Your questions answered

I keep hearing about two different methods for subsidence monitoring – InSAR and LiDAR. What is the difference?

Interferometric Synthetic Aperture Radar (InSAR) is used for establishing a trend in ground motion (i.e. changes over time). Light Detection and Ranging (LiDAR) is primarily used for establishing a baseline ground slope and drainage at a particular time.

InSAR is a satellite method used to measure ground movement at the same location over time. Measurements taken every 6 to12 days provide a reliable trend of ground motion at that specific location, similar to measurements of groundwater level changes over time from a monitoring bore location.

LiDAR provides a snapshot of relative elevation of various points on the ground, typically about 4 to 6 points within a square metre. This is then used to create a digital elevation model (DEM) to accurately draw slopes and drainage patterns to derive a baseline of ground surface.

OGIA says that InSAR is the best way to monitor subsidence over time. How confident is OGIA about this and how accurate is the data?

InSAR data collected as radar signals from satellites is currently acquired every 6 to 12 days. This signal is then converted to a change in ground elevation measurement using proprietary methods. OGIA is currently using a method and outputs developed by Tre Altamira, which is the most widely used supplier in the Surat Basin.

Measurements of ground motion from InSAR are accurate to within one millimetre per year. Conversion of radar signals may not be possible in more disturbed areas, such as parts of cultivated areas in the Condamine Alluvium. In cultivated areas the measurements are derived where more stable features are located – such as farmhouses, silos and roads. Collectively, those measurements provide a reasonable assessment of subsidence in the area.

Is InSAR data missing in cultivated areas?

An InSAR signal is collected at every location. Data obtained in heavily cultivated areas is unable to be reliably converted to ground motion at this stage. It is likely that ongoing improvements to InSAR processing methods in the future may make it possible for more signals from those areas to be converted to ground motion. OGIA is currently working with Tre Altamira and other providers to improve InSAR coverage.

The UWIR 2021  (PDF, 16.9MB) (s7.3.4) explains how the data from individual and nearby points may indicate variable ground movement. The overall trend can reliably assess changes in ground elevation over a period of time. The subsidence component can be extracted from that overall pattern noting that there are natural variations in ground movement of about ±25mm per year that are unrelated to subsidence.

If InSAR doesn’t work that well in cultivated areas, then why not use a different method?

At this stage, InSAR is the most practical method to monitor changes in ground elevation at regular time intervals across a broad area. Despite the limitation of InSAR in cultivated areas, there are sufficient measurement points to monitor overall subsidence patterns with reasonable confidence.

Who reports the findings from InSAR data and how do I access it?

OGIA gets the InSAR data directly from the supplier (Tre Altamira) and will continue to periodically report trends in ground motion.

How confident is OGIA that there is sufficient baseline data available to determine subsidence in the future?

The primary purpose of the baseline assessment is to establish a pattern of drainage and farm slope prior to CSG production, so that any future changes to landform from CSG depressurisation can be determined. Based on some investigations, OGIA concludes that airborne LiDAR is the most appropriate and fit-for-purpose technique in this situation. This is detailed in the UWIR 2021 (s7.5.1).

I hear that LiDAR accuracy is only ±50mm, so how can this be relied upon, when subsidence is also within that range?

LiDAR is more useful for establishing baseline landform and comparing overall change in slope over time. It lacks sufficient accuracy for assessing changes to ground elevation at a particular point over time. Repeated surveys for the same location may be vertically offset by 50 to 100mm. OGIA does not recommend using LiDAR data to assess or compare changes in ground motion at a particular location.

Is ground survey better to monitor subsidence and establish baseline?

OGIA undertook a pilot study to compare different methods for establishing a baseline for subsidence. This is described in UWIR 2021 (s 7.5.2). The study concluded that the surface drainage pattern across a farm field, derived from aerial LiDAR survey data is the most suitable and cost-effective method to establish background slope and baseline landform. Given the high spatial density of data points captured with aerial LiDAR, this method is also better at identifying minor slopes, localised slope reversals and depressions.

I hear that LiDAR is not very effective in cultivated areas or where water is present. Is it reliable?

LiDAR does have limitations in cultivated areas where there is canopy cover, or where water has ponded. In these situations, a combination of multiple surveys – done when crop coverage is minimal and water is not present – can mitigate that limitation and reliably establish a baseline slope at the farm scale.

Is the ‘OGIA Elevation Profile Tool’ designed to estimate subsidence?

No. The elevation profile tool is a web-based user-friendly tool designed to enable landholders to draw section lines and produce land elevation profiles ‘on the fly’ from the available data. It is not a data repository.

The profiles created from the tool provide an estimate of landform slope from an individual survey, and provides for comparison of changes in slope between surveys. The tool is not suitable for comparing changes in elevation with time at a specific location, or for directly deriving CSG-induced subsidence. Other types of data and interpretive techniques must be used in combination to derive CSG-induced subsidence.

Who captures the LiDAR data?

