Project Year 2 / Progress Report

Project Year 2 / Progress Report

For the Period September 30 -- April 1, 2003

Research Area :

Environmental Applications

Descriptive Project Title:

Remote Sensing Applications for Environmental Assessment & Forecasting

Participants

Faculty: Dr. Hamed Parsiani , Co-PI & Dr. Eric Harmsen, Research Associate

Graduate student:Maritza Tores, Leonid Tolstoy, Carlos J. Gonzalez

Undergraduate student: Daniel Rodriguez, Eli Garcia , Ruth Riechehoff,

Staff: Waleska Campos, Manager and Pieter Van der Meer, GIS Specialist and Facilitator

1. Progress of Research

1.1 Equipment used in this research:

1)A ground penetrating radar SIR-20 with 1.5 GHz antennas, Fig.1.

2)Trace system-1, 1502 Time Domain Reflectometer (TDR) tektronicx, using electromagnetic waves reflection, a graphical read-out of a pulse reflection against time.

3)Gravimetric equipment to measure soil moisture content

Fig. 1: Ground Penetrating Radar, SIR-20 Fig.2 TDR equipment Fig. 3GPR Hyperbolasignature

GPR Reflections are obtained at the boundary of two media of sufficiently different dielectric constants.. The depth resolution in sand of GPR antenna of 1.5 GHz is about 10 inches, and the approximate subsurface radiation pattern is shown in Fig. 4.

Fig. 4 1.5 GHz antenna subsurface

radiation pattern

1.2 Research Procedure:

Reflectors were laid at different depths to receive sharper reflections and be able to measure relative velocity without the aid of a depth parameter. The reflections of the GPR off of the reflector are in the form of a hyperbola, as shown in Fig. 3.

1.2.2 Sand Moisture Content Determination

A large plastic box was filled with sand which consisted of two layers of dry (upper 10 cm) and moist (lower 10 cm). A metal rod separated dry from wet sand, and a second metal rod separated the wet sand from the underlying dry sand. These metal bars that were placed at 10 cm and 20 cm acted as reflectors against the incoming electromagnetic waves. The GPR produced-image, Fig. 5, shows two large hyperbolas indicating the presence of the two metal bars. The GPR is capable of measuring the dielectric constant in the 0 – 10 cm and the 0 – 20 cm depths, but cannot measure the 10 -20 cm layer directly. The procedure described in this paper allows for the determination of the soil moisture content from the 10 - 20 cm depth using a GPR.

1.2.3 Dielectric constant determination

The curve of the hyperbola is a function of the average velocity of the electromagnetic wave having traveled from the surface of the sand to the metal rod reflector and back. In Fig. 5, the upper hyperbola help determine the average wave velocity within the dry region (upper 10 cm), and the lower hyperbola leads to the determination of the average wave velocity within the 20 cm region, which is the combined dry and wet regions. The average velocity allows the calculation of soil dielectric constant.

The depth-specific dielectric constant can not be obtained directly by GPR, with the exception of the top layer. The inverse procedure [1] is used to calculate the depth-specific dielectric constant from the GPR-obtained bulk-average dielectric constants. These dielectric constants are converted to percent moisture using an appropriate mixing model, to be considered next.

1.2.4 Moisture calculation based on a Mixing Model selection

Several dielectric mixing models for relating dielectric constant to moisture content were evaluated for use in this study (Dobson et al., 1985; Wang and Schmugge, 1980; Martinez and Byrnes, 2001; Rial and Han, 2000; Alharthi and Lang, 1987; Shutko and Reutov, 1982). The semiemperical model of Dobson et al. (1985) was the only method considered that accounted for frequency; however, it did not provide accurate estimates of moisture content from the dielectric constant. The Wang and Schmugge (1980) method performed best and was selected for use in this study. The results of the model is depicted by graph of Fig. 5.

Figure 5 shows the Wang and Schmugge mixing model which will be used to convert dielectric constant of sand to percent volumetric soil moisture content.

1.2.5 Moisture Measurement Based on Gravimetric Method:

Actual soil moisture was measured in each of the soil layers by the gravimetric method. In this method, the volumetric moisture content of the soil is obtained by eq.1.

Volumetric Soil Moisture Content= (ρb/ρw)[(Wwet – Wdry) /Wdry ] (1)

where Wwet is the soil wet weight, Wdry is the dry weight of the soil, ρb is the soil dry bulk density (1.6 gm/cm3), and ρw density of water (1 gm/cm3). The soil was dried in an oven at 105 oC for 24 hours. The soil dry bulk density was obtained by dividing the soil dry weight of an undisturbed soil core by its volume.

2. Results results

The dielectric constants obtained by GPR and the depth specific ones calculated by the inverse procedure [1] were converted to percent moisture contents, Tbl. 1. The TDR measured dielectric constants in the upper 10 cm, and the hidden 10 cm layer (the layer below the upper 10 cm layer) were also converted to percent moisture contents, Tbl.1. The mixing model of Wang and Schmugge were used to calculate the moisture content from dielectric constants obtained by both GPR and TDR.

The gravimetric results were obtained for the two depth-specific layers, Tbl. 1. The Benedetto conversion model [3] was calculated for 1.5GHz antenna, Fig. 7, and these results were also used for comparison purposes, as shown in Tbl. 1. This result was presented and to be published in [4].

