264

GIScience & Remote Sensing, 2011, 48, No. 2, p. 264–279. DOI: 10.2747/1548-1603.48.2.264

Copyright © 2011 by Bellwether Publishing, Ltd. All rights reserved.

A Semi-automated GIS Model for

Extracting Geological Structural Information

from a Spaceborne Thematic Image

A. Dadon The Remote Sensing Laboratory, Jacob Blaustein Institutes for Desert Research,

Ben-Gurion University of the Negev, Sede Boker Campus 84990, Israel

A. Peeters The Desert Architecture and Urban Planning Unit, Department of Man in the Desert,

Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev,

Sede Boker Campus 84990, Israel

E. Ben-DorThe Remote Sensing and GIS Laboratory, Department of Geography

and the Human Environment, Tel Aviv University, Tel Aviv 69978, Israel

A. Karnieli1

The Remote Sensing Laboratory, Jacob Blaustein Institutes for Desert Research,

Ben-Gurion University of the Negev, Sede Boker Campus 84990, Israel

Abstract: This paper presents a semi-automated GIS model for extracting structural

information from a spaceborne imaging spectroscopy classification of sedimentary rocks by combining the IS classification with a digital terrain model. The output consists of a database with structural attributes, specifically the dip and strike, of the geological layers. The model was evaluated statistically for its accuracy with promis-

ing results, which demonstrate its potential to support field surveys, for geological mapping, for 3D modeling of the subsurface, and for geological spatial analysis.

INTRODUCTION

Geological mapping describes the spatial attributes of the geological strata (lay-

ers), mainly the extent of exposure of the strata and their structural attributes, which

refers primarily to tectonically induced processes such as the slope and aspect of

exposed strata. Structural analysis of geological layers in geologic mapping is primar-

ily performed on sedimentary rocks originally formed as horizontal strata, which may

have subsequently remained horizontal or have been folded or fractured. Structural

analysis is primarily based on two variables: dip and strike, which define the structural

1Corresponding author: email: karnieli@bgu.ac.il

gis model for extracting geological information 265

characteristics of the geological layer—i.e., its form, spatial orientation, and trend.

The strike is the direction (azimuth) of a geological surface, measured clockwise from

the true north and ranging between 0° and 359°, whereas the dip is the angle (inclina-

tion) of a geological surface measured from a horizontal plane (perpendicular to the

strike direction). The geometric inquiry of the position and orientation of geological

layers can be performed using several methods; the most common are: (1) in situ point

measurements; (2) graphical, by constructing scaled projections and cross-sections,

commonly used in engineering drafting; (3) graphical, using spherical projections

in order to re-project the structure onto a representation plane; and (4) numerical,

using trigonometric identities (Lahee, 1961; Compton, 1985; FGDC Geological Data

Subcommittee, 2006; Tarbuck et al., 2008).

The graphical and numerical methods can be executed on a conventional geologic

map. Deriving the strike is performed by stretching lines across contours crossed by

the edges of the exposed geological layer. The azimuth of such a line (strike line),

stretched between two successive points on the same contour, defines the strike (or aspect) of the geological layer between those points. The dip may be derived based on

calculating the distance between two successive strike lines at the point of sampling,

together with the elevation difference using Eq. (1):

β = arctan (H/L) , (1)

where β is the dip angle, H is the height difference between two successive strike lines,

and L is the distance between two successive strike lines.

The true thickness of the geological layer can also be calculated using trigonom-

etry as follows (Eq. 2):

True Thickness = W sin β + H cos β , (2)

where W is the extent of the exposed geological layer (width).

These methods demand extensive field surveys or careful examination of the geo-

logic map, including diligent manual sketching. Time consuming and labor intensive

as they are, they limit the number of sampling points, especially when dealing with

large areas (Suppe, 1985). In addition, in situ measurements are not always feasible

due to inaccessible terrain or topography.

The fundamental difficulty in mapping rock structure in the field, or in using a geologic map, is that rocks are opaque and only exposed at the surface. In sites of

possible great economic interest or scientific significance, data acquisition can be enhanced by drilling, excavation, and geophysical measurements such as seismology,

gravity, magnetic, and electrical resistance. Each of these methods is expensive, time

and labor consuming, and possesses specific inherent difficulties (Tucker and Yorston, 1973). To date, most knowledge about large-scale structures is derived by studying the

rock exposure at the ground surface.

