Sinclair Knight Merz Pty Ltd
PO Box H615 Perth, Western Australia
Curtin University of Technology, Bentley Western Australia
Phone number, 618 92689672; Fax number 618 9268 4598
Email address Gstreet@skm.com.au
Stress of European style agriculture on Australian landscapes is causing long-term degradation problems which are difficult to reverse. These problems coupled with low commodity prices and narrow profitability margins underlie the need for better information for land and catchment management.
For farmers to increase productivity and be sustainable both from an ecological and economic sense they need to understand landscapes. Geophysics measures within the regolith and thus reflects sub-surface processes that may impact on land management.
Geophysical systems have been adapted, altered or new systems designed specifically for the deeply weathered Australian landscapes. Development of new-generation airborne geophysical systems has been funded because of the potential application in land management. The land management market needs to ensure that these systems are appropriate for their needs because the designers and operators experience is in mineral exploration.
In land management the interpretation approach is to use all the data and the use of many complimentary datasets. A ‘pixel by pixel’ interpretation is appropriate for datasets which may only have a few thousand pixels per farm. In the traditional mineral exploration market around 90% of the data may be discarded in the first pass of interpretation in the rush to find mineralisation targets. The different approaches are important in terms of data resolution and quality. Advances in GIS allow easier data manipulation and information extraction. The information and knowledge from interpretation of geophysical data will allow better design and testing of land management decisions.
There are currently around two million hectares of salt affected land in Western Australia. Significantly larger areas with already high watertables are considered at risk of salinisation sometime in the future. While Western Australia may have the largest problem in terms of area, there are other states in Australia where the economic impact will probably be greater. In other parts of the world sociological impacts of increasing salinity may be catastrophic.
In Australia the clearing of perennial native vegetation and its replacement with annual agricultural species has resulted in greater accessions of recharge water to the groundwater. Rises in groundwater levels of between 10cm and 50 cm per year are not uncommon after clearing. Rainfall has very low concentrations of salt – around 50 mg/L at the coast dropping to around 15mg/L in inland areas. Over many thousands of years in the poorly drained landscape the native vegetation has used the fresh water and left the salt behind stored in the regolith – the weathered layers above basement rock. Rising groundwater brings this salt to the surface where the combined effects of waterlogging and salinity kill the remaining vegetation. Once this process has occurred it is difficult to reverse.
In a three-dimensional sense most Australian landscapes in the inland agricultural areas are saline. The salt may be at surface or at depth with some potential to rise to the surface in rising watertables. Most farmers manage the soils on their land in great detail. This management usually only extends to the top 50cm to 2 m. However the surface soils are only a thin veneer and long term management depends on understanding the processes going on under the surface. In particular we need to understand the processes we believe have the potential to change surface conditions in the short (few years) to medium (few decades) time frames.
Geospatial data sets such as airphotos have long been used in order to extend point source information over large areas. Thus soil mapping and geology extensively used airphoto interpretation to join up points of ground observation. Airphotos are an accepted tool in land management and satellite images have taken over some of this role. Airphotos reflect what we can normally see on the ground but from a different angle and with a more encompassing view. They are thus easy to accept and easy to interpret. Satellite images can almost replicate what we see in airphotos and thus it is a logical step to replace airphotos with images which can take in a broader view. However satellite images can be separated into wavelength bands to reveal features not immediately obvious to the eye but actually contained in airphotos. As we get into infra-red bands things become more difficult for the untrained eye. New multispectral systems go further in discrimination of various wavelengths.
Airborne and space radar systems have also been used. In their normal role they reflect the ground surface and are easy to interpret. Some of the signal penetrates the ground and information about ground conditions can be extracted. In this role the radar is harder to interpret and understanding of electromagnetic theory becomes important.
Geophysics goes further than remote sensing. Whereas remote sensed techniques measure the reflection of radiation from the top one micron of the surface, geophysics measures physical variations in the ground beneath the surface. These are not usually variations we can see visually and are thus harder to understand. They can be interpreted however in terms of processes that helped to form the landscape and that are still active today. Geophysical techniques although firstly applied in agriculture for salinity can be used to improve whole farm productivity in the short to long term. This paper concentrates on the environmental use of airborne geophysics in particular. Airborne geophysics can be used to cover vast areas quickly and economically and thus it has a place in land management that is presently curtailed by the lack of knowledge, exposure and expertise of land managers. Airborne geophysics will be a tool that efficient farmers will come to use widely if not routinely in the future.
