Integration of Satellite Imagery, Geology and Geophysical Data

Satellite imagery is a large scale surface geological mapping too, which offers the unique opportunity to investigate the geological characteristics of remote areas of the earth surface without the need to access the area on the ground. The resolution of the technique is limited by the resolution of the imagery. This chapter explains how geological information can be extracted from satellite imagery and how this information can be merged with geological and geophysical data to build consistent geological models for the surface and subsurface. On the one hand, the interpretation of satellite imagery can generate start models prior to the start of geophysical surveys. On the other hand, geological and geophysical data can calibrate models derived from satellite imagery.


Introduction
Satellite imagery is a large scale surface geological mapping too, which offers the unique opportunity to investigate the geological characteristics of remote areas of the earth surface without the need to access the area on the ground. The resolution of the technique is limited by the resolution of the imagery. This chapter explains how geological information can be extracted from satellite imagery and how this information can be merged with geological and geophysical data to build consistent geological models for the surface and subsurface. On the one hand, the interpretation of satellite imagery can generate start models prior to the start of geophysical surveys. On the other hand, geological and geophysical data can calibrate models derived from satellite imagery.

Methodology
When studying the shape of the earth surface in connection with the rock layers and their deformation by tectonic forces, we often notice a correlation between shapes and structures at the surface and in the subsurface (Short and Blair 1986). This opens the opportunity to map the characteristics of the surface and infer characteristics of the subsurface. We can describe the surface by its shape and by its structure. The surface shape depends on topography, terrain gradient and surface lithology, which we call geomorphological properties. The surface structure is determined by lithological boundaries and fracture zones outcropping at the surface, which we call litho-structural properties. Fracture zones can also be inferred from the characteristics of recent or paleo-drainage (Short and Blair 1986). Geomorphology and litho-structure allow building a static geological model. If information is available about the elevation change with time, then the statics model can be expanded into a dynamic geological model. Figure 1 gives an over of the building blocks for geological model building. We illustrate the methodology at a simple layer-cake geological model which is deformed by a vertical fault ( fig. 2). The surface is formed by a soft sandstone layer resting on a hard limestone layer. These two layers form the near-surface. We call the layers between the bottom of the near-surface and the top of the basement subsurface. Prior to the deformation by the fault only the soft sandstone was visible at the surface. The fault has lifted part of the layer package and exposed the near-surface sandstone and limestone layers at the fault plane and made them accessible for mapping by satellites.

Satellite imagery
Earth observation satellites map the physical properties of the earth surface and nearsurface. In the context of geological mapping we distinguish two types of electro-magnetic methods (see figure 3) :  passive optical methods: use the sunlight as the source and measure the reflectance of the earth surface in the visible and infrared spectral bands. We used Landsat 7 ETM+ and the ASTER instrument from the Terra satellite.  active microwave radar methods: use a microwave source onboard of the satellite and measure the back-scatter from the earth. We used Radarsat-1 and the radar sensor from the Shuttle Radar Tomographic Mission (SRTM). Details about the acquisition and processing are provided among others by the USGS (2011) and Short (2010). For an introduction into the interpretation of satellite imagery see Sabins (1996). The visible imagery (VIS) covers the colors blue, green and red and provides information about water features, infrastructure and landuse as well as limited information about selected rock types. Infrared imagery is split into three classes : very near infrared (VNIR), which detects specifically vegetation; short wave infrared (SWIR) which is the best option for the discrimination of sedimentary rocks; and finally thermal infrared (TIR). The thermal infrared radiation from the earth surface represents the property of the surface material to convert the solar spectrum into heat radiation. We distinguish between a warm response from dark materials such as non-sedimentary rocks and cool response from ground moisture or voids, where evaporation absorbs energy. In general optical imagery does not penetrate the earth surface.