More recently, remotely sensed data offers an attractive solution to overcome

some of the difficulties associated with traditional methods of mapping rock struc-

ture. Combined with innovative image processing techniques it has become a sig-

nificant data source for geological exploration (Lillesand and Kiefer, 2000). The main

266 dadon et al.

advantages are the potential to collect data for large-scale areas, to obtain data on

areas that are difficult to reach or not accessible via ground collecting methods, and to extract information in a fast and reliable way. In recent years, hyperspectral imagery or

imaging spectrometry (IS) has become a powerful tool in the Earth sciences. The high

spectral resolution of IS enhances ground-collected geological data and classification of geological layers, used for geological mapping and spatial analysis. This quality

enables the acquisition of image data in many narrow contiguous spectral bands,

thus producing detailed spectral reflectance of each pixel. IS techniques can be used to extract sub-pixel information and provide quantitative data on pixel components

(Elchi, 1987). For example, most minerals within rocks can be identified according to their spectral signatures, which result from electronic and vibration processes that are

stimulated by photons reaching mineral atoms and molecules. Although the transmit-

ted electromagnetic radiation enables the exploration of only the upper 50 µm of the

exposed surface, geology was the prime discipline for which IS was first used and suc-

cessfully applied. Examples for the use of IS in geology can be found in Hunt (1977),

Goetz et al. (1983, 1985), Clark and Roush (1984), Kruse (1988), Clark et al. (1990, 1993, 2002), Kruse et al. (1990, 1993, 1997, 2002), Beyth et al. (1993), Ben-Dor et al. (1994, 2002), Clark (1999), Van der Meer (1999, 2006), and Kruse and Boarman (2000).

Geographic information systems (GIS) are commonly used for extracting spatial

information from classified remotely sensed images. The integration of remote sensing and GIS has an added value for geologists due to the ability to extract and incorporate

spatial and structural attributes of geological layers with additional variables of the

site, such as topography, to understand physical processes in their spatial context, and

to obtain surface and subsurface information. These applications have been imple-

mented in various fields of geology, such as structural studies, geomorphology, and lithological mapping (Gupta, 1991).

Remotely sensed data has been traditionally used for 2D digital mapping of geo-

logical themes, such as rocks and minerals (Lillesand and Kiefer, 2000). Currently, with the development of sophisticated image processing and GIS techniques, these

data are also used for automated mapping and physical modeling of geological

features (Meentemeyer and Moody, 2000; Apel, 2006). Applications include 3D

visualization of the terrain and the subsurface, query and analysis of structural rela-

tionships and automated extraction of geological objects and their attributes, such as

in Meentemeyer and Moody (2000), Masumoto et al. (2004), and Apel (2006). Many

researchers have addressed this topic by using stereo SPOT images, photogrammet-

ric or multispectral image data in combination with digital elevation models, such as

Buchroithner (1984), Lang et al. (1987), Deffontaines and Chorowicz (1991), Lang

and Paylor (1994), Reimer et al (1996), Bilotti and Shaw (2000), Donnadieu et al.

(2003), and Rinaldi (2007).

It should be noted that in order to make remotely sensed data accessible for GIS

polygon-based analysis—e.g., for extracting structural information—the former must

first be translated into a vector topology, a process that is often based on prior manual digitizing of geological unit polygons.

The following presents the development, application, and verification of a GIS-based semi-automated model for extracting structural information, namely the dip and

gis model for extracting geological information 267

strike of the geological strata from remotely sensed data, without the need for prior

manual digitizing of polygons representing the geological layers.

METHODOLOGY

Input Data

The model is based on two data sources: (a) a classified geocoded geologic map derived from IS data; and (b) a digital terrain model (DTM) from which contour lines

are extracted. The geologic map includes a classification of geological strata repre-

sented as geocoded raster classes. The classes correspond to different geological layers.

