Research Proposal
1. Project No.
2. Title: Evaluation of Ground Penetrating Radar for Measuring Soil Water Content Distribution
3. Focus Categories: AG, CP, COV,GW, HYDROL, IG, MET, MOD, WQN, WU,
4. Keywords: Soil moisture, unsaturated zone, soil water relations, remote sensing, ground penetrating radar (GPR).
5. Duration Period: March 1, 2002 – February 28, 2003
6. FY 2002 Federal Funds:$19,998
7. FY 2002 Non-Federal Funds:$55,057
8. Principal Investigators:
Eric Harmsen, Department of Agricultural and Biosystems Engineering, University of Puerto Rico at Mayaguez.
Hamed Parsiani, Department of Electrical and Computer Engineering, University of Puerto Rico at Mayaguez.
9. Congressional District:N/A
10. Statement of the Critical Problem.
Soil moisture content is needed when performing hydrologic analyzes, developing environmental models, and for properly scheduling irrigation. Currently, the moisture content is obtained by intrusive methods, which are limited to point measurements. The groundwater penetrating radar method holds promise because it is non-intrusive and provides a spatially continuous two or three-dimensional estimate of the soil moisture content.
11. Statement of Results or Benefits
This project will provide information needed for applying the GPR method for estimation soil moisture in Puerto Rico. The spatially continuous form of the GPR output would be useful for developing subsurface flow models, and for model calibration and validation. The Project will provide training to M.S. graduate student and an undergraduate student on the Mayagüez Campus.
Nature, Scope and objectives of the research (1 page)
Soil moisture is an important hydrologic variable needed to assess a variety of processes related to water resource management. Knowledge of soil moisture variation with time is necessary for conducting soil water balances and for parameterization of numerical models. These tools are used to estimate evapotranspiration from land cover, deep percolation for groundwater impact studies, and for quantifying surface runoff and sedimentation. Numerical models for predicting soil water movement are also important in solute transport studies.
The goal of this study is develop a model for accurate converting GPR images into moisture content distribution. The to develop a methodology for deriving soil water related parameters and soil water content for use with traditional hydrologic analysis tools (e.g., water balance and numerical models). Specific objectives are:
- Conduct field study to obtain shallow (1 meter) GPR measurements at various locations where the vertical moisture content distribution is know with high accuracy. The vertical moisture content will be determined independently of the GPR using a time domain reflectometry (TDR) soil moisture measurement device.
- Develop and apply improved image resolution and noise reduction techniques to estimate moisture content from the GPR data.
- Conduct a second field study for the purpose of validating the techniques developed under No. 2.
- Disseminate information in professional journals (2 papers per year).
Methods, Procedures, and Facilities (4 pages)
This proposal described the field study and related laboratory analyses. The GPR image resolution and noise reduction techniques will be developed under a separate project, and therefore are not described here.
Year 1
The purpose of the field study is to measure soil moisture content to a depth of 1 meter usinga GPR at locations were the vertical soil moisture distribution is known. This will be accomplished by collecting spatially continuous GPR data along transects adjacent to TDR access tubes. The experimental layout is shown in Figure 1, which will be established on the University of Puerto Rico-Mayagüez Campus.
The study area will consist of four groups of TDR access tubes aligned along two perpendicular axes. The access tubes spacing will increase exponentially from the inner tubes to the outer-most tubes. At the center point, a sprinkler will be installed to apply water to the surface as needed. Water from the sprinkler will be applied in a pattern similar to precipitation pattern shown in Figure 2. Note that the maximum depth of water is applied at the sprinkler location and gradually decreases to zero at the outer edge. The set-up shown in Figure 1 will provide four access tubes which will receive the same depth of water, and consequently, the vertical moisture content distribution will be approximately identical as well. The area will be carefully selected so that the soil is laterally homogeneous and will be covered with turf grass. Vertical soil variations may exist, and in fact, are desirable so that the GPR method can be tested with several soil textures. During the study, water will only be applied during periods of low wind speed (< 1 m/sec) so as to maintain a symmetric application of water over the circular area.
