The Importance of Accurate Altimetry in AEM Surveys for Land Management Ross Brodie Geoscience Australia ross.c.brodie@ga.gov.au Richard Lane Geoscience Australia richard.lane@ga.gov.au SUMMARY Airborne electromagnetic (AEM) systems are increasingly being used for mapping conductivity in areas susceptible to secondary salinity, with particular attention on near-surface predictions (ie those in the top 5 or 10 metres). Since measured AEM response is strongly dependent on the height of both the transmitter loop and receiver coil above conductive material, errors in measurements of terrain clearance translate directly into significant errors in predicted near-surface conductivity. Radar altimetry has been the standard in airborne geophysical systems for measuring terrain clearance. In areas of agricultural activity significant artifacts up to five metres in magnitude can be present. One class of error, related to surface roughness and soil moisture levels in ploughed paddocks and hence termed the paddock effect, results in overestimation of terrain clearance. A second class of error, related to dense vegetation and hence termed the canopy effect, results in underestimation of terrain clearance. A survey example where terrain clearance was measured using both a radar and a laser altimeter illustrates the consequences of the paddock and canopy effects on shallow conductivity predictions. The survey example shows that the combination of the dependence of AEM response on terrain clearance and systematic radar altimeter artefacts spatially coincident with areas of differing land-use may falsely imply that land-use practices are the controlling influence on conductivity variations in the near surface. A laser altimeter is recommended for AEM applications since this device is immune to the paddock effect. Careful processing is still required to minimise canopy effects. Key words: AEM, radar altimeter, laser altimeter, conductivity, land management. INTRODUCTION Airborne geophysical survey aircraft carry a radar altimeter as a ground-proximity-warning device. The data are usually recorded digitally along with other survey information. These data are generally used to construct a digital elevation model (DEM) by subtracting the aircraft terrain clearance from the aircraft altitude measured with a barometric altimeter or satellite navigation system. Various corrections for antenna locations on the aircraft, aircraft attitude and the geoid to ellipsoid separation (in the case of satellite positioning) complete the processing. In the case of airborne electromagnetic (AEM) surveying, terrain clearance measurements are used to define part of the system geometry that is an input to the process of converting the electromagnetic measurements to conductivity. Shallow conductivity predictions are extremely sensitive to the terrain clearance of the transmitter loop and receiver coil elements of the system. The two different radar altimeters discussed in this paper operate at around 4.3 GHz, corresponding to a wavelength of around 7 cm. The manufacturers typically quote an accuracy for the analog output of 0.6 m plus 2-3 % of the terrain clearance, which would lead to an overall accuracy of 2-4 m in airborne geophysical applications. Spatial filtering of closely spaced readings along flight lines reduces random noise, resulting in an improvement in accuracy of the processed terrain clearance measurements to around 1-2 m. Laser ranging devices, either profilometers or scanners, are increasingly being installed on survey aircraft (Stone and Simsky, 2001). Rapid sampling of these devices enable ground returns to be discriminated from returns from above ground features such as vegetation or buildings by extracting the maximum range within a window of the data along a flight line. Canopy effects occur when the radar altimeter signal is reflected off the canopy of vegetation rather than the ground beneath. These effects have been recognised in radar altimeter measurements in geophysical survey applications (Beamish, 2001; Markham and Morris, 2002). Beamish (2001) notes the effects on conductivity predictions. A second class of effect, termed the paddock effect, mentioned in Richardson (2000) is described here. A possible reason for the paddock effect is that surface roughness of the order of the wavelength of the radar signals affects the shape and amplitude of the radar returns. A reduction in amplitude, or shift in the timing, of the peak return would produce an increase in the measured travel time of the pulse and result in overestimation of the terrain clearance. A notable source of surface roughness at the relevant scale is ploughing of paddocks prior to cropping. Ploughed paddocks give rise to artifacts that are confined to specific paddocks, hence the name used to describe this effect. Roughness is present to varying degrees across all landscapes, but the gradational nature of the majority of these roughness variations limits the visual impact of the effect on a DEM and the ability to outline the affected areas.