Collection of Airborne Geophysics
Airborne geophysical data is collected by low flying aircraft usually at a level of between 20 metres and 150 metres above the ground. They fly in what is termed a loose drape formation attempting to follow the ground contours as closely as possible. By necessity these aircraft are fairly slow flying although they must maintain a reasonable speed to maintain platform stability and be economic. They generally operate at speeds of close to 120 knots or around 60 metres per second.
Typical aircraft utilised today for magnetic and radiometric work are small single or twin engine commuter aircraft and more recently agricultural crop duster type. Electromagnetic aircraft are usually much larger to accommodate the transmitter loop which is strung around the nose, tail and wingtips. All these aircraft are modified to some degree to enable good quality data to be collected.
Figure 1 Fletcher crop duster type aircraft operating at very low level. Note magnetometer sensor is located in a ‘stinger’ at tail of the aircraft and radiometric crystal sensors in the belly of the aircraft (photo courtesy UTS Geophysics).
Data is collected along grid lines which can be as close as 25 metres apart or up to 3 kilometres. In choice of line spacing it is essential to first define the spatial variability that is likely to be measured. For most catchment management applications a maximum line spacing of 200 metres is recommended. The latest radiometric surveys for soil mapping have been conducted at line spacings as close as 25 metres and 20 metre altitude using crop duster type aircraft.
Data is collected in computers on-board the plane. All data is linked to a position from GPS signals. The GPS is usually linked to satellite differential correction signals and therefore the accuracy of position is usually better than 1 metre in X and Y and around 2 metres in Z or vertical direction. There are currently around 12 airborne geophysical aircraft from four contracting companies operating in Australia.
Geologists use airborne magnetic surveys to map the variations in magnetic mineral content in the rocks and thus to map geology. Very few mineral exploration programs do not have magnetic surveys as an integral component. In salinity geological structures are often important in causing groundwater levels to rise. Engel at al (1987) and Street and Engel (1988) showed that lateritic weathering of rocks with low quartz content created vertical zones of low permeability in the regolith. These act as barriers to groundwater flow causing watertables to rise upslope. Rocks with low quartz content are often more magnetic than quartz rich rocks. They are therefore easy to map with modern magnetic methods.
Modern magnetometers measure the magnetic field every tenth of a second with newer system now capable of measuring every millisecond. Most instruments take a measurement of the total magnetic field every 5 to 6 metres along line.
Figure 1 Diagram showing measured magnetic field in Southern Hemisphere over a dyke which has higher magnetic mineral susceptibility than surrounding rocks.
Once data is collected it is processed to remove changes in the diurnal magnetic field and processed to produce contours or image maps of the total magnetic field. The Earth’s total magnetic field varies in intensity from around 90,000 nanoteslas near the magnetic poles to around 20,000 near the equator. In between these extremes the field over a magnetic body is asymmetrical with a low to the south and a high to the north of centre. As it is easier to interpret symmetrical anomalies the data is usually reduced to pole or transformed to give a shape that would be expected at polar regions. Further processing can be employed to enhance narrow features or edges that will be of interest to the interpreter or user of the data. The most common is a first vertical derivative filter.
Figure 2 Magnetic data image over 15,000 hectares of farmland near Narrogin in Western Australia. The data has been processed to enhance linear features such as dolerite dykes which act as barriers to groundwater flow.
Figure 3 Magnetic data interpreted in GIS to show dolerite dykes in green; faults in red; granite basement in pink and grey.
The data can then be interpreted to produce maps of solid geology (figure4). In an interpretation for land management purposes it is essential to do the work at a scale that is appropriate for farmers. In most of the deeply weathered Australian landscapes hydrogeological systems is highly related to the underlying geology. It is highly variable and complex and there can be two or more completely different hydrological systems operating within a paddock and significant changes can occur within less than 100 metres. Interpretation of magnetic data can be used to map the underlying geology and this can be related to the regolith. Scale of interpretation needs to be at the same level at which remedial measures will be applied.
Airborne radiometric systems measure the gamma radiation (or radioactivity) spectrum. The energy of each gamma ray can be measured and therefore a full spectrum of energy levels recreated by processing of the measurements. Individual peaks in energy levels can be related to the presence of natural radioactive isotopes of potassium (K), thorium (Th) and uranium (U). Thus individual peaks due to K, U and Th are separated and measured to give concentrations. K is usually between 0 and 3%, while Th and U are in parts per million. Images of K, T, U and total count can be made either as counts per second or as isotopic concentrations. In addition it is common to produce a ternary image whereby the values for K, Th and U are assigned to the red, green and blue guns on the computer screen.
The radiometric signal comes predominantly from the top 30cm up to 2m below the surface. Thus the relationships of these isotopes can be used to improve soil mapping. Heavy waterlogging will adsorb gamma rays and thus the best surveys are carried out in dry conditions or at least in times when there have been a few days without rain.