Fig. 3. Spectral overview of electro-magnetic satellite imagery
Microwave radar uses electro-magnetic waves in the mm to m range. At hard surfaces microwaves are almost completely back-scattered. Their travel time can be used to determine the distance between the satellite and the surface, which is used to generate digital elevation models (DEM). For soft, non-conductive surfaces the microwaves penetrate into the subsurface; the back-scattered signal is generated from volume scatter in the subsurface. A special application of microwave radar is the estimation of gravity anomalies. Sandwell and Smith (2009) have studied anomalies in the orbits of radar satellites and inferred Bouguer gravity anomalies using a geoid model. These gravity anomalies can be interpreted for the thickness of the sedimentary cover above the crystalline basement.
The following examples illustrate the information obtainable from the various satellite imagery sets. Figure 4 shows an example of multi-spectral data from ASTER for each range : visible ( fig. 4.a), short wave infrared ( fig. 4.b) and thermal infrared bands ( fig. 4.c) of the island of Bahrain with a colour scale from blue (low reflection intensities) via green and yellow to red (high intensities). The visible green band penetrates water for a few meters (blue shades). It is relatively insensitive to different rock types (red shades). The short wave infrared band does not penetrate water and therefore shows uniform blue color, whereas it allows the discrimination of the different rock types in the oval anticlinal structure covering most of the island (yellow and red shades). The thermal infrared band shows the thermal properties such as warm response from the built-up areas (dark red). Cool response (greenish areas in the coastal areas) is observed for wet coastal salt flats called sabkha.  . 5.a), which shows mountain ranges (dark purple), gravel planes (yellow-gray), salt lakes (blue) and vegetation (green). The radar back-scatter intensity from Radarsat-1 ( fig. 5.b) is displayed from blue (total absorption in conductive salt brine) via intermediate volume back-scatter (green and yellow) to strong surface back-scatter from hard rock (brown and white). Figure 5.c shows the gravity anomaly inferred from radar satellite imagery. Negative anomalies correspond to low density rocks as from thick sedimentary cover at the foot of the www.intechopen.com Atlas mountains whereas positive anomalies correlate with dense metamorphic and basement rocks in the Atlas ranges.

Extraction of information from satellite imagery
Satellite imagery is provided as sets of digital images, one image for each spectral band. Each image displays the measured values as intensity. The information contained in the satellite imagery can be extracted using either single bands or combinations of bands. Single bands are usually displayed as maps coding measured amplitude as colour. Figure 6 shows radar data from the very dry desert in south-west Egypt : radar data from the shuttle radar tomographic mission (SRTM, Jarvis et al., 2008) interpreted for a DEM using topographic colour coding, i.e. green for low elevations, yellow to brown for high ground ( fig. 6.a). Raw radar data from Radarsat-1 reveal high amplitudes from the backscatter at hard sandstone (white and brown colors in fig. 6.b). Low radar back-scatter (fig. 6.c) correspond to volume back-scatter from microwaves penetrating sand sheets. They reveal buried paleo-rivers (blue colors in fig. 6.c) following the interpretation by El-Baz and Robinson (1997) and Robinson et al. (2007). Dual band images use combinations of two bands from Landsat for the north-western desert in Egypt ( fig. 7) thereby enhancing subtle features in the data that would not be imaged by a single band alone. The ratio of the infrared bands 7 and 4 ( fig. 7.a) highlights for example clay minerals which fill karst holes (red and cyan) in an otherwise homogeneous limestone plateau (dark blue). The band difference of the very near infrared band 4 and the green band 2 for the same area highlights difference in lithology between the pure limestone (yellow to red) and the more sandy cover (blue tones) towards the top of the image ( fig. 7.b). Multi-band images use three or more bands combined in continuous colour or red-greenblue (RGB) images. RGB images provide significantly more shades than single or dual band images: for 8 bit imagery an arithmetic combination of 3 bands provides 256 shades whereas an RGB image offers 16.8 million colors (Guo et al., 2008).

Integration of satellite imagery and geology
In this section we describe techniques to extract geological information from satellite imagery and how to integrate this information with geological data. Satellite imagery is interpreted for surface topography and lithology as well as for surface and subsurface structure with the goal of generating geological models for the near-surface and subsurface Insley, 2007, andLaake et al., 2008). The techniques are illustrated through a series of case histories starting with simple layer-cake geology. www.intechopen.com

Layer-cake geology (Qattara Depression, Egypt)
The surface north of the Qattara depression in Egypt is dominated by flat layering of hard and soft rocks at the surface ( fig. 9). The raw DEM ( fig. 9.a) shows a platform (yellow), which is located between rough hills (brown tones) and a sharp escarpment towards the Qattara Depression (green to blue). The terrain classification map ( fig. 10.c) allocates the different elevations to three classes. The lithological analysis of the multi-spectral image ( fig. 9.b) allows a clear separation into rock types : two types of limestone (blue tones), two types of sandstone (yellow to orange) and evaporites of the sabkha at the bottom of the depression (cyan). Ideally the surface lithology interpretation is validated in the field (see fig. 10) using GPS tracked lithological analysis. The combination of terrain classification and lithology suggests that the plane represents the top of the sandstone formation, which continues also below the limestone layers, which form the higher ground. We will use this geological model to estimate statics for seismic data processing in section 5.