The geologic map is a result of an automated supervised classification that was applied to an IS image. While multispectral data can also be used as an input for the classifi-

cation, in the current case IS was used in order to enhance classification of strata and consequently enhance the accuracy of extracted structural data. The presented model

made use of Earth Observing-1 (EO-1) Hyperion spaceborne imaging spectrometer

data (Folkman et al., 2001). Hyperion data has potential advantages over multispectral

instruments such as Landsat-ETM+ due to its high spectral resolution data that pro-

vides an enhanced level of information for atmospheric correction in order to derive

surface reflectance and achieve better classification results. The classification of the IR image is not presented in the current paper and is detailed in Dadon et al. (2010, 2011).

Both the classified image and the DTM are combined in a semi-automated process to extract structural attributes for each geological layer.

Structure of Model

The model was developed using ESRI’s ArcGIS® Desktop software package, ver-

sion 9.3. It can be divided into two major, automated components. The first component of the model combines the geological information (represented as polygons) and the

topographical data (DTM) to automatically create a GIS layer of intersecting contour

lines that define the relation between the two. The intersecting lines are used in the second component of the model to automatically derive strike lines that are used for

calculating the values of the strike and dip for each intersecting contour line. The fol-

lowing explains in detail the processes involved in the two components (steps A to J

are shown in Fig. 1):

Component I. A. Classification. The first step consists of image processing and includes an automated supervised classification of IS data. The classification is not an integral part of the current model. Any classification of remotely sensed geological data that results in classified geological layers can be used as an input into the GIS model. The IS classification that was used in the current paper consists of geocoded raster classes that correspond to the different geological layers. The IR classification process is detailed in Dadon et al. (2010, 2011).

B. Filtering. A majority filter was applied to the geocoded raster classes in order to reduce speckles and smooth the geological classification map. Two options exist in the ArcGIS® Majority Filter tool: one is applied to the four orthogonal neighboring

cells and the other is applied to the eight closest cells. The second option was used in

the current model in which the kernel of the filter is a 3 × 3 window.

268 dadon et al.

C. Vectorization. Conversion of the smoothed geocoded raster classes from raster-

to-vector format. The vectorized classes are represented as GIS polygons that corre-

spond to the different geological strata. The vectorized output is exported into separate

GIS layers of polygons, each representing a different geological class (stratum).

D. Spatial Adjustment and Polygon Smoothing. Classification results follow sur-face attributes that are not always uniform in their nature and thus, in some cases,

Fig. 1. Schematic flow chart representing the automated model. Component I = intersection of contour lines extracted from DTM with polygons of stratum. Component II = calculation of

strike and dip values.

gis model for extracting geological information 269

floating pixels that do not correspond to a geological unit appear, despite the applica-

tion of the majority filter. When automatic vectorization is applied to the classification, these single pixels appear as “holes” of extremely small areal features within the poly-

gons. These might result in errors in step (F), in which strata polygons are intersected

with the DTM. The vectorized GIS layers are first spatially adjusted to allow correct geographical location. This is followed by removing or filling the “holes” using GIS geoprocessing tools to form smooth and uniform polygons.

E. Extracting Contour Lines. The DTM data are introduced into the model and

contour lines and elevation values are extracted for a 10 m elevation interval. The

interval in which the contour lines are extracted depends on the vertical accuracy of

the DTM that is used. The contour intervals and consequently the height difference

value that is inserted in Eq. (1) in step (J) can be adjusted according to the DTM’s

vertical accuracy.

F. Intersecting the Strata and Elevation Data. Contour lines are intersected

together with the polygon edges for each geological stratum to create contour line seg-

ments that lay completely within the polygons of the stratum.

Post-processing. To ensure that erroneous lines are not used in the following pro-

cess, minimal post-processing is required to remove small residual line segments.

Component II. G. Extracting the Start and End Points (x, y) of Contour Lines.

The x, y coordinate values that represent the intersection points of the contours with

the polygon edges—x, y start and x, y end—are extracted and merged into one layer.

H. Converting Points into Strike Lines. The intersection points (start and end

points) between the contour lines and between the polygon edges are used to auto-

matically create polylines connecting the start and end points of each contour segment.

These represent the strike lines illustrated in Figure 2.

I. Deriving the Strike. Strike values are calculated based on the azimuth of the

strike lines.

J. Deriving the Dip. The dip is based on calculating the average distance between

two successive strike lines, together with the elevation difference between the succes-

sive lines using Eq. (1). The true thickness of the geological layer can also be calcu-

lated using Eq. (2).