Figure 1. Experimental Layout
Figure 2. Distribution Pattern and Precipitation Profiles from a Typical Double Nozzle Sprinkler under Fvorable Conditions (From Keller J. and R. Bliesner, 1990).
There will be ten sampling events conducted over a six month period. During a sampling event, the GPR will be moved along the transects indicated in Figure 1. TDR measurements will be made at six depths within each access tubes. The GPR and TDR measurements will be taken on the same day within several hours of each other. The replicated data point will allow for statistical comparisons between the TDR and GPR-derived moisture contents.
Prior to starting the sampling events, the TDR will be calibrated for the soil conditions. Based on our experience in using TDRs in Puerto Rico, it will be necessary to calibrate each distinct soil horizon. The calibration will be made within the experimental area, but away from the locations of the access tubes and GPR transects. The calibration tests will be performed by collecting soil samples next to five access tubes, at up to six depths corresponding to the vertical TDR measurements. The calibration will be performed over a one to two month period. This will allow large variations in soil moisture content to be achieved.
The soil samples will be analyzed for the gravimetric soil moisture content. The moisture contents by weight (w) will be converted to the volumetric moisture content by multiplying by the soil bulk density and dividing by the density of water. Bulk densities will be obtained from the analysis of undisturbed soil cores collected at up to six depths and at five locations.
The soil characteristic curves relating moisture content to pore water pressure (or suction potential) will be developed using the undisturbed soil cores. Soil characteristic data will be obtained using the pressure plate extraction technique (Klute, 1986). Using the soil characteristic curves, the soil moisture contents from the TDR and GPR can be converted to pore water pressure. The pore water pressure data along with the reflectivity data, will be studied very carefully with the goal of developing algorithms for direct determination of pore pressure. Developing soil characteristic curves from undisturbed soil samples is time consuming and expensive. If the pore water pressure could be estimated directly from GPR data, this would be of great value.
Year 2
During Year 2, a field study will be conducted to validate the GPR measured moisture contents (and possible pore water pressure) at another location. (What is actually being validated is the image processing algorithms developed during the first year.) A new site will be selected and an identical experimental system will be installed (see Figure 1). Ten sampling events will be conducted over a six month period. The TDR will be calibrated as was done during the first year. Bulk densities and soil characteristic curves will be developed as was the case during the first year. Statistical comparisons of the TDR and GPR will be performed.
During the second year, a study will be conducted to assess to what extent the soil characteristic curves are hysteretic. Hysteresis refers to the condition in which multiple pore water pressures are possible for a single value of the moisture content, depending upon the wetting history of the soil. In the 0 to 1 atm range, the draining/wetting analysis required to measure hysteresis, will utilize Tempe cells.
Related Research (3 pages)
The temporal distribution of soil water can be described by the Richard’s equation (Hillel, 1980):
∂ ∂t = - ∂(K(h) ∂h/∂x) /∂x - ∂(K(h) ∂h/∂y)/∂y - ∂(K(h) ∂h/∂z)/∂z - ∂K(h)/∂z + S equ. 1
where = volumetric soil moisture content; t = time; x,y,z = spatial coordinates; h = soil water pressure; K = unsaturated hydraulic conductivity; and S = Source or sink. Volumetric moisture content () can be related to soil pressure (h) via the soil characteristic curve. Expressions relating to h developed by van Gunuchten, Haverkamp and Brooks and Corey have been presented by Lappala (1987). As an example, the Haverkamp expression is presented below:
= (r) (hb/h)requ. 1
where soil moisture content; = soil porosity; r = residual moisture content; hb = air inlet pressure; = a pore size distribution indext that is a function of soil texture; and h = soil pressure.
Because of its non-linearity, equation 1needs to be solved numerically. Numerical techniques require discritizing the problem domain into small cells or elements. The initial conditions in a model such as equation 1, requires the specification of moisture content (or pore water pressure) at every model cell. Since field data is typically derived from point measurements, data between points must be interpolated. Interpolation of data introduces potential error which can lead to erroneous results, especially in the case of non-linear models. Large prediction errors resulting from small errors in the initial conditions has been referred to as the butterfly-effect (Lorenze, 1963).