Anticline (Awali, Bahrain)
The outline of the island of Bahrain is determined by the topographical and lithological structure of the Awali anticline ( fig. 11). We use satellite imagery from the ASTER sensor for the discrimination of clastic and carbonatic rocks (Laake et al., 2006). The continuous colour image from short wave infrared and visible bands (ASTER 631 RGB in fig. 11.a) allows discriminating the coastal farmland in the north (green) from the carbonates of the anticline (purple tints) and coastal sabkha (cyan). The structure of the anticline can be delineated using the difference of visible and short wave infrared data ( fig.  11.b), which can be traced even in the built-up area of Manama city in the north. The outer contour of the anticline appears highlighted in the west through the strong signature of the coastal sabkha (dark red). Draping the continuous colour image over the vertically exaggerated DEM generates a strong structural impression, which is useful for structural interpretation ( fig. 11.c).

Mapping of basins from gravity and radar data (Illizi Basin, Algeria)
The mapping of basins from satellite imagery is the only remote sensing technique which infers deep geological structures. Bouguer gravity anomalies inferred from satellite radar data give an indication of the thickness of the sedimentary cover above the basement. However, this interpretation requires additional information for example from magnetic data to constrain the model. Following the concept of geomorphology, which correlates surface and subsurface geology and litho-structure, surface structural lineaments can be used to infer subsurface structures. In figure 12 we show structural maps from satellite imagery for the Illizi Basin

Mapping of fracture zones below sand cover (Gilf Kebir, Egypt)
When sand dunes cover fracture zones in the near-surface, satellite radar data can map ancient river courses, from which fracture related weak zones in the near-surface can be derived ( fig. 13). We start with the outcrops of faults in the hard rock surrounding the sand dune field mapping straight valleys in the hard sand stone. The topography of the study area ( fig. 13.a) is composed of the flat Gilf Kebir plateau (brown tones) with gentle slopes (yellow) and a large plane (green to white). This corresponds to a surface lithology ( fig. 13.b) of hard sandstones (greenish-brown) and basement rocks (dark brown) as well as belts of sand dunes (white), which are locally discoloured by hematitic iron (purple). The shape of the escarpment and the valleys intersecting the Gilf Kebir plateau reveal fairly straight fault lines (black lines). Below sand neither topographical nor optical satellite imagery can reveal buried fracture zones. Therefore we use satellite radar which can penetrate dry sand for up to 20 m to map the clay contained in buried paleo-river beds. The radar data ( fig. 13.c) do not only map the fracture zones in the sandstone of the plateau (brown lineaments), it also shows W-E and SW-NE trends in the paleo-river courses (low radar intensities in blue) which continue from the lineaments in the rock outcrops. The overlay of the paleo-river courses from radar data on the geomorphology delineates the outcropping fracture zones across all terrains ( fig. 13.d).

Mapping of glacial moraine structures (Pechora Basin, Russia)
The last case study concerns the mapping of post-glacial structures in the Pechora basin in northern Russia using vegetation and water features as indicators; the lithology is not exposed in the study area. The structures studied comprise different types of moraines as well as drainage features in front of the glaciers (Laake 2009, Astakhov et al. 1999. We distinguish terminal and lateral moraines, which are composed of gravel and glacial till, from ground moraines, which are characterized by undulating terrain. Moraine ridges are often drier than the surrounding terrain because the elevated gravels cannot support a shallow water table. In contrast to this, ground moraines deposit more glacial till, which provides the basis for a shallow water table with very wet terrain and lakes. Significant differences in the ground moisture attract different plant species, which can be distinguished by their different response in the very near infrared and visible bands of satellite imagery. The topography allows delineating of the lateral and terminal moraines as long as they are still elevated above the surrounding plane ( fig. 14.a). Rendering the geomorphological map on the DEM in 3D improves the detection of these moraines. The location of lakes, which is indicative of ground moraines, is obtained from the landuse analysis of short wave infrared and visible green data. (fig. 14.c). The combination of all maps yields a clear delineation of the glacial moraines particularly when rendered in 3D ( fig. 14.d). In this case four glacial stages can be interpreted (numbers in fig. 14) : an initial stage where the entire area was covered by a thick ice shield (1), which deposited ground moraines over the entire study area. The second (2) and the third (3) phase comprise two distinguishable glaciers, where the glacier from phase two created the partly broken moraine wall in the west. The glacier in phase three covered the centre of the study area thereby partly levelling the lateral moraines from phase two. After melting, this glacier left the huge Lake Komi between the lateral and terminal moraines, which forms an extensive swamp today. The final glacier advance (4) covered only the northern third of the study area, leaving an irregular line of terminal moraines and extensive planes of ground moraines behind.

