Ken Lawrie

Resolution analyses for selecting an appropriate airborne electromagnetic (AEM) system

The choice of an appropriate airborne electromagnetic system for a given task should be based on a comparative analysis of candidate systems, consisting of both theoretical considerations and field studies including test lines. It has become common practice to quantify the system resolution for a series of models relevant to the survey area by comparing the sum over the data of squares of noise-normalised derivatives. We compare this analysis method with a resolution analysis based on the posterior covariance matrix of an inversion formulation. Both of the above analyses depend critically on the noise models of the systems being compared. A reasonable estimate of data noise and other sources of error is therefore of primary importance. However, data processing and noise reduction procedures, as well as other system parameters important for the modelling, are commonly proprietary, and generally it is not possible to verify whether noise figures have been arrived at by reasonable means. Consequently, it is difficult – sometimes impossible – to know if a comparative analysis has a sound basis. Nevertheless, in the real world choices have to be made, a comparative system analysis is necessary and has to be approached in a pragmatic way involving a range of different aspects. In this paper, we concentrate on the resolution analysis perspective and demonstrate that the inversion analysis must be preferred over the derivative analysis because it takes parameter coupling into account, and, furthermore, that the derivative analysis generally overestimates the resolution capability. Finally we show that impulse response data are to be preferred over step response data for near-surface resolution. In this paper, we compare two ways of analysing resolution in AEM surveys: using the sensitivities, and using the posterior covariance matrix. We apply these to compare two response types, impulse response and step response, and to compare two AEM systems, SkyTEM and TEMPEST, for hydrogeophysically relevant models.

Using AEM and GMR Methods for Non-Invasive, Rapid Reconnaissance Mapping and Characterisation of Groundwater Systems in the Kimberley Region, Northern Australia

The East Kimberley Region of north-western Australia has been identified as a priority for potential agricultural development. Within this region, the Ord Bonaparte Plain is remote, with limited access in an area of great cultural and environmental sensitivity. Initially, spatio-temporal mapping using remote sensing (and potential field) data, combined with data on the deeper basin geology was used to plan an airborne electromagnetics (AEM) survey. The relatively resistive nature of the basin sediments has enabled the AEM to map the hydrostratigraphy to depths of 300-500m, except in the coastal zone affected by seawater intrusion. Two overlying aquifers, separated by a faulted, ‘leaky’ aquitard, have been identified. The AEM and remote sensing data were subsequently used to plan a ground magnetic resonance (GMR) survey. The latter has enabled a water table map to be constructed in an area with almost no drilling, while also enabling key aquifer properties to be determined. The target aquifer has a high free water content and high transmissivity. The GMR results have been validated by drilling, borehole Nuclear Magnetic Resonance (NMR), and induction logging. Integration of AEM, GMR and temporal (Landsat) remote sensing data has enabled rapid mapping and characterisation of the groundwater system in a data-poor, culturally and environmentally sensitive area. These data have also revealed complex faulting within and bounding the aquifer system, delineated the sea-water intrusion interface, and mapped groundwater dependent ecosystems. These data have been used to target drilling and pump testing that will inform groundwater modelling, water allocations and development decisions.

An Integrated Hydrogeophysical Approach to Exploring for Groundwater Resources in Southern Northern Territory

In Australia’s semi-arid and arid interior, groundwater resources provide water supply security for agriculture and community consumptive use and are critical for underpinning economic development. . The Southern Stuart Corridor Project in central Australia, is an inter-disciplinary study which aims to better characterise regional groundwater systems and identify the location, quantity and quality of new groundwater resources. The main aims of the project are(1) to de-risk investment in development of a potential agricultural precinct in the Western Davenport Basin, and expansion of horticulture in Ti-Tree Basin, (2) to identify future water supplies for Alice Springs and Tennant Creek, and (3) for regional water supplies for mineral resource development. The project is funded by Geoscience Australia (GA) as part of the Exploring for the Future (EFTF) Programme. The project integrates airborne electromagnetic (AEM), ground geophysics (ground magnetic resonance (GMR) and borehole geophysics (Induction, gamma and nuclear Magnetic Resonance (NMR)) with drilling and pump testing; hydrochemistry and geochronology; and geomorphic, geological, hydrogeological and structural mapping and modelling. Advancements in temporal remote sensing technologies for surface hydrology, vegetation and landscape mapping are also used to facilitate the identification of recharge and discharge zones and groundwater-dependent vegetation. This paper reports on initial AEM inversion results for the Alice Springs, Ti-Tree Basin, Western Davenport and Tennant Creek areas and the use of a machine learning approach for rapid geological and hydrogeological interpretation of the AEM data. These machine learning approaches have the potential to significantly reduce interpretation time and facilitate the rapid delivery of project results.

