Use of a Computer Model to Simulate Soil Moisture Content in Irrigated Fields

Transcription

1 Use of a Computer Model to Simulate Soil Moisture Content in Irrigated Fields Mukesh Mehata 1, Saleh Taghvaeian 1, Sumon Datta 1, and Daniel Moriasi 1 Dept. of Biosystems and Agricultural Engineering, OSU, Stillwater, OK USDA-ARS Grazinglands Research Lab, El Reno, OK Annual Oklahoma Governor s Water Conference and Research Symposium Dec 5, 18

2 Outline Introduction Objectives Materials and Methods Results Conclusions Acknowledgement References

3 Introduction Why use model? To implement an efficient water management strategy in irrigated fields, soil moisture status must be carefully monitored throughout the growing season. However, obtaining information about soil moisture can be expensive, timeconsuming and labor-intensive Models, if validated, can be used to simulate the fluctuations in soil moisture under variable field conditions Reduce environmental contamination Improve profitability Effective irrigation management strategies Saving limited freshwater resources

4 Objectives To measure soil moisture dynamics at four soil depths in irrigated fields located in central Oklahoma, using a previously tested soil moisture sensor. To evaluate the performance of HYDRUS model by comparing the simulated soil water content with sensor readings in the same fields and depths

5 Materials and Methods Study Area Fort Cobb Reservoir Experimental Watershed, Caddo county, Central Oklahoma 5 irrigated fields 3 crops: peanut (PT), Soybean (SB), and Cotton (CN) Irrigation system: Center Pivot Climate: Semi-arid with annual rainfall of 75 mm (1997-1). June to October 18 (Precipitation = 5 mm)

6 Materials and Methods Step 1. Field Measurements Hourly based soil moisture data were collected at 1, 3, 51, and 71 cm depths using Acclima TDR 315 sensors (tested previously). Soil samples for textural analysis: -1, 1-, -3, 3-1, 1-51, and 51-71cm Rainfall and irrigation amounts were measured by rain gauges. Step. Reference Evapotranspiration (ET o ) Calculation FAO-5 Penman-Monteith (PM) ET o equation (R. G. Allen et al., 1998; Suleiman & Hoogenboom, 9; Walter et al., ) Hourly ET o was calculated by REF-ET software (R. Allen, )

7 Materials and Methods Step 3. Estimating crop evapotranspiration (ET c ) FAO-5 single crop coefficient (Kc) approach was used to estimate ET c based on ET o by applying crop-specific Kc values adjusted based on local climatic condition. ET = K ET c Step. Partitioning ET c to Evaporation (E p ) and Transpiration (T p ) E = ET e p T = ET E LAI = leaf area index; K = radiation extinction coefficient Values of these variables were obtained for peanut from Singh (), for soybean from Setiyono et al. (8), and for cotton from Aggarwal et al. (17). c c k. LAI p c p

8 Materials and Methods Step 5. Hydrus-1D Simulation for Root Water Uptake (RWU) Hydrus-1D v. was developed by Šimůnek et al. (8). Numerically solves Richards equation Assign number of materials & depth of the soil profile Input number of timevariable boundary records

9 Materials and Methods Other inputs to HYDRUS model Single porosity models: Van Genuchten (198) s closed form model Water flow boundary condition: Atmospheric BC with surface runoff for upper BC and free drainage for lower BC Root water uptake model : Used Feddes (198) parameters Root distribution from Fan, McConkey, Wang, and Janzen (1)

10 Materials and Methods Statistical Indicators for Model Performance Root mean square error (RMSE) Statistical Indicators Mean bias error (MBE) Nash Sutcliffe model efficiency (NSE)

11 Results: PT. M1-cm S1-cm I/P(cm). M3-cm S3-cm I/P(cm) M51-cm S51-cm I/P(cm). M71-cm S71-cm I/P(cm) Depth (cm) RMSE MBE R r NSE

12 Results: SB M1-cm S1-cm I/P(cm) M3-cm S3-cm I/P(cm) M51-cm S51-cm I/P(cm) M71-cm S71-cm I/P(cm).. Depth (cm) RMSE MBE R r NSE

13 Results: SB M1-cm S1-cm I/P(cm) M3-cm S3-cm I/P(cm)...5 M51-cm S51-cm I/P(cm).5 M71-cm S71-cm I/P(cm) Depth (cm) RMSE MBE R r NSE

14 Results: CN M1-cm S1-cm I/P(cm) M3-cm S3-cm I/P(cm)... M51-cm S51-cm I/P(cm). M71-cm S71-cm I/P(cm) Depth (cm) RMSE MBE R r NSE

15 Results: CN M1-cm S1-cm I/P(cm) M3-cm S3-cm I/P(cm) M51-cm S51-cm I/P(cm) M71-cm S71-cm I/P(cm).. Depth (cm) RMSE MBE R r NSE (-.1) (-.1) -1.

16 Conclusions The study compared simulated water content (ϴ v ) with sensor readings at high temporal resolution. The statistical indicators showed the performance of the model at top soil layer was reasonably good. The simulations from deeper soil layers did not produce satisfactory results except one field. Some differences between simulated and measured may be due to high level of heterogeneity in agricultural fields that can not be captured in input data, and thus cannot be blames on poor model performance HYDRUS can be considered as a useful tool for estimating top soil layer ϴ v.

17 Acknowledgement Natural Resources Conservation Service, U.S. Department of Agriculture, under number 9-3A Agricultural Research Service, U.S. Department of Agriculture, under Agreement No

18 References Setiyono, T., Weiss, A., Specht, J. E., Cassman, K. G., & Dobermann, A. (8). Leaf area index simulation in soybean grown under near-optimal conditions. Field Crops Research, 18(1), 8-9. Singh, A. (). Growth and physiology of groundnut. Groundnut research in India, Aggarwal, P., Bhattacharyya, R., Mishra, A. K., Das, T., Šimůnek, J., Pramanik, P., Sudhishri, S., Vashisth, A., Krishnan, P., & Chakraborty, D. (17). Modelling soil water balance and root water uptake in cotton grown under different soil conservation practices in the Indo-Gangetic Plain. Agriculture, ecosystems & environment,, Simunek, J., Van Genuchten, M. T., & Sejna, M. (5). The HYDRUS-1D software package for simulating the movement of water, heat, and multiple solutes in variably-saturated media. Version.. HYDRUS Software Series 3. Riverside, Cal.: University of California-Riverside, Department of Environmental Sciences. Fan, J., McConkey, B., Wang, H., & Janzen, H. (1). Root distribution by depth for temperate agricultural crops. Field Crops Research, 189, 8-7.

19 Thank You