Integration of satellite imagery and geophysics
The geological information extracted from satellite imagery can be used to build geological models from basin scale (several 10 5 sqkm) to survey scale (few 10 3 sqkm) well before any geophysical data are acquired on the ground. These models can be used for the estimation of logistic risks for personnel and vehicles as well as for the estimation of the quality of the acquired geophysical data (Laake and Insley (2004aand b, Laake and Cutts 2007, Coulson et al. 2009). In turn, geophysical data can be employed to calibrate geophysical characteristics of near-surface layers inferred from satellite imagery. We distinguish the following types of geophysical surveys : 1. Frontier exploration aiming at identification of new basins in large unexplored areas.
Satellite imagery can assist defining the outline of potential basins and the definition of scouting surveys. 2. Structural imaging focuses on potential structures once the outline and character of a basin has been identified. Satellite imagery can support the design and logistic planning of geophysical surveys on the ground and can provide estimates for the quality of the acquired data before the data are acquired. 3. Reservoir characterization targets the most comprehensive study of the subsurface geological structure and fluids for individual reservoirs and therefore requires the best data quality. Satellite imagery can provide detailed models of the surface and nearsurface which provide input to data quality estimation before and during acquisition. For data processing satellite imagery can supply input to processes that correct for noise related to near-surface properties.

Fig. 15. Workflow for application of satellite imagery to geophysical methods
An outline of the information satellite imagery can provide for geophysical methods and the parameters studied is shown in figure 15. This section focuses on the aspects of survey design and data quality estimation. The design of geophysical surveys requires an understanding of the logistics and data quality aspects of the target area to correctly estimate the effort required to provide the desired quality of the final subsurface geological product. Satellite imagery can provide the information about the surface and near-surface before the start of the data acquisition (Cutts and Laake, 2009a and b). The method uses satellite imagery to generate geomorphological and litho-structural models of the surface and near-surface as described in the previous section to derive logistic planning and data quality estimation maps. Figure 16 shows the result of the method for a survey in the Western Desert of Egypt. DEM and multi-spectral data provide topography ( fig. 16.a) and lithology class maps ( fig. 16.b). The planning of the logistics requires information about limitations for access and maneuver for personnel and vehicles. The terrain classification provides the locations and steepness of escarpments as well as the surface roughness related to hard limestone. The combination of these terrain attributes defines the logistics risk ( fig. 16.c). A major obstacle for data quality for this survey is the scattering of seismic waves at the numerous escarpments resulting from the different weathering of limestone and sandstone. Fig. 16.d shows an estimate map for the scatter risk associated with escarpments.

Planning of safe operation in sand dunes
In high sand dunes, both logistics for wheeled vehicles and high absorption for seismic waves may impact the survey design severely (Laake and Insley, 2004a). The analysis of the topography can assist in outlining the dunes and characterizing their shape ( fig. 18). Digital elevation models derived from radar or stereo optical data (ASTER in our case) provide the slope index, which can be represented as an index for safe operation. In seismic operation surface gradients above 15 degrees are considered unsafe for operation, whereas in areas with 10 to 15 degrees slope an inspection of the terrain would be advised. Figure 18 shows a composite image of seismic line crossing a dune field. The visible band ASTER image ( fig.  18.b) gives a photographic impression of the surface, whereas the DEM ( fig. 18.c) provides an impression of the elevations. The slope risk map ( fig. 18.d) is obtained from the gradient of the DEM.