A novel approach to comparing AEM inversion results with borehole conductivity logs

Borehole conductivity logs, besides being useful for identifying, interpreting and correlating geological formations, also find widespread use as auxiliary information in the inversion of airborne electromagnetic (AEM) data. One of the quality checks often applied to AEM inversion results is a comparison between the conductivity structures revealed by borehole conductivity logs in the survey area and the AEM inversion model closest to the borehole, often called an ‘FID point comparison’. Another use of borehole conductivity logs is found in modern AEM inversion procedures, where the borehole conductivity information is included as prior information in a laterally constrained inversion. In most former and present practices, AEM layer conductivities are compared with the measured conductivity in the borehole. However, the borehole conductivity is essentially an apparent conductivity – it is a measured data value – while the AEM layer conductivities are model parameters resulting from inverting AEM data. To avoid comparing data and model parameters we suggest a conceptually clear approach based on an inversion of the borehole conductivity data to obtain a borehole conductivity model, which in turn can be compared with the AEM model. Furthermore, the AEM forward response of the borehole model can, in a consistent way, be compared with the AEM data. In both approaches, we keep track of uncertainty and define quantitative, uncertainty-normalised measures of the difference between borehole and AEM values, and we find simple functional relationships between the two. The methodology is demonstrated on the AEM data and conductivity logs of the Broken Hill Managed Aquifer Recharge (BHMAR) project. Two consistent methods of comparing borehole induction logs with models from inversion of AEM data have been developed: one in model space and one in data space. The two methods are applied to the Broken Hill Managed Aquifer Recharge project conducted by Geoscience Australia on AEM data from the SkyTEM system.

Response to comments by Adam Smiarowski and Shane Mulè on: Christensen, N., and Lawrie, K., 2012. Resolution analyses for selecting an appropriate airborne electromagnetic (AEM) system, Exploration Geophysics, 43, 213–227

We analyse and compare the resolution improvement that can be obtained from including x-component data in the inversion of AEM data from the SkyTEM and TEMPEST systems. Except for the resistivity of the bottom layer, the SkyTEM system, even without including x-component data, has the better resolution of the parameters of the analysed models.

Using airborne EM and borehole NMR data to map the transmissivity of a shallow semi-confined aquifer, western NSW.

The Broken Hill Managed Aquifer Recharge (BHMAR) project aimed to define key groundwater resources and aquifer storage options in the lower Darling River floodplain of western NSW. The project was multi-disciplinary and utilised airborne electromagnetics (AEM), borehole nuclear magnetic resonance (NMR) and LiDAR DEM data and lithological, hydrostratigraphic and hydrochemical information to develop a suite of hydrogeological and groundwater property maps and products. This abstract discusses the methods and results of estimating the transmissivity of the semi-confined target aquifer. Hydrostratigraphy and hydraulic texture classes were mapped by interpreting the AEM data in conjunction with borehole geophysics and lithological information. Aquifer transmissivity was statistically derived by combining borehole NMR hydraulic conductivity estimates with the mapped 3D distribution of texture classes and hydrostratigraphic units. Using a statistical and GIS approach, the derived aquifer thicknesses in the key areas ranged from 20 - 40 m and the lower and upper transmissivity bounds ranged from 1 to 10 m2/d, and 10 m2/d to 1000 m2/d, respectively.

Optimizing Airborne Electromagnetic (AEM) Inversions for Hydrogeological Investigations using a Transdisciplinary Approach

High-resolution hydrogeophysical data are increasingly acquired as part of investigations to underpin groundwater mapping. However, optimization of AEM data requires careful consideration of AEM system suitability, calibration, validation and inversion methods. In modern laterally-correlated inversions of AEM data, the usefulness of the resulting inversion models depends critically on an optimal choice of the vertical and horizontal regularization of the inversion. Set the constraints too tight, and the resulting models will become overly smooth and potential resolution is lost. Set the constraints too loose, and spurious model details will appear that have no bearing on the hydrogeology. There are several approaches to an automatic choice of the regularization level in AEM inversion based predominantly on obtaining a certain pre-defined data misfit with the smoothest possible model. However, we advocate a pragmatic approach to optimizing the constraints by an iterative procedure involving all available geological, hydrogeological, geochemical, hydraulic and morphological data and understanding. In this approach, in a process of both confirming and negating established interpretations and underlying assumptions, the inversion results are judged by their ability to support a coherent conceptual model based on all available information. This approach has been essential to the identification and assessment of MAR and groundwater extraction options in the Broken Hill Managed Aquifer Recharge project.

Use of the Resolve Airborne Electromagnetic (AEM) Data in the Assessment of the Salinity Hazard and Risk to Iconic River and Wetland Ecosystems, Murray River, Se Australia

An AEM survey using the RESOLVE frequency domain system has been acquired along a 450 km reach of the Murray River in SE Australia. the AEM data were inverted using the holistic Inversion method, enabling key elements of the hydrogeological system in the shallow sub-surface (top 20-50m) to be mapped with high confidence levels. the AEM data have been used in conjunction with remote sensing, and hydrogeological and hydrogeochemical data obtained from drilling, to determine that healthy vegetation along the Murray River is generally associated with the presence of significant river ‘flush zones’ where fresh groundwater is present at shallow depths, and groundwater salinity is low. the study has also found that the corollary is true: where the river is ‘gaining’, and salt stores are high, vegetation health is generally in decline. Similarly, the AEM data show there is a marked decline in vegetation health towards the western edge of the iconic Gunbower State forest. This appears to be associated with salt being mobilised from irrigation districts on the western margins of the Gunbower forest. In the areas where the river flush zones are discontinuous, and the salt stores and water tables are closer to surface, there is also a risk of salt ingress to the river. in these areas, the data identify areas for targeted salinity management, including sites for potential Salt interception Schemes. This study fills important knowledge gaps particularly the distribution of key elements of the hydrostratigraphy, salinity extent, and the relationships between vegetation health, salinity and groundwater processes. in particular, the project has successfully integrated AEM, remote sensing, and lithological and hydrogeological data from drilling, to identify reaches of the River Murray and areas of iconic wetland ecosystems at risk from groundwater salinisation. these datasets provide geospatial context for targeted salinity and groundwater management actions.