Accurate Altimetry in AEM Surveys METHOD AND RESULTS Artefacts in the terrain clearance profiles are often difficult to recognise. The presence of these artifacts is often much clearer in DEM images. The DEM shown in Figure 1 definitively illustrates paddock effects. These data, from the Dirranbandi irrigation district in Southern Queensland, were acquired as part of a magnetic and radiometric survey in April-May 2001 at 60 metre nominal terrain clearance and 100 m line spacing. The artefacts are evident as bluish polygonal features, and are of the order of 3 to 5 m, which corresponds to errors in the radar altimetry of +3 to +5 m. Incidentally, the red polygonal features that might also be interpreted to be artefacts are actually large above-ground water storage dams and are thus not artefacts. Figure 2. DEM produced from the AEM survey flown with greater terrain clearance and wider line spacing than the survey referred to in Figure 1. Terrain clearance was measured using a radar altimeter. A profiling (as opposed to scanning) laser altimeter recorded terrain clearance data in parallel with the radar altimeter on the AEM survey. The laser altimeter, installed in this case on an experimental basis, was of early 1990's vintage. It was set to record at 0.2 second (~15m) intervals and suffered from intermittent null returns and some noise in the form of mainly negative directed spikes. Figure 1. DEM produced from a low-level magnetic and radiometric survey. Terrain clearance was measured using a radar altimeter. A survey employing the AEM system referred to by Lane et. al. (2000) was conducted over the same area from June to August of 2001. The AEM system utilised a different make of radar altimeter. The survey was flown at a larger nominal terrain clearance of 120 m and with a broader line spacing of 250 m. The DEM generated from the AEM survey (Figure 2) contains the same artefacts as the magnetic and radiometric survey DEM, although they are not as clearly resolved owing mainly to the wider line spacing and perhaps the greater terrain clearance. Local maxima and minima threshold rejection filters were iteratively applied to the laser altimeter data over shrinking width windows such that any terrain clearance variation in excess of that physically achievable by the aircraft over smooth ground was removed. This filtering removed the null returns and the majority of the noise and vegetation returns. The magnitude of errors introduced by variations in aircraft attitude was investigated but found not to be significant at the metre scale. The laser altimeter data proved to be immune to the paddock effect within the precision of the measurement system. Consequently the DEM generated through use of the filtered laser altimeter terrain clearance data is free of paddock effects (Figure 3). The canopy effect is also substantially reduced in the laser altimeter DEM. An image of the difference between the radar and laser altimeter measurements (Figure 4) highlights the canopy effects (red colour) and paddock effects (blue colour). ASEG 16th Geophysical Conference and Exhibition, February 2003, Adelaide.
Accurate Altimetry in AEM Surveys This raises the issue as to whether paddock effects are restricted to a small number of aircraft installations, and to consider why they have not been noted previously to any significant degree. Results from a number of surveys have been reviewed, and paddock effects have been observed in surveys involving different aircraft and manufacturers radar altimeter instruments. They are however most evident when the terrain clearance and line spacing are reduced. The trend in airborne survey specifications toward smaller line spacing and terrain clearance, the increase in use of AEM for shallow conductivity mapping in land management applications over areas of cropping and the direct use of terrain clearance in conductivity transformations all combine to explain why paddock effects have become increasingly evident and important in recent years. Figure 3. DEM produced from the AEM survey referred to in Figure 2. In this instance, terrain clearance was measured using a laser altimeter. Figure 4. Difference between laser and radar altimeter measurements for the AEM survey. A portion of survey line 30670 is shown together with selected canopy effect features (C1, C2) and paddock effect features (P1, P2). The electromagnetic measurements were transformed to conductivity using program EMFlow (Macnae et al., 1998), firstly using the radar altimeter terrain clearance measurements and secondly using the laser altimeter measurements. An image of the difference in the shallow conductivity predictions between the two datasets is shown in Figure 5. In general, conductivity is overestimated in the case of paddock effects and underestimated in the case of canopy effects. The errors introduced by these terrain clearance artifacts diminish with depth. This is illustrated in profile form for line 30670 in Figures 7 and 8. Figure 5. Difference in conductivity predictions using laser and radar altimeter measurements for the interval 0 to 4 metres below surface. The conductivity image for 0-4 metres below surface based on laser altimeter measurements (Figure 6) is considered to be as free from artifacts in terrain clearance measurements as is practical in this instance. There are conductivity anomalies shown in Figure 6 partially coincident with paddock effects. ASEG 16th Geophysical Conference and Exhibition, February 2003, Adelaide.
