Indirect Gpr Applications To Agricultural Investigations

Agricultural utilization of remote sensing to detect, identify, locate, map, predict, or estimate a buried feature or object that may affect production or management is an increasing application for GPR. This section will discuss the various ways that radar has been used for these indirect purposes. Let us begin by looking at studies involving GPR to detect and monitor groundwater.

GPR investigations to detect and monitor groundwater have several components. It has been used to estimate soil water content (Serbin and Or, 2004) during irrigation and drainage (Galage-dara et al., 2005); to identify subsurface flow pathways (Collins, et al., 1994; Gish et al., 2002; Kow-alsky et al., 2004; Gish et al., 2005) and nitrogen loss (Walthall et al., 2001); to estimate moisture contents in the vadose zone (Alumbaugh et al., 2002); to survey perched water on anthropogenic soils (Freeland et al., 2001); to estimate volumetric water on a field scale (Grote et al., 2003); and to map spatial variation in surface water content to compare GPR to time domain reflectometry (Huisman et al., 2002).

Serbin and Or (2004) reported using a GPR with a suspended horn antenna to obtain continuous measurements of near surface water content dynamics. These measurements were made in sand over silt loam textures. They concluded that radar enabled them to verify radar measurements at well-defined spatial scales and detailed temporal resolutions not available by other remote sensing techniques. Galagedara et al. (2005) went one step further to estimate water contents with GPR under irrigation and drainage conditions. Specifically, they were interested in determining the optimal ground wave sampling depth under irrigation and drainage situations.

Identifying subsurface flow pathways is a very important application of GPR because of what (e.g., fertilizers, pesticides) may move with the soil water. Several investigations (Collins et al., 1994; Gish et al., 2002, 2005; Kowalsky et al., 2004) documented potential subsurface water movement. Collins et al. (1994) was one of the first to do so and will be discussed as an example.

A large portion of north-central Florida is located in a "bare-karst" region. The limestone can be exposed at the surface or within a depth of a few meters. This area is known for sinkholes opening "overnight!". A series of sinkholes (mother with two daughter dolines) did open up in a matter of a few days in a field and Collins et al. (1994) used this area for their investigation.

The study site was in a pasture in which horse riding and jumping took place (Collins et al., 1994). GPR was used to identify the size of the caverns and determine subsurface flow patterns in order to locate other potential sinkholes in the immediate area. Three grids (macro, medi, and micro) were created, and GPR transects were made. Collins et al. (1994) were able to determine the size of the dolines and create three-dimensional diagrams of subsurface flow. They reported that the surface topography had preferential flow toward the dolines, and the subsurface flow patterns were more complex. There are subsurface "depressions," and the flow patterns were toward these

FIGURE 3.2 Ground-penetrating radar (GPR) data showing a paleo-sinkhole. The sand thickness is approximately 1 m; underlying is an argillic horizon that drapes the limestone. Notice how the radar bands point down. The limestone is shown by the hyperbolas.

low areas. These subsurface low areas are not obvious by viewing only the surface topography. A conclusion one may arrive at after looking only at the surface diagram is that soil water and nutrient movements would be toward the dolines. But by studying the subsurface topography of the clay horizon and assuming it was acting as an aquatard, one can assess that solute movement would not necessarily be in the direction of the dolines but may follow another pathway. This research would not have been possible without the use of GPR.

Using GPR in karst environments has been one of the most successful applications of this geophysical technology. Paleo-sinkholes can be identified by radar in optimum soil conditions (Figure 3.2). Optimum conditions include relatively dry sands over a moist argillic horizon that "drapes" the limestone. Breaks in the draping of the argillic horizon may indicate the existence of a void or a cavity that has been filled in with collapsed material (Figure 3.3).

Gish et al. (2002) also did a study involving GPR to evaluate its use in identifying subsurface flow pathways in a 7.5 ha agricultural production field in Maryland. They commented that understanding subsurface stratigraphy was critical to obtain accurate estimates of water fluxes in

FIGURE 3.3 An area showing the sands above the argillic horizon that drapes the limestone. The depth to limestone varies greatly in short distances. This is located in an area in Florida known as "bare" karst.

agricultural fields. They concluded that geo-referenced GPR data sets have great potential to locate soil horizons that control subsurface water pathways. Gish et al. (2005) continued with their interest in identifying subsurface flow pathways by studying such occurrences in a corn (Zea mays L.) field. They measured subsoil water contents that supported the GPR-identified preferential flow pathways. The impact of the subsurface water pathways was observed by the increase in corn yield during a drought season. Their conclusion was that subsurface pathways exist and influence soil moisture and corn grain yield patterns.