The UWIR 2021 includes an obligation on Arrow Energy for the annual collection of LiDAR data over the western Condamine Alluvium. Arrow Energy engages a specialist contractor to acquire the LiDAR data and generate a ground digital elevation model (DEM). The point cloud – i.e. the raw data, aerial imagery, DEM derived from the raw data, and metadata (contextual information about the data) – is then provided to OGIA.

How can landholders access the data?

The DEM provides information about the ground elevation on a 1×1-metre grid and is the most practical dataset for download and further use – due to its manageable file size.

DEMs are available for download from the national LiDAR repository, ‘ELVIS’. The associated metadata is made available through the Queensland Spatial Catalogue. Aerial imagery can be viewed on the Queensland Globe, using the ‘historical image’ functionality.

Raw LiDAR data – i.e. the point cloud – is an extremely large file and is not practical for direct download. However, if needed, the raw data can be made available upon request to OGIA – noting that this may involve significant processing time.

Given the large and complex nature of the datasets, system requirements and data quality checks, it can be months from when the data is received before it is published. Coordination from multiple organisations is also needed to utilise existing systems and procedures for similar datasets. OGIA will continue to explore ways to streamline this further for greater efficiency.

I have noticed a sudden jump in the recorded elevation from the LiDAR data. Why is that?

Given the scale of the data acquisition, multiple flights over several weeks are necessary, with some overlap between adjacent flight lines. When data from different periods is stitched into a regional coverage, anomalies may occur. Thus far, this appears to affect less than 1% of the dataset. OGIA is continuing to develop tools to identify errors and anomalies in the data provided. When interpreting slope changes at a field scale, the effect of such anomalies can be addressed by using data from individual flight lines.

Modelling presented in the UWIR 2021 is at a regional scale. How does that help in farm-scale assessment?

The predictions of subsidence in the UWIR 2021 are based on a regional-scale model at 1.5km grid resolution. The model is informed by local geological, hydrogeological, geomechanical and monitoring data, where available. The regional assessment helps to identify areas where more detailed assessment may be required at the farm scale.

Farm-scale assessments help in understanding the timing and differential subsidence at a scale appropriate to managing farming practises. An important consideration is the configuration of CSG wells and production schedules in and around the farms.

Would farm-scale modelling assist in farm-scale assessment?

Farm-scale model predictions may be useful in assessing how slopes in a farm may change over time. However, farm-scale modelling may not always be necessary, depending upon the scale of subsidence and planned production.

Does the UWIR 2021 model accurately match the subsidence that has already occurred?

The OGIA subsidence modelling is matched to the overall regional subsidence pattern that has already occurred. OGIA’s overall modelling approach is detailed in the UWIR 2021 (s7.4.2).

Does OGIA use a numerical model or analytical model in predicting subsidence?

The UWIR 2021 (s7.4.2) describes a combination of numerical and analytical models used by OGIA for making subsidence predictions. This includes an additional uncertainty analysis.

A research paper (Wu et al, 2018), referenced by some landholders in relation to modelling for subsidence, states that ‘subsidence is difficult to predict using the analytical methods due to the complex response of pore pressure of the entire geological profile to the gas extraction’. The context for this research paper is very different. It refers to a model that predicts depressurisation as well as compaction through a single model. In contrast, OGIA has developed separate models for those two elements because of the complexity associated with the Surat Basin.

OGIA’s groundwater model (UWIR, s6.4) has evolved in the last 10 years to become a reliable tool for predicting regional depressurisation patterns (i.e. groundwater impacts). It is a very complex numerical model that considers dual-phase flow and other CSG-specific processes to predict depressurisation. That depressurisation then provides input to a separate subsidence model. OGIA’s subsidence model uses a combination of numerical and analytical modelling techniques.

Does the UWIR subsidence modelling factor in the compaction of surrounding formations, or just the coal seam alone?

OGIA’s model considers both the coal seam formation and surrounding formations.

How confident is OGIA in the predictions, and how likely is it that the predictions will change?

Similar to groundwater flow model predictions (UWIR 2021, s6.1), the two primary factors affecting predictions of subsidence are the model itself and the resource development profile, which comprises the location of proposed wells and timing of development. A change to either of these factors will result in a change to predictions. In most cases, it is generally a change in development profile that causes a change to predictions.

Changes in a development profile will likely affect the pattern and timing of subsidence at a local scale, even though the overall magnitude and extent of regional subsidence may not change substantially unless there are significant changes to planned production.

Does the UWIR 2021 model account for vertical or directional well configuration?

The UWIR 2021 regional model represents all wells based on their intake locations. New tools developed by OGIA allow simulation of directional wells for farm-scale modelling.

Does the UWIR 2021 model account for existing as well as proposed future development?

The predictions of groundwater impacts and subsidence impacts are based on existing and future well production.

For existing wells, the location and production timing data is used in the UWIR 2021 from the time each well began production. For future wells, the location and timing of production is supplied by the industry on an annual basis.

Is production from all wells turned on simultaneously in the UWIR 2021 model input?