Prediction of accessibility and data quality in sabkha
In arid coastal areas often sabkha called salt flats pose a severe threat to personnel and equipment (Cutts and Laake 2009b). The salt flats may be subject to seasonal inundation, which may change their accessibility significantly. Analysis of the multi-spectral satellite Landsat imagery can allow the detection of halite minerals found in the surface crusts of sabkha. Thermal infrared imagery provides information about the different thermal properties of wet and dry sabkha. The visible band Landsat image in figure 19.b gives an indication of the location of the wet sabkha. The risk that the sabkha surface would not support heavy vehicles (see fig. 19.a) is directly correlated with presence of sabkha and can be derived from a sabkha detection map ( fig. 19.c). In addition to its impact on the logistics sabkha also affects the vibrator data quality. High distortion of the vibrator signal, an undesirable data property, is directly related to the presence of sabkha ( fig. 19.d).  Cutts and Laake 2008] www.intechopen.com

Estimation of the vibrator baseplate-to-ground coupling
Hard rock terrain affects the coupling of the vibrator baseplate to the ground (Laake and Tewekesbury 2005). Good ground coupling is achieved when the entire baseplate is in contact with the ground for example on gravel ( fig. 20.c). If the baseplate rests on a big boulder, the coupling surface from the baseplate to the ground may be reduced to a very small area resulting in substantial distortion of the transmitted signal and in severe damage to the baseplate ( fig. 20.a).
In this case the so-called point loading risk is correlated with hard limestone. The limestone prediction map draped over the DEM is shown in figure 20.b. But even for gravel the ground coupling may be compromised if the force level of the vibrator shaker and the frequency exceed the levels at which the ground supports the weight of the vibrator. The time signal of the baseplate amplitude ( fig. 20 bottom) shows the increasing frequency of the shaker force with time along with measured signal distortion. At three moments of the sweep photos of the shaker were taken. As long as the coupling of the baseplate to the ground is perfect the signal distortion is low ( fig. 20.d and e). When the ground surface breaks dust is whirled up by the baseplate motion and the distortion goes up ( fig. 20.f). Usually frequencies higher than the frequency at which the baseplate breaks through the surface show high phase distortion and deteriorate the data quality. The high distortion for very low frequencies (corresponding to fig. 20.d) is associated with motions of the entire vibrator and are not considered here. Soft terrain may have an impact on both the source coupling as well as on the receiver signal output (Laake 2005a). Due to higher absorption in soft sand, the signal level on sandy surface ( fig. 21.b and d) is generally lower than on hard gravel plain ( fig. 21.a and c). Satellite imagery can discriminate gravel with shallow evaporite pans on the higher ground from sandy terrain on the lower ground ( fig. 21.e). The quality of the source signal also correlates with the terrain character : signals from the sandy terrain show higher distortion ( fig. 21.f) and lower ground stiffness ( fig. 21.g) than signals from the gravel plain. Interpretation of satellite imagery can provide estimates for the quality of seismic signals obtained from the terrain characterization. Fig. 21. Correlation of source and receiver data quality with terrain, Kuwait: a -gravel plain, b -undulating sandy surface, c, d -shot record displays, e -surface lithology, f -total harmonic distortion, g -ground stiffness [after Laake 2005a]

Improvement of static corrections
Near-surface geological models can be generated from the interpretation of satellite imagery for geomorphology and lithology (Laake et al. 2008). Using standardized seismic velocities the geological model is converted into an elastic model, from which estimates for statics corrections can be computed (Laake and Zaghloul 2009). The analysis of the DEM delivers the topographic classification into three layers (see fig. 22): the platform (sandstone), which occupies most of the study area, as well as the lower platform of the Qattara Depression and the rough higher layer (limestone). The interpretation of the shortwave infrared data reveals t h a t t h e p l a t f o r m i s c o m p o s e d o f s a n d s t o n e, whereas the higher ground consists of limestone ( fig. 22 top). When choosing standardized P-wave velocities of 2200 m/s for sandstone and 3300 m/s for limestone we can compute estimated statics values ( fig. 22  bottom). The estimated values are compared against refraction static corrections from a 3D seismic survey ( fig. 23).

Consistent deep geological models from data integration
In the final section we will show the benefits of merging interpreted satellite imagery with geological and geophysical data on the western shore of the Gulf of Suez in Egypt by demonstrating the correlation of the geomorphology with the subsurface litho-structure . The processing of multi-spectral imagery from Landsat for lithology and drainage reveals fault structures which are masked by recent wadi deposits (Laake 2010, Laake et al. 2011. The approach follows the idea that tectonic movements deviate the courses of wadis in a characteristic way.