3.5 GPR APPLICATIONS TO STUDY PLANTS

Literature on the use of GPR applications to study plants has been extremely limited. One of the main limitations has been the wavelength and resolution associated with commercially available antennae. Until recently, the antennae were not able to distinguish very small objects such as plant roots. Truman et al. (1988b) used GPR to assess root concentrations. This may have been the earliest research in the application of GPR to evaluate plant roots.

More recently, Butnor et al. (2003) reported they were able to measure loblolly pine (Pinus taeda L.) root biomass to a depth of 30 cm with the aid of a digital signal processed GPR. Correlation coefficients were highly significant (r = 0.86, n = 60, p < 0.0001) between the GPR estimates and the measured root biomass. They concluded that GPR could decrease the number of cores needed to determine tree root biomass and biomass distribution. Hruska et al. (1999) had done a similar study in the Czech Republic. They looked at the three-dimensional distribution of oak trees (Quercus petraea (Mattusch.) Liebl) with DBH = 14 to 35 cm to estimate the coarse root density. The GPR unit employed in this study was able to give a resolution of approximately 3 cm in all directions. They reported satisfactory results.

Wielopolski et al. (2002) were concerned with increased CO2 in the atmosphere. They believed that a considerable portion of CO2 could be sequestered in plant roots. But a suitable means to measure the root morphology, distribution, and mass without destroying the roots' environment was not available. From these root characteristics, they wanted to access below-ground rates and limits of carbon accumulations. Thus, using a 1.5 GHz impulse GPR system and off-the-shelf software, they were able to image root systems (morphology and dimensions) in situ. They concluded from their study that GPR could image a 2.5 mm root twig buried in sand (under ideal conditions), but that further work is required to improve the images of the plant root system. Also, they believed that with future developments, it would be possible to routinely image roots 2 to 3 mm in diameter.

3.6 GPR GOLF COURSE STUDIES

A golf course was the study site of a very interesting investigation performed by Boniak et al. (2002). A suitable playing surface on a golf course is important for play and aesthetics. Surface watering and subsurface drainage are necessary. When the underground drainage system is not properly working, substantial damage may be done to the golf course resulting in a loss of playing time. The exact location of the underground tile drainage system is not known on many golf courses. To correct a problem, considerable destruction may be done to the course. Thus, the objective of their study was to locate and map the tile drainage system under a putting green using the radar. By doing so, it was expected that less physical destruction would take place to repair the tile because the exact locations of trouble areas could be accurately located with the GPR. Several intriguing findings were a result of this study.

Initially, the 900 MHz antenna was used because of its ability to detect shallow features with high resolution, but it was not successful because of a recent application of granular fertilizer. The fertilizer (high salt content) created noise and distortion to the radar data. As a result, the 400 MHz antenna was used for the study. A discontinuity was located where the golfers would exit the greens.

This resulted in soil compaction (higher bulk density and moisture content) in the area and affected the propagation velocity of the radar signals.

The conclusions of Boniak et al. (2002) were that GPR is very effective in locating and mapping underground drainage tile beneath golf greens. One suggestion that they offered was that a high-accuracy Global Positioning System (GPS) be used to geo-reference each location in the course of doing the radar transects. In another publication, Chong et al. (2000) also used GPR to determine the root zone of golf greens.

3.7 INCORPORATION OF GPR AND GPS INTO GIS

In the last ten years, three-dimensional modeling of radar data has increased due in part to inexpensive, commercially available, and user-friendly software. Applications of geographic information systems (GISs) have increased for the same reasons. Therefore, more users are employing the technologies. But integrating all three—GPR, GPS, and GIS—has not been routinely appreciated. One of the first to explore this opportunity was Tischler (2002).

As discussed in Tischler (2002) and Tischler et al. (2002), GIS provides a means of storing, manipulating, analyzing, and displaying spatially distributed data in a two-dimensional or three-dimensional view. Combining the efficiency and practicality of GPR with the visual appeal and interpretive power of GIS is the next reasonable step in the development of both technologies. This can be accomplished using GPS. However, few methods exist for combining geo-referenced GPR data with GIS data sets, which would reduce time and costs while increasing the interpretive quality of the information.