Large crystal volumes will trap more gamma rays and thus are essential in obtaining good data. Low altitude is also critical and the best data (figure 7) is obtained at lowest possible safe flying heights using crop duster type aircraft (figure 1).
Figure 1 Cartoon showing the configuration in radiometric surveys with inset showing a typical gamma ray spectrum. Note potassium peak (red) is the largest with the uranium peak (blue) usually the lowest.
Figure 2 (a) Terrain image, (b) potassium, (c) thorium and (d) uranium for an area in wheatbelt of Western Australia
Figure 3 Radiometric ternary image. Note that picture is ‘fuzzy’ because of wide line spacing 150 metres plus small crystal volume Data is even ‘fuzzier’ in northwest corner where line spacing was 300 metres. (Area around 6500 ha)
Figure 4 Improved image due to better processing and larger crystal volume over similar size area and geological setting to the area in figure 5 (Area 6500 ha).
The relationship between the isotopes and total elemental composition are fairly rigid and thus isotopic concentration of radiogenic K40 is closely related to total K. The elements K, U and Th behave differently under different conditions of acidity etc that are involved in soil formation. Thus the concentrations of radiometric isotopes are related to the processes of soils formation and to some degree to soil types.
Classification techniques can then be used to separate out areas with similar radiometric character. Elevation and slope data can assist in this classification. Other geospatial data including geophysics and remote sensed data can assist in separating classes.
Figure 5 Perspective image of radiometric data draped on geophysical survey derived DTM from an ultra high-resolution survey carried out over 7500 ha area near Wyalkatchem in Western Australia. The survey specifications were 25 metre line spacing and 25 metre flying height. Crystal volume was 32 litres. The number of gamma rays detected at this height makes sampling at half a second or every 30 metres. Soil relationships are easy to see in these data. Spikes are due to radar altimeter reflecting from trees and can be removed (image supplied by UTS Geophysics)
The concept of using airborne geophysics for land management in areas prone to salinity has been around for around 15 years. It was pioneered in Victoria and WA in late 1980s with analogue mineral exploration airborne electromagnetic (AEM) systems. The survey results using these early systems were poorly resolved partly due to the sampling of the system and partly due to the specifications of the surveys.
Cook and Kilty (1992) trialed a more detailed approach using a helicopter system in South Australia but the system used, plus the available software for inversion at that time were inadequate for resolving shallow conductivity changes.
Digital technology was introduced into AEM in the late 1980s. The author and others received a grant under the National Soil Conservation Program (NSCP) in 1987 to develop a low cost AEM system for salinity mapping. This lead to the optimisation of the QUESTEM mineral exploration system for salinity mapping.
Under the National Soil Conservation and National Landcare Programs QUESTEM surveys were conducted in the early 1990s at East Yornaning, WA, 1990; Kent River, WA, 1992; Carnamah, WA, 1992; Wanilla SA, 1993, Boscabel, WA, 1992; Cressy-Longford, Tasmania, 1993; Loddon-Campaspe, Vic, 1993; Jemalong-Wyldes Plains, NSW, 1993; Pittsworth, Queensland, 1993; and Esperance, WA, 1994.
A finding of these surveys was that better discrimination of conductivity depth relationships would probably assist in better understanding the relationships between surface soil salinity and the salt storage patterns reflected by the conductivity measured with AEM systems.
A new AEM system was needed that was optimised for conductivity changes in the top 20-50 metres of the ground. The SALTMAP system (Street and Roberts, 1995) was developed with improved signal sampling rates and design features that would enable better discrimination of conductivity depth relationships. This was trialed at Broomehill in 1995.
Further surveys were carried out to develop farm plans with farming groups in Trayning, Wyalkatchem, Nungarin, Moora and Towerrinning areas in WA in 1996 and 1997. The concepts behind the development of land management plans were developed from surveys done for farmers (see Anderson et al, 1995). As farmers are closer to the implementation of land management decisions and are business managers they require a technically sound approach (Nulsen et al, 1995).
In 1997 the Federal Government made available funds through the National Dryland Salinity Program in the National Airborne Geophysics Project to undertake “airborne electromagnetic, magnetic, gamma-ray spectrometric and digital elevation model surveys to assist in the planning for control of dryland salinity and to evaluate the applicability of airborne electromagnetic surveys for this purpose”.
Figure 1 Image of shallow regolith conductivity from first SALTMAP survey at Broomehill. High conductivities are in hot colours red to white and low conductivity in purple to blue. The highest conductivities are mostly but not always in the lower landscape.