Mapping of fault outcrops at surface from satellite imagery
The geological setting of the study area is determined by rift faulting in NW -SE direction and transform faults in the perpendicular direction (Darwish andEl-Araby 1993, Alsharhan andSalah 1993). The natural colour Landsat 321 RGB image ( fig. 24.a) shows the almost featureless beige gravel plain between the Red Sea Mountains in the west and the Gulf of www.intechopen.com Suez in the East. The only visible features are SW -NE running wadis delineated through their light sediments. The inverted natural colour image directs the eye to the wadi courses, which reveal several anomalously straight sections ( fig. 24.b). When we use all data from Landsat and spectrally enhance the resulting image boundaries along the main tectonic directions are imaged ( fig. 24.c): parallel to the coast NW -SE trending linear anomalies point to outcropping parallel fracture zones (dashed lines). The dominating wadi in the southern part of the study area is confined between two straight SW -NE trending lineaments which point to faults running perpendicular to the rift orientation (dotted lines). This map also reveals differences in the mineral composition of the bedrock of the Red Sea Mountain granites. The erosion fans from these two granites may actually be used as tracers to highlight the structural boundaries on the gravel plain. The lineaments are also highlighted in the drainage map, which results from the ratio of the thermal and pan-chromatic bands ( fig. 24.d).

Mapping of faults in near-surface from shallow geophysical data
In the study area, the fault outcrop pattern mapped from satellite imagery is calibrated with surface features and shallow seismic data from a 3D seismic survey. The only tree in the study area has been found at the intersection of two fault outcrops ( fig. 25.a) indicated by the arrow in the lithology map ( fig. 25.b). The surface structure becomes evident when draping the litho-structural map over the vertically exaggerated DEM ( fig. 25.c). Arrows indicate the outcrop of faults at the surface. The correlation with faults in the near-surface is achieved by extracting weak zones of low surface wave velocity from seismic 3D data ( fig.  25.d). Surface wave velocity analysis provides an iso-velocity horizon which corresponds to a shallow lithologic horizon. This horizon shows very low velocities along the rift parallel fault outcrop zones mapped by remote sensing. In the direction orthogonal to the rift faults a high-velocity structure below the wadi is revealed but no sharp boundaries. This may be due to the transform character of these faults which does not provide an impedance contrast that could be detected from the seismic.

Mapping of faults in the subsurface using surface tamplates
The characteristics of drainage patterns delineated from satellite imagery may also be used to generate shape templates for the detection and extraction of similar structures in subsurface seismic data ( fig. 26). We use the drainage map to define the template for a wadi ( fig. 26.a) consisting of the braided stream and the fan delta. In our case the braided stream is confined by the SW -NE trending faults perpendicular to the rift orientation, whereas the fan delta part shows so little elevation change that the fan delta crosses the boundary faults ( fig. 26.b).
The shape template is used in the processing of the shallow seismic data to search for the correct seismic attribute suited to map the faults which are hidden in most other seismic attributes. In our case the instantaneous frequency attributes were identified for mapping the SE -NE faults because this attribute highlights locally high spectral amplitudes which were interpreted as resonances within the palao-wadi structures ( fig. 26.c). Once the shallow paleo-wadi has been identified the same methodology is applied to the deeper data; a total of four paleo-wadis spanning the entire rift history were mapped ( fig. 26.d).

Mapping of shallow drilling risk related to faults
Finally the surface and shallow subsurface mapping can be merged into a shallow structural map with indications of shallow drilling risks related to outcropping fault zones ( fig. 27). The map basis is a landuse map (Landsat 742 RGB) onto which the near-surface topography is projected. The fault lines delineate areas of shallow drilling risk related to fault induced weak zones, which might lead to collapsing boreholes and/or loss of circulation. www.intechopen.com

Conclusion
Satellite imagery enables the investigation of the properties of the earth surface in remote areas and over large areas through the mapping of physical properties from satellite based sensors. Optical imagery delivers information about land-use, water features and surface rocks. Microwave radar imagery maps surface roughness when back-scattered at hard surfaces and paleo-drainage when penetrating dry sand cover. Microwave radar distance measurements can be converted into surface topography maps or digital elevation models. The joint interpretation of radar elevation measurements and the geoid shape delivers satellite based gravity anomaly maps, which reflect the thickness of the sedimentary layers above the crystalline basement. The interpretation of satellite imagery can assist in the planning of surface seismic surveys through assessment of logistic and data quality risks early in the planning specifically when exploring in frontier areas. Interpretation of satellite imagery can provide estimates of the source and receiver data quality and static corrections. Consistent geological models can be generated from the interpretation of satellite imagery for geomorphological and lithostructural properties integrated with geological and geophysical data. Fig. 28 and table 1 give an overview of the geological features detectable from satellite imagery and their impact on seismic data quality.