Accurate Altimetry in AEM Surveys We are not claiming that these are not real. In fact it is reasonable to hypothesise that where land is preferentially cultivated on the basis of a particular soil type there is likely to be some correspondence between variation in conductivity and land use. CONCLUSIONS In general, conductivity is overestimated in the case of paddock effects and underestimated in the case of canopy effects. The errors introduced by these terrain clearance artifacts diminish in significance with depth. The effects are much more evident in conductive terrain than resistive terrain. The magnitude and significance of the effects will vary with the characteristics of the AEM system used. The combination of the dependence of AEM response on terrain clearance and systematic radar altimeter artefacts spatially coincident with areas of differing land-use may falsely imply that land-use practices are the controlling influence on conductivity variations in the near surface. Laser altimeter devices are immune to paddock effects and are therefore recommended for airborne geophysical surveys where shallow conductivity mapping or derived digital elevation models are considered important. Rapid sampling laser altimeters would be advantageous for removal of the canopy effect. ACKNOWLEDGEMENTS This work was carried out under the auspices of the Nation Action Plan for Salinity and Water Quality supported by State and Commonwealth governments. We would like to acknowledge the Queensland Department of Natural Resources and Mines and the Bureau of Rural Sciences for their support for this work to be undertaken. Figure 6. Conductivity predictions using laser altimeter measurements for the interval 0 to 4 metres below surface The profiles in Figure 7 also show that these effects are of greater magnitude and significance in conductive terrain than in resistive terrain. The canopy effect can be accommodated in the conductivity transform by addition of a false resistive layer at surface. In resistive terrain, this will be barely noticeable, but will result in a large artefact in areas with conductive near-surface conditions. The paddock effect is more complex. It can result in a situation of incompatibility between response, terrain clearance and conductivity. This occurs when the observed response exceeds the inductive limit determined using the erroneously large terrain clearance. In this case, the conductivity transformation increases the near-surface conductivity above the true value in an attempt to accommodate both the observed response and the erroneously large terrain clearance. After accounting for the AEM system s characteristics (noise levels, sensitivity to near surface conductivity, geometry and terrain clearance) this is more likely to be significant, in conductive areas. The paddock and canopy effects and the resultant errors in the derived conductivity information shown here are spatially associated with land use. It is therefore possible that interpretation of the erroneous conductivity data may result in incorrect linkages between land use and conductivity being made. This paper is published with the permission of the CEO, Geoscience Australia. REFERENCES Beamish, D., 2001, The canopy effect in airborne EM: Paper ELEM13, Proceedings of the 7 th Meeting Environmental and Engineering Geophysics, Birmingham, England, September 2 6, 2001, European Section of the Environmental and Engineering Geophysical Society (EEGS). Lane, R., Green, A., Golding, C., Owers, M., Pik, P., Plunkett, C., Sattel, D., and Thorn, B., 2000, An example of 3D conductivity mapping using the TEMPEST airborne electromagnetic system:exploration Geophysics, 31, 162-172. Markham, K.J., and Morris, W.A., 2002, Creating and correcting a digital terrain elevation model using radar altimetry and GPS data acquired from aeromagnetic surveys: Poster P7, Proceedings of the Symposium on the Application of Geophysics to Environmental and Engineering Problems (SAGEEP), Las Vegas, Nevada, USA, February 10-14, 2002, Environmental and Engineering Geophysical Society (EEGS). Macnae, J.C., King, A., Stolz, N., Osmakoff, A. and Blaha, A., 1998, Fast AEM data processing and inversion: Exploration Geophysics, 29, 163-169.
Accurate Altimetry in AEM Surveys Richardson, L.M. 2000, Errors in digital elevation models derived from airborne geophysical data: Australian Geological Survey Organisation Record 2000/37. Stone, P.M., and Simsky, A., 2001, Constructing high resolution DEMs from airborne laser scanner data: Extended Abstract, ASEG Conference, Brisbane, 2001. Figure 7. Profiles for line 30670. Features C1, P1, P2 and C2 are described in the text and shown on Figures 4 and 5. (a) DEM calculated radar (thin line) and laser (thick line) altimeter measurements. (b) The difference between radar and laser altimeter measurements. (c) Conductivity predictions for the interval 0 to 4 metres below surface, based on radar (thin line) and laser (thick line) altimeter measurements. (d) The difference between the conductivity measurements in (c) as a percentage of the conductivity calculated using laser altimeter measurements. A strong negative correlation is observed between the difference in altimeter readings and the errors in conductivity predictions. Figure 8. Profiles for line 30670 as per Figure 7 except that the conductivity values in (d) refer to the interval 40 to 44 metres below surface. The impact of radar altimeter artifacts is significantly reduced from that observed at shallower depths (Figure 7).