The model can be applied repeatedly on each geological layer in an automated

mode. The final database consists of the extracted dip and strike attributes, represent-ing structural information for the different geological layers.

APPLICATION OF THE MODEL TO A CASE STUDY

Terrestrial cover such as aeolian sands, alluvial, and talus cover, as well as thick

vegetation remain obstacles for geological mapping by remote sensing methods, as

well as for conventional field mapping. Thus, several preconditions were set for select-ing the case study: (a) locations with arid conditions for high reflectance values; (b) low vegetative cover; and (c) locations with sedimentary layers.

The following case study was selected. Input data consisted of: (a) a classified IR image, which covers the area of the Dana National Geological Park, Jordan (35º18´30´´

E, 30º40´50´´ N). The park is situated at the eastern edge of the Dead Sea transform

fault system, which separates the Arabian and African tectonic plates (Fig. 3); and (b)

a DTM covering the same area that was extracted from the National DTM of Israel.

270 dadon et al.

The DTM was compiled by the Geological Survey of Israel from 1:50,000 topographic

maps using a raster-based scanning process detailed in Hall (1993, 2009). The DTM

provides the elevation data with a 25 m spatial resolution. Elevation intervals between

pixels are 1 m.

The study site consists of a diverse geological setting ranging from Precambrian

to Quaternary in age and comprises Precambrian igneous rocks and sedimentary rocks

including Cambrian massive sandstones, dolomites, cretaceous sandstones, and lime-

stone (Bender, 1974; Rabb’a, 1994). Additionally, the area is known for its ore min-

erals, such as copper, which was mined even in ancient times. In fact, from the Late

Neolithic period up to medieval times, the area was considered a major mining center

of the region (Rothenberg, 1997). To conclude, the combination of low vegetative

Fig. 2. Layer of intersecting contour lines for a polygon representing a stratum with marked

x, y start and end points and the strike lines derived from the points.

gis model for extracting geological information 271

Fig. 3. Location of case study.

272 dadon et al.

cover and high reflectance values, due to arid conditions, with the diverse geological setting and mineralogical assembly make the research area a quite suitable test site for

remote sensing of surface geology and structural research.

The IR image was classified into 15 geocoded raster classes, each class represent-ing a geological layer. These classes were evaluated for their accuracy in Dadon et

al. (2011). Results showed that all 15 formations were recognized within their spatial

arena successfully, and that the supervised classification coincided closely with the geological formations digitized from a conventional map. For example, the classes

Umm Ishrin sandstone (IN) and Na’ur Limestone (NL) demonstrated the highest

accuracy levels, with 86% and 88% user’s accuracy, respectively. As explained in the

introductory section of the paper, structural analysis of geological layers in geological

mapping is primarily performed on sedimentary rocks; therefore only sedimentary

geological layers were used.

The geocoded raster classes were converted into 15 polygon layers as explained

in step (C) of the methodology (Fig. 4). The polygons of each geological class were

intersected with the DTM and the strike and dip were calculated for the intersecting

contour lines.

MODEL VERIFICATION

To evaluate the performance of the model, the automated dip and strike results

were evaluated statistically in comparison with field data—i.e., compared to strike and dip measurements as they appear on a conventional geological hard-copy map.

Sampling points were extracted at locations that coincide with field dip and strike cal-culations (Fig. 5). Table 1 presents extracted (automated) strike and dip values vs. map

values (field data included in the conventional hard-copy geologic map of the Jabal Hamra Faddan area [Rabb’a, 1994]) at the points of sampling.

The Student’s t-test was applied to evaluate the level of similarity between the

automated and the field values for both the dip and the strike. The statistical results for the strike evaluation (t

0.05 (11) = 1.4, p = 0.19) show, by conventional criteria, that the

difference between extracted (automated) and field values (map) is considered not to be statistically significant (p > 0.05), therefore indicating that the automated strikes are

similar to the measured field strikes. Respectively, statistical results for the dip evalua-

tion (t0.05 (11)

= 1, p = 0.34) indicate that the difference between extracted and field dips is considered to be statistically not significant (p > 0.05).