Traditional point methods for measuring the soil moisture content include the gravimetric, time domain reflectometry (TDR), resistance, capacitance and neutron scatter methods (Selker et al., 1999). Due to the heterogeneous nature of soils and their properties, these methods have a high degree of inherent uncertainty.
The ground penetrating radar (GPR) approach can provide a closer approximation to the true one, two or three-dimensional distribution of soil moisture content (REF). The method also has the advantage that it is not intrusive or destructive, a negative characteristic of the traditional methods.
Various researchers have begun using the GPR method for estimating soil moisture content in shallow soils (e.g., Graves et al., 1996; Lesmes et al., 1999; Lesmes et al., 2000; and Dannowski and Yaramanci, 1999). Van Dam and Van Den Berg (2001) have indicated that GPR is valuable for understanding sedimentary structure because small textural-property variations in sedimentary structures leads to changes in capillary pressure and water content.
References
Daniels, D. J. 1996. Surface-Penetrating Radar. Institute of Electrical Engineers. Short Run Press Ltd., Exeter. pp 300.
Dannowski, G. and U. Yaramanci. 1999. Estimation of water content and porosity using combined radar and geoelectrical measurements. TechnicalUniversity of Berlin, Department of Applied Geophys., Ackerstr. 71-76. D-13355 Berlin, Germany.
Hillel, D. 1980. Fundamentals of Soil Physics. Academic Press, New York.
Keller J. and R. Bliesner, 1990. Sprinkler and Trickle Irrigation. Van Nostrand Reinhold Publisher.
Klute, A. 1986. Methods of Soil Analysis Part 1 Physical and Mineralogical Methods. Agronomy Monograph 9, Second Edition. Agronomy Society of America and the Soil Science Society of America.
Lappala, E. G., R. W. Healy, and E. P. Weeks, 1987. Documentation of Computer Program VS2D to Solve the Equations of Fluid Flow in Variably Saturated Porous Media. U.S. Geological Survey. Water Resources Investigations Report 83-4099.
Lorenz, Edward N. "Deterministic Nonperiodic Flow." Journal of Atmospheric Science. 20 (1963): 130-141.
Selker, J. S., C. K. Keller and J. T. McCord, 1999. Vadose Zone Processes. Lewis Publishers, Boca Raton, USA.
Greaves, R. J., Lesmes, D. P., Lee, J. M., and Toksöz, M. N., 1996, Velocity variations and water content estimated from multi-offset, ground-penetrating radar, Geophysics, 61, 683-695.
Lesmes, D. P., Herbstzuber, R., and Wertz, D., 1999, Terrain permittivity mapping: GPR measurements of near-surface soil moisture. Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems, March 14 - 18, Oakland, CA, 575-582.
Lesmes, D. P., Japitana, J., Wertz, D., and Stevens, N., 2000, Ground-based geophysical investigations of soil moisture variations and soil heterogeneity during the Southern Great Plains Hydrology Experiment 1999, Spring AGU, Washington D.C.
Van Dam, R. L. and E. H Van Den Berg. 2001. Understanding GPR reflection in sediment. Abstract of a talk presented at the conference: GPR in sediments: applications an interpretations. August 20-21, 2001, London, UK.
Training Potential (1 page)
Hamed: 1 MS grad students (2 years)
Eric: 1 undergraduate student (2 years)
1 Graduate student (1 year)
Investigator’s Qualifications (4 pages)
ERIC W. HARMSEN - CURRICULUM VITAE
PERSONAL:
Address: / Agricultural and Biosystems Engineering DepartmentUniversity of Puerto Rico
PO Box 9030
Mayagüez, PR 00681-9030
Telephone: / (787)832-4040 ext. 3112 / E-mail:
EDUCATION
1991 Post Doctorate, Soil Science, North CarolinaStateUniversity
1989 Ph.D., Agricultural Engineering, University of Wisconsin-Madison
1984 M.S., Agricultural Engineering, MichiganStateUniversity
1981 B.S., Agricultural Engineering, MichiganStateUniversity
PROFESSIONAL AFFILIATIONS
Associate Research Editor, Journal of Soil and Water Conservation;
Reviewer for Journal of Irrigation
Member Soil and Water Conservation Society
Member American Society of Agricultural Engineers
EXPERIENCE
1999-Present: Assistant Professor Agricultural BiosystemsEngineering Department, University of Puerto Rico, Mayagüez, PR.