Four models were developed (Tischler, 2002). Model 1 consisted of categorical data that predict the presence of sand or an argillic horizon based on GPR amplitude readings. Model 2 was a numerical raster model that predicted GPR amplitude values continuously throughout the field. Models 3 and 4 were both numerical raster models that predict the depth to argillic horizon but differ in their methods of generation. The GPR processing steps taken, specific objectives, and the quantitative analysis performed on the data, differentiate all the models. Summarizing his results, Tischler commented that Model 1 met the objectives initially defined but the statistical correlation was not as strong as expected. Model 2 was much more visually appealing than Model 1, but the end value of his predicted AA (change in amplitude) was not a good variable to model. Models 3 and 4 were similar in generation and display. Both models predicted the same variable, depth to argillic, and are raster models. Figure 3.4 and Figure 3.5 are examples of Models 1 and 2, respectively.

3.8 CONCLUSIONS

There is no doubt that geophysical techniques have an essential function in the agricultural research world. Only GPR was discussed in detail in this chapter, but a similar chapter could be dedicated to EM use for agricultural purposes, especially in regions where the soil conditions for GPR are limited.

Over the twenty plus years that GPR has been routinely used by agriculturists, many of the studies have had direct or indirect applications to agriculture production and management. Some of the direct applications include mapping bedrock depth in a glaciated landscape; conducting microanalyses of soil and karst; high-resolution mapping soil and rock stratigraphy; interpreting a fragipan; improving interpretation of water table depths and groundwater flow patterns; determining forest productivity on a glacial drift soil; and assessing Bt horizons in sandy soils and ortstein. Indirect applications of GPR include detection, identification, location, mapping, predicting, or estimating a buried feature or object that may affect production or management in an incidental fashion. This has been a significant and increasing use for GPR and includes many investigations in detecting and monitoring groundwater and nutrients. Some of these are estimating soil water content during irrigation and drainage; identifying subsurface flow pathways; approximating moisture contents in the vadose zone;

FIGURE 3.4 Model 1. The darker gray areas are voxels with a value of 1, indicating the argillic horizon. The ligher gray areas 1 are voxels with a value of 0, indicating sand (Ap and E horizons). The white areas are meant to help visualize the layer transition. (From Tischler, M. A., 2002, Intergration of ground-penetrating radar data, global positioning systems, and geographic information systems to create three-dimensional soil models, M.S. Thesis, University of Florida. Gainesville. With permission.)

FIGURE 3.5 Model 2. Integration of ground-penetrating radar (GPR) and Global Positioning System (GPS) data into geographical information system (GIS) application as viewed in ArcScene. Darker areas are soils that are dominantly sandy in texture. Lighter areas are soils high in clay content. Each layer is a krigged interpolation of GPR values from the corresponding soil depth. (From Tischler, M. A., 2002, Intergration of ground-penetrating radar data, global positioning systems, and geographic information systems to create three-dimensional soil models, M.S. Thesis, University of Florida. Gainesville. With permission.)

FIGURE 3.5 Model 2. Integration of ground-penetrating radar (GPR) and Global Positioning System (GPS) data into geographical information system (GIS) application as viewed in ArcScene. Darker areas are soils that are dominantly sandy in texture. Lighter areas are soils high in clay content. Each layer is a krigged interpolation of GPR values from the corresponding soil depth. (From Tischler, M. A., 2002, Intergration of ground-penetrating radar data, global positioning systems, and geographic information systems to create three-dimensional soil models, M.S. Thesis, University of Florida. Gainesville. With permission.)

surveying perched water on anthropogenic soils; determining volumetric water on a field scale; and mapping spatial variation in surface water content to compare GPR to time domain reflectometry.

The use of GPR to study plants, specifically plant roots and their biomass, has not received as much attention as the other applications. This application has slowly increased in recent years, mainly due to the development of high-frequency antennae. One of the latest applications of GPR data as well as GPS data has been to incorporate these data into GIS. This application will continue to grow as more users of GPR and GIS cooperate in their research and use the data to solve emerging agricultural issues.

0 0

Post a comment