holding capacity were determined at several intervals of the 120 cm profile. Two modifications were introduced into the original version of PAMII to improve modeled soil moisture conditions late in the growing season (when the observed soil moisture conditions were drier than the modeled values) and after heavy rainfall events (when the observed soil moisture conditions were higher than the modeled values). The canopy resistance function was modified so canopy resistance would not start to increase until the soil water content was below 50% of plant available water. In addition, the model was modified to allow infiltration of water to continue independent of the top-zones water content. Overall both modifications reduced the RMSE of modeled soil water by 9 mm or 2% of field capacity over all site-years. The output in this dataset is from the modified version of the model.

completed

notPlanned

Land Surface > Soils > Soil Moisture/Water Content

Land Surface > Soils > Soil Water Holding Capacity

Atmosphere > Atmospheric Water Vapor > Evapotranspiration

GCMD

2007-04-01

Publication

intellectualPropertyRights

vector

eng

utf8

farming

-107.72

-97.17

49.5

52.82

The original study focused on growing season weather effects on spring wheat. Therefore, the dataset spans the growing season at each site-year, daily observations from planting to a date at crop maturity or later as per below. Melfort: 2003 (05-May to 16-Sep), 2004 (07-May to 30-Sep), 2005 (06-May to 05-Sep), 2006 (09-May to 30-Aug) Regina: 2003 (22-May to 14-Sep), 2004 (19-May to 30-Sep), 2005 (02-May to 28-Aug), 2006 (05-May to 23-Aug) Swift Current: 2003 (17-May to 15-Sep), 2004 (29-Apr to 06-Sep), 2005 (26-Apr to 15-Aug), 2006 (27-Apr to 23-Aug) Winnipeg: 2003 (16-May to 17-Sep), 2004 (29-May to 06-Sep), 2005 (05-May to 17-Aug), 2006 (28-Apr to 16-Aug) Carman: 2004 (18-May to 15-Sep), 2005 (04-May to 23-Aug), 2006 (12-May to 12-Sep)

The accuracy of actual evapotranspiration (ETa) by the water balance method was assessed as the total measurement error of initial and final soil moisture content determination plus error in precipitation measurement during each measurement interval. The accuracy of each water balance ETa measurement (error bars showing total measurement error) and the corresponding ETa value from the model is available as a graph through the data distributor

XLS

2003

DRI Data Legacy

http://www.drinetwork.ca/

Distributor

0.6

http://odin.meteo.mcgill.ca/dri_download/pam2_wheat_output.zip

attribute

Soil Moisture and Evapotranspiration data accuracy

Root mean square error of soil moisture

53mm -(10% of field capacity)

Mean bias error of soil moisture

0 mm

R square of soil moisture

0.84

Accuracy of Actual Evapotranspiration

The accuracy of actual evapotranspiration (ETa) by the water balance method was assessed as the total measurement error of initial and final soil moisture content determination plus error in precipitation measurement during each measurement interval. The accuracy of each water balance ETa measurement (error bars showing total measurement error) and the corresponding ETa value from the model is available as a graph through the data distributor

Not applicable

Automated weather stations (Campbell Scientific, Logan, UT; Spectrum, Plainfield, IL) were installed at each location at seeding to collect hourly and daily air temperature, rainfall, wind speed, relative humidity, solar radiation, soil temperature and soil moisture until harvest. Air temperature and relative humidity were measured at 1.8 m height with a radiation shielded probe (CS 500, Campbell Scientific., Logan, Utah). Incoming solar radiation was measured at 2.0 m with a silicon pyranometer (Model SP-LITE, Kipp & Zonen, Netherlands). Other measured parameters were wind speed at 2.5 m (Cup Anemometer, Model 3102, RM-Young Co. Traverse City, MI) and rainfall (Tipping Bucket Rain Gauge, Model TE-525mm, Texas Electronics, Houston, TX). The data loggers were programmed to log each sensor every 10 s and to output both hourly and daily averages, sums, and maximum and minimum values. Prior to seeding, six random soil samples were collected at depths of 0-15, 15-30, 30-45, 45-60, 60-90, and 90-120 cm to determine initial soil moisture. Immediately following harvest each plot was sampled at depths of 0-15, 15-30, 30-45, 45-60, 60-90, and 90-120 cm to determine final soil moisture content. Soil moisture was monitored every 10 to 15 days at each growing location using a neutron probe (Troxler Laboratories, Triangle Park, NC). Neutron access tubes were installed to a depth of 150 cm into each plot of AC Barrie and Superb between the 3rd and 4th rows and 1 m meter inside the plot perimeter. Neutron readings were taken at 12.5, 22.5, 37.5, 52.5, 75, and 105 cm to provide a neutron count for the corresponding horizons of 0-15, 15-30, 30-45, 45-60, 60-90, and 90-120 cm. A calibration curve of moisture content was developed for each site using gravimetric soil samples collected near the access tubes. The calibration equation was used to produce volumetric moisture content data from the neutron counts from the six access tubes. Soil water content at the 0-15 cm depth was determined gravimetrically. Soil moisture data were averaged from the six measurements for each date at each location. Upper atmospheric elements were interpolated from the regional Global Environmental Multiscale (GEM) model gridded output. Documentation for the high-resolution CMC grib data is available online (http://weatheroffice.ec.gc.ca/grib/index_e.html). Upper atmospheric data included air temperature, wind speed, dew point depression and geopotential height for the standard levels of 1000 mb, 850 mb, 700 mb and 500 mb. Site specific upper atmospheric data was obtained for Melfort, Swift Current and Carman. The Regina airport was used for the Regina site and Glenlea was used for the Winnipeg, University of Manitoba research site. Roughness length for the wheat plots was approximated as 10% of crop height (Oke, 1987, Boundary layer meteorology, 2nd ed, Methuen & Co., London). Since crop height was not monitored, a maximum height of 1.2 m was assumed. Roughness length was constant for the entire growing season since PAM2nd does not account for changes in roughness length as the crop grows. Terrain drag coefficients (drag due to terrain relief) were obtained at each station by interpolating values from a 381 km grid (Cressman, 1960, Mon. Weather Rev. 88: 327-342.) down to a 47.6 km grid (Raddatz and Khandekar, 1977, Boundary Layer Meteorol. 11:307-327).