Five catchments around Australia were surveyed and the results interpreted at Willaura, Victoria; Balfes Creek, Queensland; Toolibin Lake, WA; Liverpool Plains, NSW and Chapman Valley WA (Street et al, 1998, Pracilio et al, 1998, Street and Pracilio 2000, George et al 2001)
Prior to the NAGP it was already recognised that SALTMAP was inadequate for the application and that recent advances in electronic switching and computing enabled a much more powerful system to be built which had both a mineral and environmental application.
The TEMPEST system (Lane et al, 1998) was built a trial surveys conducted over the Toolibin Lake catchment (Lane and Pracilio, 2000) towards the end of the NAGP. The system is now being used in the Salinity Mapping and Management Support Project.
The extensive development work has resulted in a system that is applicable to both mineral exploration and environmental markets. The downside is that for environmental applications the cost has risen and the specifications are more suited for mineral exploration than shallow environment work.
Information Extraction for Land Management
Information extraction for land management is complex. Initially, the role for geophysicists is important because they understand the limitations of the data. However, once geophysicists set the boundaries, there is no reason why end users should not be able to successfully extract information from the data. Thus in land management the interpreters are environmental scientists who understand what information they want to extract from the data.
In using geophysics for land management the interpreter wants to know what each piece of data means. This is not an anomaly hunting exercise but a pixel by pixel analysis. Geographic Information Systems are invaluable in such data analysis.
The keys in extracting information from geophysical data are
- the land manager asking the right question;
- the geophysicist providing the right data in the right format; and,
- the interpreter (information extractor) using the right tools in the extraction process.
The people in the process all have to understand what information the land manager requires, the role for geophysics in supplying that information and the role for other data sources.
A suite of products derived from interpretation of airborne geophysical data could be:
- Salt Hazard maps incorporating an understanding of the hydrogeological causes of salinity at local scale
- Water Resource Target Maps
- Soil Maps
- Salt Hazards
Salt hazard mapping
Salt hazard mapping entails first interpreting the data suite in terms of the regolith aquifer processes, which are causing salt degradation in the survey area. These processes are then summarised as a set of salt hazard models based on three key factors: salt storage beneath the surface; an increase in recharge; and a change in transmissivity of regolith aquifers. Salt hazard models are used to create salt hazard maps that identify or predict each salt hazard in terms of its hydrogeological causes. The salt hazards are then used as part of the information accessed by the land management planner to develop a better approach to managing the landscape. Salt hazard mapping ensures that remediation techniques are targeted at addressing the causes of salinity rather than just the symptoms. In the first instances this work was done manually on light tables with multiples of maps. Now most of the information extraction can be done in GIS.
The salt hazard interpretation involves firstly an analysis of where the salt is within the landscape base don airborne electromagnetic surveys. This data is related to soil types from radiometric data and geology from magnetics to build models of the causes of saline groundwater discharges or rising groundwater seen in bore data and aerial photos. Drainage and topography also are incorporated and an assessment of likely severity carried out based on catchment size, groundwater levels, salt content of catchment and recharge potential of soils. Full manipulation of the multitude of datasets has only been possible with the introduction of GIS. Some of the salt hazard interpretation process can be automated in GIS (Anderson-Mayes, 2000)
Figure 1 salt hazard map for an area of 100,000 hectares in the West Australian wheatbelt showing local hazards (small areas) and regional hazards due to palaeochannels mapped with geophysics. The salt hazards have been interpreted down to paddock scale to guide farmer’s land management decisions.
Water Resource Target Maps
A land manager also want know likely sources of groundwater. Although the data used to answer this question is the same, the interpreter will use it in a different way and put different weighting on the information from different data sets.
Most of the water in areas with granite basement is contained in the regolith above the basement. In looking for water in the regolith electromagnetic data is used to map salt content and water quality. It can also give the thickness of the regolith aquifer. Radiometrics can be used to map the areas with highest recharge potential based on soil types. Other datasets such as magnetics and digital terrain models are used to locate the best target sites in the landscape. For example the interpreter may look for a site towards the lower landscape and upslope of a potential permeability barrier. The water resource target map interpretation can be largely automated in GIS (Street and Pracilio, 2000).
Figure 1 Water resource target map for the Toolibin Lake Catchment (55,00 ha) in Western Australia. Areas in blue are most prospective for fresh water. Map created using radiometrics, electromagnetics, DTM, magnetics and bore data.
Getting good soils information is an important step in making a land management decision and airborne radiometric surveys are the only easily collected data set that measure differences in the soil profile.
Figure 1 Soil map produced by classification of radiometric data seen in figure 6 with additional input from EM and detailed point sourced soil data. Although the classification can divide up the landscape into classes that can be related to soil types the classification processes actually loses some of the information in the radiometric data.