Less correlative dip and strike values were mainly detected for extracted points

located next to faults, such as in points no. 0 and 1 (Fig. 5), and in geological layers

with poorly defined edges (e.g., layers covered with debris, aeolian sands, alluvium, colluvium, etc.). As the dip and strike strongly rely on the spatial relation between the

original edges of the stratum and the topography, the authenticity of the stratum edges

is significant. Consequently, extracted points located in a stratum bounded by a fault appeared as a discrepancy.

DISCUSSION AND CONCLUSIONS

The main advantage of the presented model lies in its ability to extract in-lab

structural data, which can serve as preliminary information, reducing the time-, labor-,

g

is m

od

el f

or e

xt

ra

ct

ing

g

eo

lo

gic

al in

fo

rm

at

ion

273

Fig. 4. A. Supervised classification of Hyperion image in raster format after applying a majority filter. B. Output of raster-to-vector conversion prior to polygon smoothing, including detail of map in vector format. Note the “holes” of extremely small areal features within the polygons that are removed at the following step. The classes displayed represent the geological units: Al = alluvium; Als = aeolian sand; Plg2 = fluvial gravel; NL = Na’ur limestone; KS = Kurnub sand-stone; IN = Umm Ishrin sandstone; BDS = Burj dolomite, shale; SB = Salib arkosic sandstone; MM = Minshar monzogranite; GR = Ghuwayr volcanic; HK = Hunayk monzogranite; AS = As Sabra granodiorite; FN = Finian granite.

274 dadon et al.

and cost-consuming processes involved in field surveying, drilling holes, and manual digitizing. In addition, it creates one framework for automated recognition and mul-

tiple geoprocessing tasks that can be repeatedly run on different images. Existing data

can be readily modified and new data and parameters can be incorporated in the pro-

cess. However, it is important to note that the method is not a stand-alone one, and is

suggested as a preliminary tool for understanding regional structural settings to assist

field surveys, or in cases where accessibility is limited. The limitations of the model are mainly set by:

1. The accuracy of automated recognition. To enhance the extraction of geological

structural data, the supervised classification of geological layers must result in a very highly accurate classification. The precision of the polygon edges that represent the

Fig. 5. Polygons representing different strata and sampling points for verifying the model, over-

laid on a 3D view of the topography. The points were extracted at locations that coincide with

field (map) dip and strike calculations.

gis model for extracting geological information 275

geological layer is particularly significant for assuring the accuracy of the results. An enhanced classification will also result in smoother polygons and reduce the “holes” within them. Currently, user interaction is still required for post-processing of small

residual line segments of intersected contour lines. An enhanced classification will permit a fully automated process.

2. The exposure of the strata. The method is currently applicable only to sedimen-

tary rocks, in moderate topography and when stratum edges represent their true inter-

section with the topography—i.e., when strata are well exposed, are without debris

or aeolian or alluvial cover, and are free of soil and vegetation. The method is not

applicable for geological layers bounded by faults and in cases where classification results are truncated at the edges of the layer. Additional research sites with different

geological settings are required, in the future, to develop a more generic model.

3. The accuracy of the DTM. The results are influenced by the vertical and hori-zontal accuracy of the DTM. The vertical accuracy, for example, will define the eleva-

tion accuracy of the extracted contours at different intervals. Low horizontal accuracy

might introduce errors at the points of intersection between contours and strata. The

accuracy of the model will obviously benefit from a DTM with a higher spatial resolu-

tion. Alternatively, other sources of elevation data can be used as input for the model,

such as local ground surveying data, photogrammetric data, or airborne laser scanning

data such as LIDAR.

While preliminary results demonstrate the workability of the model and agree-

ment with variables appearing in the conventional geologic map, the model will

improve following further research in the field of IS and the introduction of higher resolution DTMs. For example, advanced airborne IS sensors such as the Airborne

Visible/Infrared Imaging Spectrometer (AVIRIS) may be utilized.

In addition, it is important to note that a conventional geologic map is created not

simply according to on-surface spatial appearance of rock formations, but rather relies

Table 1. Extracted (automated) Strike and Dip Values vs. Map Values at Points of

Samplinga

No. of point Strike, automated Strike, map Dip, automated Dip, map

0 124 120 33 60

1 60 60 25 60

3 66 45 13 10

8 17 20 16 16

10 56 55 11 9

12 60 80 13 10

15 75 70 16 15

19 127 125 24 20

22 130 110 16 15

26 75 44 17 16

27 15 12 15 13

aPoint numbers are presented in Figure 5.

