Modeling Soil Moisture using the Versatile Soil Moisture Budget and AquaCrop

Emmanuel RoTimi Ojo and Paul Bullock, University of Manitoba, Winnipeg, MB, Canada

Contact First Author: Emmanuel RoTimi Ojo; Timi.Ojo@umanitoba.ca

Abstract ID#: 34987

 

English Abstract:
There is an increasing focus on soil moisture determination using remote sensing. However, many of these remotely sensed retrievals are sensitive to soil moisture levels only in the top few centimeters of the soil. Soil moisture through the entire root zone is important for many agronomic and hydrological applications. The VSMB and AquaCrop are one-dimensional soil water balance models that simulate soil moisture at user-defined depths. Nine permanent soil moisture monitoring sites were established in Manitoba, Canada using Stevens Hydra probes installed at 5, 20, 50 and 100 cm to provide observed soil moisture data for the root zone across a range of soil textures and agricultural crops. A meteorological station that monitors rainfall, temperature, radiation, humidity and wind speed was also installed at each site. A preliminary analysis was conducted on the VSMB model using the 2013 growing season data for model calibration and the 2014 growing season data for model validation. The Normalized Root Mean Square Error (NRMSE) for total root zone soil moisture (0 – 130 cm) ranged from 2.6 to 11.0% in 2013 and 6.5 to 37.5% in 2014. It is expected that over time, the NRMSE for the VSMB output will decline as more data becomes available for model calibration across a wider range of soil moisture conditions. Soil moisture simulation using AquaCrop is being investigated. Improving the accuracy of soil moisture estimation from models is essential for interpolating root zone soil moisture across the landscape from soil profile information and point-based monitoring networks.