Depth / Gravimetric / GPR / TDR / Benedetto
0 – 10 cm / 0. 87 / 0 / 3.02 / 0
10 – 20 cm / 16.03 / 15.04 / 14.71 / 14.5

Table 1. Volumetric moisture (%) by gravimetric, GPR and TDR methods

3. Conclusion:

The percent moisture content obtained by GPR corresponds well to other methods, and verifies the results of the Benedetto model for 1.5 GHz antenna. Furthermore, the calculation of the depth-specific moisture average was realized, using the inverse procedure.

The average moisture content calculated at specific depths, based on bulk-average GPR measurements must be extended, in the future, to more layers and tested with TDR. Experimental work, however, becomes more laborious as the number of distinct moisture layers is increased.

4. References:

[1] Eric Harmsen & Hamed Parsiani, 2003. “ Inverse Procedure for Estimating Vertically Distributed soil Hydraulic Parameters using GPR,” NOAA-CREST-EPSoR Joint Symposium for Climate Studies Conference, Mayaguez, Puerto Rico, Jan. 2003.

[2] Wang, J. R and T. J. Schmugge, 1980. An empirical model for the complex dielectric permittivity of soils as a function of water content. IEEE Transactions on Geoscience and Remote Sensing, Vol. GE-18, No. 4:288-295.

[3] A. & F. Benedetto, “GPR Experimental Evaluation of Subgrade Soil Characteristics for Rehabilitation of Roads,” Ninth International Conference on Ground Penetrating Radar, Proceedings of SPIE Vo. 4758, 2002.

[4] Hamed Parsiani & Eric Harmsen & et al. “Validation of an Inverse Procedure for Estimating for Estimating Soil Moisture Content using GPR”, 4th NOAA annual conference on Expanding Opportunities in Oceanic & Atmospheric Sci.-2003”, March 30-April 1, 2003, Tallahassee Fl.

5. Immediate Research Plans

Soil test area set up, and soil parameter measurements per area:

Hamed, the colored text is just my comments and you can delete it after reading it.

I think we should do the following

1. Conduct a test in the Finca on the sand with the bars sufficiently spaced so that we can look at four or five layers of different soil moisture. Include additional sensors (tensiometer, theta-probe, profile probe, etc.)
2. Conduct a similar series of tests on the Loam.
3. Dry/wet layer test
4. Variable moisture (four or five layers)

1. Conduct a series of tests on the Clay
2. Dry/wet layer test
3. Variable moisture (four or five layers)

The above work could be completed by the end of the summer. We could then do the test at the Experiment Stations during the first semester. In those cases, we should wait until climatic conditions produce significant variations in the vertical soil moisture use irrigation to produce the desired variability.

The laboratory results will provide information necessary to accurately predict the vertical distribution of the water potential (negative pressure), hydraulic conductivity and soil water flux. The data will also be used in the dielectric mixing model, where we are currently used generic data for the sand. The SoilCon data can also be used for data fusion purposes, however, to me this task is still vague in my mind.

** Another tie-in with Ramon’s stuff would be to estimate moisture content using GPR at the four weather stations in Lajas corresponding with the weather station data and the satellite image.

** In one of these reports we should say that we are going to collect data as part of the ATLAS Mission.

Please call me after you have read this.

The test area will contain three sub-areas of sand, silt and clay, respectively. The soil will be analyzed for pH, electrical conductivity, point of zero net charge, anion exchange capacity, cation exchange capacity, exchangeable cations, organic C, and free iron-aluminum oxides. The physical properties that will be determined are texture, hydraulic conductivity, specific surface of both the soil and the clay fraction, aggregate stability, bulk density, porosity and the soil suction vs. moisture content.

Due to limited facilities at UPRM, most of these analyses have already been sent to the SoilCon-Ltd, and we are expecting results soon.

Actual measurement of soil moisture considering more moisture layers in larger areas of sand, silt and clay.

Future Research Plans

A future study will consist of measuring spatially continuous soil moisture contents in naturally occurring clay and silty loam soils. Data will be collected at the UPR Experiment Stations at Isabela, Lajas and Fortuna, PR. As in the initial phase of work, these soils will be analyzed for their physical and chemical properties.

1) Soil Moisture Measurements:

Point measuring devices 1-5 will be used to create measurements of soil moisture at different points on the three sub-areas of sand, silt, and clay. Spatial measurements will be done on the same areas using the SIR-20 1.5 GHz, and 900 MHz antennas separately. Images of subsurface will be obtained directly from the SIR-20, and by interpolation in the case of each of the 1-5 sensors.

2) Sensor Data Fusion for Automatic Moisture Determination:

Literature search is in progress in the area of data fusion and decision, considering both parametric and information theoretic methods.

3) Prepare new course modules: on the subject of ground penetrating radar, understanding and operation of the SIR-20.

Existing Coordination with CREST Members

Soil moisture content is an essential parameter in climate change assessments and forecasting, which is the prime emphasis of this collective research. Close collaborative work with a graduate student of Dr. Ramon Vasquez has been underway in the understanding of the problems related to the soil dielectric constant measurements.

Planned Coordination with CREST Members

The moisture data once obtained by SIR-20 will be made available to Dr. Vásquez research group for their atmospheric model needs. The data will also be made available to other CUNY-CREST members through UPRM group leader Dr. Vásquez. Dr. Eric Harmsen who is an agricultural engineer and a soil expert is our new Research Associate with NOAA-CREST, collaborating in the area of Field Measurements and Validation of soil moisture data.