1994-1999: Chief Engineer/Manager of Environmental Modeling, ICF Kaiser Engineers, Pittsburgh, Pennsylvania.
1991-1994: Manager, Pittsburgh Groundwater Modeling Section, International Technology Corporation, Pittsburgh, Pennsylvania.
1989-1991: Post Doctoral Study, Department of Soil Science, North CarolinaStateUniversity.
1985-1989: Research Assistant (Ph.D. Research), Department of Agricultural Engineering, University of Wisconsin-Madison.
1988: Teaching Assistant, Department of Agricultural Engineering, University of Wisconsin-Madison.
1985-1988: Software Developer, Diversified Software Co., Owner, Madison, WI.
1984-1985: Environmental Engineer, State of Wisconsin Department of Natural Resources
1983-1984: Building Plan Examiner, State of Wisconsin Department of Industry Labor and Human Relations, Safety and Buildings.
1983: Consultant, Department of Agricultural Engineering, MichiganStateUniversity.
1983: Instructor, Department of Agricultural Engineering, MichiganStateUniversity.
1981-1983: Research Assistant, Department of Research Assistant (M.S. Research) Agricultural Engineering, MichiganStateUniversity
1981-1983: Teaching Assistant, Department of Agricultural Engineering, MichiganStateUniversity.
1980-1981: Soil and Water Technician, Department of Agricultural Engineering, MichiganStateUniversity.
1980: Computer Programmer, MichiganStateUniversity Cooperative Extension Service.
ERIC W. HARMSEN - CURRICULUM VITAE, Page 2
LIST OF PUBLICATIONS:
Harmsen, E. W., M. R. Goyal, and S. Torres Justiniano, Estimating Evapotranspiration in Puerto Rico. Puerto Rico Journal of Agriculture. In Press
Harmsen, E. W. and S. Torres Justiniano, 2001. Estimating Island-Wide Reference Evapotranspration for Puerto Rico Using the Penman-Monteith Method. ASAE Paper No. 01-2174. 2001 ASAE Annual International Meeting, Sacramento Convention Center, Sacramento, CA, July 30-August 1.
Harmsen, E. W. and S. Torres Justiniano, 2001. Evaluation of prediction methods for estimating climate data to be used with the Penman-Monteith method in Puerto Rico. ASAE Paper No. 01-2048. 2001 ASAE Annual International Meeting, Sacramento Convention Center, Sacramento, CA, July 30-August 1.
Harmsen, E. W., 2001, “Reevaluation of Predictions of Pumpkin and Onion Water Consumption at Two Locations in Puerto Rico.” Proceeding of the SixthCaribbeanIslands Water Resources Congress. February 22-23, 2001 – Mayagüez, PR.
Munster, C.L., Skaggs, R.W., Parsons, J.E., Evans, R.O., Gilliam, J.W., E. W. Harmsen. "Aldicarb Transport in a Drained Coastal Plain Soil." ASCE Journal of Irrigation, 1995.
Harmsen, E. W., J. W. Gilliam, R. W. Skaggs, and C. L. Munster, 1991, "Variably Saturated 2-Dimensional Nitrogen Transport," presented at the 1991 International Meeting of the American Society of Agricultural Engineers (ASAE), Chicago, IL. ASAE Paper No. 912630.
Harmsen E. W., J. C. Converse, and M. P. Anderson, 1991, “Application of the Monte-Carlo Simulation Procedure to Estimate Water-Supply Well/Septic Tank-Drainfield Separation Distances in the Central Wisconsin Sand Plain,” Journal of Contaminant Hydrology, 8(1991); 91-109.
Harmsen, E. W., 1983, “A Portable Chamber to Measure Plant Water Use: Design Considerations and Analysis,” M.S. thesis, MichiganStateUniversity, East Lansing, MI.