However, the radiometric highs and lows mean nothing until the interpreter goes through the process of turning the data into information in a form that is useable. The radiometric data however is an excellent spatial data set on which to ‘hand’ more detailed soil information collected laboriously by digging pits and soils analysis.
Recent research has been aimed at developing computer-assisted interpretation techniques to address this problem (Anderson-Mayes, 2000). Simple spatial analysis techniques implemented on a geographic information system (GIS) can be used to improve information extraction.
There is valuable information in airborne geophysical data that can be used to gain knowledge about sub-surface processes which allow a land manager to make optimal decisions to maximise productivity. The keys are the quality and resolution of the data and ensuring the interpreter asks the right questions of the data.
Airborne geophysics can measure sub-surface variations in the landscape that will impact on the short, medium and long term sustainability of agriculture. In the short term, radiometric data can be used for improved soil information and therefore optimise crop placement and management. In the medium term a suit of geophysics can guide farmers to areas prospective for water bores. In the long term a better understanding of the processes that lead to salinity can be obtained allowing better placement of remedial actions in land management.
There are already considerable areas of Australia that have been covered by airborne geophysical surveys. Much of these data is too poor resolution for most land management purposes. There are however large parts of Victoria, WA, NSW and Queensland agricultural areas covered by surveys with line spacings of 200 metres which could be utilised. Most of these data is either freely available or very cheap to obtain. There is an opportunity for farmers to utilise these data to direct farm productivity.
The future of airborne geophysics as a land management tool now rest with those who manage the land. Too much development has been technology driven in the past and there has been too much focus on developing systems that are optimised for mineral exploration despite the funding for these systems being based on environmental applications. The next generation of systems must be able to incorporate a full suite of geophysical tools, operate at low level and at low cost.
Anderson S., Street, G and Anderson H. (1995) The use of airborne geophysics to understand salinity processes and develop a farm plan on a farm in the south-west of Western Australia. Proceedings 4th National Conference on the Productive use of Saline Lands, Albany WA .
Anderson-Mayes, A. (2000). Enhancing Interpretation of Multivariate Airborne Geophysical Data for Dryland Salinity Studies. PhD Thesis (unpublished), University of Queensland.
Engel, R., McFarlane D., and Street, G.J. (1987). The influence of dolerite dykes on saline seeps in south-western Australia. Aust J Soil Res 25, 125-136.
George, R.J., Beasley, I., Gordon, D., Heislers, D., Speed, R., Brodie, R., McConnell, C., and Woodgate, P. (1998) Evaluation of airborne geophysics for catchment management. National Report to AFFA on National Airborne Geophysics Project.
Lane, R., Golding, C., Green A., Pik, P., and Plunkett, C., (1998). The TEMPEST Airborne EM system. Poster paper presented at the 13th ASEG Conference, Hobart 8-12 November, 1998.
Lane, R. and Pracilio, G. (2000). Visualisation of sub-surface conductivity derived from airborne EM. Proceedings of SAGEEP 2000 Environmental and Engineering Geophysics Society.
Nulsen, R.A., Beeston, G., Smith R., and Street G. (1995). Delivering a technically sound basis for catchment and farm planning. Proceedings WALIS Forum '96 Perth WA.
Pracilio, G., Street, G.J., Nallan Chakravatula, P., Nash, C., Owers, M., Triggs, D., and Lane, R. (1998). National Dryland Salinity program. Airborne Geophysical Surveys to assist in planning for salinity control – 3. Lake Toolibin SALTMAP Survey – Interpretation Report – December 1998. (110 pages, plus appendices and maps.) Unpub report for National Airborne Geophysics Project, AGSO
Street, G.J., & R Engel, (1990). Geophysics and dryland salinity in Environmental and Groundwater Geophysics, Society of Exploration Geophysicists, Tulsa, Ok. USA.
Street, G.J., & G P Roberts, (1994). SALTMAP - High resolution airborne EM for electrical conductivity profiling. Proceedings of Resource Technology ‘94. University of Melbourne, Australia.
Street, G.J., Pracilio, G., Owers, M., Triggs, D., and Lane, R. (1998). National Dryland Salinity program. Airborne Geophysical Surveys to assist in planning for salinity control – 1. Willaura SALTMAP Survey – Interpretation Report – June 1998. (92 pages, plus appendices and maps.) Unpub report for National Airborne Geophysics Project, AGSO
Street, G.J., and Pracilio, G. (2000), Catchment reinterpretation Toolibin Lake. Unpub report for Conservation and Land Management WA.