276 dadon et al.

on interpretation of stratigraphic relations combined with other information such as

drill holes. Interpretations of stratigraphic relations were not added to the classification data in the current research and may be a basis for further research on geological data

modeling and mapping.

Despite existing limitations, the model demonstrates the potential of integrating

GIS and image processing techniques to generate a comprehensive process for auto-

mated extraction of geological structural information. Furthermore, the model can be

applied to geological polygons attained by other means, such as manual digitizing. The

model and the constructed database serve as a basis for geological spatial analysis and

3D physical modeling and visualization. It provides field researchers with additional information needed to quantitatively study the relationship between the spatial loca-

tion and the geophysical characteristics of geological layers or minerals. The overlay

method, fundamental to a GIS system, allows the integration of thematic data such as

climate, soil, vegetation, hydrology, and land use, which can be added as GIS layers to

facilitate the comprehensive analysis of an area.

ACKNOWLEDGMENTS

This work was supported in part by an Eshkol Scholarship and by an Ilan Ramon

Scholarship from the Ministry of Science and Technology, The State of Israel. In addi-

tion, the authors wish to acknowledge the help of Wolfgang Motzafi-Haller.

REFERENCES

Apel, M., 2006, “From 3D Geomodeling Systems towards 3d Geoscience Information

Systems: Data Model, Query Functionality and Data Management,” Computers

& Geosciences, 32(2):222–229.

Bender, F., 1974, Geology of Jordan, Contribution to the Geology of the World, Berlin,

Germany: Bebrueder Borntraeger, 196 p.

Ben-Dor, E., Kruse, F. A., Lefkoff, A. B., and A. Banin, 1994, “Comparison of Three Techniques for the Utilization of GER 63 Channel Scanner Data of Makhtesh

Ramon, Negev, Israel,” Photogrammetric Engineering and Remote Sensing,

60:1339–1354.

Ben-Dor, E., Patkin, K., Banin, A., and A. Karnieli, 2002, “Mapping of Several Soil Properties Using DAIS-7915 Hyperspectral Scanner Data—a Case Study over

Clayey Soils in Israel,” International Journal of Remote Sensing, 23(6):1043–

1062.

Beyth, M., Henkel, J., and R. Geerken, 1993, Applying Image Processing Technique

and Reflectance Measurements Combined with Detailed Field Work for Analyz-

ing TM Data of an Arid Area, Southern Israel and Southwest Jordan, Jerusalem,

Israel: Geological Survey Israel Report GSI/25/93, 29 p.

Bilotti, F. and J. H. Shaw, 2000, “Quantitative Structural Analysis with Stereoscopic

Remote Sensing Imager,” AAPG Bulletin, 84(6):727–740.

Buchroithner, M. F., 1984, “Geological Mapping of Remote Mountainous Regions

using Metric Camera Imagery,” The Austrian Journal of Earth Sciences, 77:115–

149.

gis model for extracting geological information 277

Clark, R. N., 1999, “Spectroscopy of Rocks and Minerals and Principles of Spectros-

copy,” in Manual of Remote Sensing, Rencz, A. N. (Ed.), New York, NY: John Wiley & Sons, 3–58.

Clark, R. N., King, T. V. V., Klejwa, M., and G. A. Swayze, 1990, “High-Spectral Resolution Spectroscopy of Minerals,” Journal of Geophysical Research,

95(B8):12653–12680.

Clark, R. N. and T. L. Roush, 1984, “Reflectance Spectroscopy: Quantitative Analysis Techniques for Remote Sensing Applications,” Journal of Geophysical Research,

89(B7):6329–6340.

Clark, R. N., Swayze, G. A., Gallagher, A. J., King, T. V. V., and W. M. Calvin, 1993, “The U.S. Geological Survey, Digital Spectral Library: Version 1: 0.2 to 3.0

Microns,” U.S. Geological Survey Open File Report, 1340:93–592.