Harmsen, E. W., T. L. Loudon, G. A. Peterson, and G. E. Merva, 1982, “A Portable Chamber Technique for Evapotranspiration Measurement,” American Society of Agricultural Engineers, Paper No. 82-2598.
Harmsen, E. W., V. F. Bralts, T. L. Loudon, and F. J. Henningsen, 1982, “Computerized Irrigation Scheduling,” American Society of Agricultural Engineers Paper No. TSR82-201.
Hamed Parsiani Resume
BUDGET JUSTIFICAITON - YEAR 1Salaries and Fringe Benefits / PRWRERI / Remote Sensing Project / UPR
PD Academic Release Time, 8.33% (1 credit) / $0 / $4,165
Co-PD, Academic Release Time 8.33 / $0 / $4,998
$0
$0
$0
One (1) Graduate Student ($9,000 per year) / $0 / $9,000
Undergraduate Student - 1 student @ 10hrs/wk @ $6.5/hr*40 wks / $2,600
$0
$0
Total Salaries and Wages / $2,600 / $9,000 / $9,163
Fringe Benefits, 9.2% for E5, E6, E10-E12 grades, for additional compensation / $239 / $1,686
Fringe Benefits, 1.55%, Undergraduates / $40
Total Salaries and FB / $2,880 / $9,000 / $10,849
Other Direct Costs
a. Subcontract
b. Consultant
c. Equipment
Sprinkler Irrigation System / $150
GPR with Antennas / $30,000
TDR Soil Probe (PR1/6d 1-m profile probe) (quantity 1) / $1,665
TDR access tubes (quantity 6) / $1,380
SoilMoisture Soil Core Sampler; 2-1/4" diameter (quantity 1) / $583
Brass Sampling Cores - 6 cm long (quantity 30) / $390
Soil Pressure Plate Extraction Kit (pressure plates, extractors, compressor, manifold) / $11,000
Refrigerator and drying oven (for soil samples) / $1,300
Analytical Balance / $650
Total Equipment / $17,118 / $30,000 / $0
d. Supplies
$0
Total Supplies / $0
e. Travel
Local Travel / $0
$0
$0
Total Travel / $0
Total Direct Costs / $19,998 / $39,000 / $10,849
Indirect Cost (UPR: 48%) / $0 / $5,208
Subtotal -Estimated Costs / $19,998 / $39,000 / $16,057
BUDGET JUSTIFICAITON - YEAR 2
Salaries and Fringe Benefits / PRWRERI / Remote Sensing Project / UPR
PD Academic Release Time, 8.33% (1 credit) / $0 / $4,165
Co-PD, Academic Release Time 8.33 / $0 / $4,998
$0
$0
$0
One (1) Graduate Student ($9,000 per year) / $9,000
Undergraduate Student - 2 student @ 10hrs/wk @ $6.5/hr*40 wks / $5,200
$0
$0
Total Salaries and Wages / $5,200 / $9,000 / $9,163
Fringe Benefits, 9.2% for E5, E6, E10-E12 grades, for additional compensation / $478 / $1,686
Fringe Benefits, 1.55%, Undergraduates / $81
Total Salaries and FB / $5,759 / $9,000 / $10,849
Other Direct Costs
a. Subcontract
b. Consultant
c. Equipment
Antenna / 3000
Sprinkler Irrigation System / $500
Total Equipment / $500 / 3000 / $0
d. Supplies
$0
$0
Total Supplies / $0
e. Travel
Travel to professional meeting (PI) / $1,800
$0
Total Travel / $1,800
Other Direct Cost: Paper publication costs / $2,000
Total Direct Costs / $10,059 / $12,000 / $10,849
Indirect Cost (UPR: 48%) / $0 / $5,208
Subtotal -Estimated Costs / $10,059 / $12,000 / $16,057
Budget Explanation
Matching Funds Committement Letter
Some literature I need to read:
Water balance and numerical modeling studies require soil property information. Specifically, soil particle and bulk density, soil porosity, residual moisture content (i.e., wilting point),moisture content at the air inlet pressure (e.g., field capacity), and soil particle size. Soil water content with time is also required.
1