Clark, R. N., Swayze, G. A., Livo, K. E., Kokaly, R. F., King, T. V. V., Dalton, J. B., Vance, J. S., Rockwell, B. W., Hoefen, T., and R. R. McDougal, 2002, “Sur-

face Reflectance Calibration of Terrestrial Imaging Spectroscopy Data: A Tutorial Using AVIRIS,” in Proceedings of the 10th Airborne Earth Science Workshop,

Pasadena, CA: JPL Publication 02–1.

Compton, R. R., 1985, Geology in the Field, New York, NY: John Wiley & Sons, Inc., 416 p.

Dadon, A., Ben-Dor, E., Beyth, M., and A. Karnieli, 2011, “Examination of Space-

borne Imaging Spectroscopy Data Utility for Stratigraphic and Lithologic Map-

ping,” Journal of Applied Remote Sensing (in press).

Dadon, A., Ben-Dor, E., and A. Karnieli, 2010, “Use of Derivative Calculations and Minimum Noise Fraction Transform for Detecting and Correcting the Spectral

Curvature Effect (Smile) in Hyperion Images,” IEEE Transactions on Geoscience

and Remote Sensing, 48(6):2603–2612.

Deffontaines, B. and J. Chorowicz, 1991, “Principles of Drainage Basin Analysis from

Multisource Data: Application to the Structural Analysis of the Zaire Basin,”

Tectonophysics, 194(3):237–263.

Donnadieu, F., Kelfoun, K., Van Wyk de Vries, B., Cecchi, E., and O. Merle, 2003, “Digital Photogrammetry as a Tool in Analogue Modelling: Applications to

Volcano Instability,” Journal of Volcanology and Geothermal Research, 123(1–

2):161–180.

Elchi, C., 1987, Introduction to the Physics and Techniques of Remote Sensing, New

York, NY: John Wiley & Sons Inc., 413 p.FGDC Geological Data Subcommittee, 2006, “Digital Cartographic Standard for Geo-

logic Map Symbolization” [http://ngmdb.usgs.gov/fgdc_gds/index.php], retrieved

March 2010.

Folkman, M. A., Perlman, J., Lushalan, L., and P. Jarecke, 2001, “EO-1/Hyperion

Hyperspectral Imager Design, Development, Characterization, and Calibration,”

SPIE Conference Proceedings, 4151:40–51.

Goetz, A. F. H., Rock, B. N., and L. C. Rowan, 1983, “Remote Sensing for Explora-

tion: An Overview,” Economic Geology, 78(4):573–590.

Goetz, A. F. H., Vane, G., Solomon, J. E., and B. N. Rock, 1985, “Imaging Spectrom-

etry for Earth Remote Sensing,” Science, 228(4704):1147–1153.

Gupta, R. P., 1991, Remote Sensing Geology, Berlin, Germany: Springer-Verlag, 356

p.

278 dadon et al.

Hall, J. K., 1993, “DTM Project Scheme of 1:50,000 Topographic Sheet Mnemonics for Israel,” Geological Survey of Israel Current Research, 8:47–50.

Hall, J. K., 2009, “The 25m DTM (Digital Terrain Model) of Israel,” Israel Journal of

Earth Sciences, 57:145–147.

Hunt, G. R., 1977, “Spectral Signatures of Particulate Minerals, in the Visible and

Near-Infraread,” Geophysics, 42:501–513.

Kruse, F. A., 1988, “Use of Airborne Imaging Spectrometer Data to Map Minerals Associated with Hydrothermaly Altered Rocks in the Northern Grapevine Moun-

tains, Nevada and California,” Remote Sensing of Environment, 24:31–51.

Kruse, F. A. and J. W. Boardman, 2000, “Characterization and Mapping of Kimberlites and Related Diatremes Using Hyperspectral Remote Sensing,” in Aerospace Con-

ference Proceedings, IEEE 2000, Big Sky, MT, USA.

Kruse, F. A., Boardman, J. W., and J. F. Huntington, 2002, “Comparison of EO-1 Hyperion and Airborne Hyperspectral Remote Sensing Data for Geologic Appli-

cations,” in Aerospace Conference Proceedings, IEEE 2002, Big Sky, MT, USA.

Kruse, F. A., Kierein-Young, K. S., and J. W. Boardman, 1990, “Mineral Mapping at Cuprite, Nevada with a 63-Channel Imaging Spectrometer,” Photogrammetric

Engineering and Remote Sensing, 56:83–92.

Kruse, F. A., Lefkoff, A. B., Boardman, J. B., Heidebrecht, K. B., Shapiro, A. T., Barloon, P. J., and A. F. H. Goetz, 1993, “The Spectral Image Processing System

(SIPS)—Interactive Visualization and Analysis of Imaging Spectrometer Data,”

Remote Sensing of Environment, 44:145–163.

Kruse, F. A., Richardson, L. L., and V. G. Ambrosia, 1997, “Techniques Developed for Geologic Analysis of Hyperspectral Data Applied to Near-Shore Hyperspec-

tral Ocean Data,” in Proceedings, ERIM 4th International Conference, Remote

Sensing for Marine and Coastal Environments, Ann Arbor, MI: Environmental

Research Institute of Michigan (ERIM), I:I-233–I-246.

Lahee, F. H., 1961, Field Geology, 6th ed., New York, NY: McGraw-Hill Inc., 926 p.Lang, H., Adams, S. L., Conel, J. E., McGuffie, B. A., Paylor, E. D. and R. E. Walker,

1987, “Multispectral Remote Sensing as Stratigraphic and Structural Tool, Wind

River Basin and Big Horn Basin Areas, Wyoming,” AAPG Bulletin, 71(4):389–

402.

Lang, H. and E. Paylor, 1994, “Spectral Stratigraphy: Remote Sensing Lithostrati-

graphic Procedures for Basin Analysis, Central Wyoming Examples,” Natural

Resources Research, 3(1):25–45.

Lillesand, T. M. and R. W. Kiefer, 2000, Remote Sensing and Image Interpretation,

New York, NY: John Wiley & Sons, Inc., 724 p.Masumoto, S., Yonezawa, G., Shiono, K., Raghavan, V., and T. Nemoto, 2004, “Con-

struction and Visualization of a Three-Dimensional Geologic Model Using

GRASS GIS,” Transactions in GIS, 8(2):211–223.

Meentemeyer, R. K. and A. Moody, 2000, “Automated Mapping of Conformity between Topographic and Geological Surfaces,” Computers & Geosciences,

26:815–829.

Rabb’a, I., 1994, The Geology of the Al Qurayqira (Jabal Hamra Faddan). Map Sheet

No 3051 II. Amman, Jordan: Hashemite Kingdom of Jordan, Bulletin 28.

gis model for extracting geological information 279

Reimer, W., Miller, H., and H. Mehl, 1996, “Mesozoic and Cenozoic Palaeo-stress

Fields of the South Patagonian Massif Deduced from Structural and Remote

Sensing Data,” Geological Society: London (Special Publications), 108:73–85.

Rinaldi, M., 2007, “Remote Sensing Capability in Structural Geology Analysis of

Different Geodynamic Settings: The Example of Al Qarqaf Arch (Libya),”

Scientifica Acta 1(1):43–46.

Rothenberg, B., 1997, “Researches in the Southern Arabah 1959–1990. Summary of

Thirty Years of Archaeo-Metallurgical Field Work in the Timna Valley, the Wadi Amram, and the Southern Arabah (Israel),” Arx, 2–3:5–42.

Suppe, J., 1985, Principles of Structural Geology, Englewood Cliffs, NJ: Prenctice-

Hall, 537 p.

Tarbuck, E. J., Lutgens, F., and D. Tasa, 2008, Earth: An Introduction to Physical

Geology, Upper Saddle River, NJ: Prenctice-Hall, 638 p.

Tucker, P. M. and H. J. Yorston, 1973, Pitfalls in Seismic Interpretation, Tulsa, OK: Society of Exploration Geophysicists Monograph Series, No. 2, 50 p.

Van der Meer, F., 1999, “Can We Map Swelling Clays with Remote Sensing?,” Inter-

national Journal of Applied Earth Observation and Geoinformation, 1(3):234–

235.

Van der Meer, F., 2006, “The Effectiveness of Spectral Similarity Measures for the

Analysis of Hyperspectral Imagery,” International Journal of Applied Earth

Observation and Geoinformation, 8(1):3–17.