The existing wells are switched on in the model based on when they started producing. Future wells are also switched on at different times, based on the information provided by the industry in its annual development plans.

Does the UWIR 2021 model assume uniform or average coal thickness and permeabilities?

The thickness of coal is not uniform in the UWIR model. Variable coal thickness is considered in the model, based on data from nearby CSG wells and statistical interpolation of such data.

Variabilities in the permeability and porosity of the coal seams and the underlying and overlying formations are also considered in the UWIR modelling. It is not uniform and varies spatially.

Will farm-scale modelling predict more subsidence?

Farm-scale modelling is not currently within the legislative mandate of OGIA. If farm-scale modelling is undertaken in the future, it is likely to provide a better understanding of subsidence magnitude and pattern over time at a local scale. It will also provide better understanding of the influence of changes in production from nearby CSG wells. The overall magnitude of subsidence is likely to be comparable to the regional predictions.

Will subsidence be uniform and will CSG wells create a pattern of subsidence hollows around the wells?

OGIA does not use the term uniform as it is not the right term in the context and prone to misinterpretation. Instead, the UWIR 2021 refers to a term relatively uniform to represent how a subsidence pattern changes and compares over time.

More subsidence is likely to occur around the CSG wells at the commencement of production, before merging with subsidence from other nearby CSG wells. This results in a relatively more uniform pattern due to the interference effect. Closer well spacing will create relatively more uniform subsidence. This is likely to occur within the first few months of production, which is consistent with InSAR monitoring observations from historical CSG development areas. As development progresses, a gentle slope of subsidence is likely to develop. This is typically less than 10mm per kilometre but may be more in some instances.

Will directional wells cause non-uniform subsidence?

Compared to single CSG wells, directional well configurations are likely to result in a relatively more uniform pattern compared to vertical wells. This is because of the closer proximity of sub-surface intakes compared to vertical well configurations. However, an isolated configuration of directional wells with no other nearby production will create an inward-sloping subsidence pattern.

Who is going to undertake farm-scale modelling and when will that be available to landholders?

Farm-scale modelling is not currently within the legislative mandate of OGIA. OGIA continues to develop improved methods for subsidence modelling at both the regional and local scales. Farm scale modelling may be undertaken by the tenure holder to assist their farm scale assessments.

How does the Horrane fault affect subsidence?

The Horrane fault plays a key role in the pattern of depressurisation in the western parts of the Condamine Alluvium. This may directly affect subsidence. OGIA predicted higher subsidence gradients across the faulted areas. It is also an area of further research by OGIA.

Will the Condamine Alluvium be subject to compaction and will this be more prevalent in the western margins where the transition zone (between the Alluvium and the Walloon Coal Measures) is either thin or absent (p46, UWIR 2021)?

Cumulative CSG-induced compaction from all the underlying geological layers beneath the Condamine Alluvium will be reflected at the surface. Along the western margins of the Condamine Alluvium, it is the Walloon Coal Measures depressurisation pattern, not the Condamine Alluvium, that will predominantly drive subsidence at the surface. Compaction within the Condamine Alluvium is expected to be a negligible proportion of the overall subsidence. Groundwater extraction from the Condamine Alluvium in the last few decades has likely caused far more compaction in the Condamine Alluvium in relative terms.

Has OGIA factored the impact of subsidence on specific farms into the model?

OGIA has not made any farm-specific assessment of subsidence because it is not within OGIA’s current legislative mandate. At the regional scale, OGIA has used subsidence monitoring data to history-match its regional model to best represent historical subsidence. OGIA’s uncertainty analysis ensures observed subsidence is factored in the predictions and provides probabilistic estimates of future subsidence.

Has the coal shrinkage been factored in the subsidence modelling?

Temporal changes associated with coal shrinkage have not currently been considered explicitly in the UWIR subsidence modelling because they are unlikely to materially affect that modelling. Any shrinkage contributing to subsidence is reflected in the subsidence monitoring data to which the modelling is calibrated.

Subsidence arising from shrinkage is an ongoing field of research. In April 2023, OGIA published an update on its research on coal shrinkage  (PDF, 1.5MB). OGIA is also developing methods to explicitly represent coal shrinkage in future modelling.

Has OGIA’s modelling work been reviewed?

OGIA seeks advice on and review of all its scientific work underpinning the UWIR. The technical advisory panel consists of various hydrogeology and modelling experts primarily from research organisations.

The ongoing review occurs at various stages to seek feedback on proposed methodologies, interim results and findings, and final outputs. A formal review of the model concluded that it is fit for purpose.

Not all referenced sources in the UWIR 2021 are on the OGIA website. When and where will these be available?

All key information, results and conclusions are summarised in the UWIR. OGIA prepares separate reports to provide further technical details as companion documents. Those that are necessary for the UWIR approval are published together with the UWIR. Others providing additional contextual and technical information on approaches and methods used during UWIR cycles, are published online as they are finalised.

Publication of some referenced reports are delayed due to significant disruptions and diversion of OGIA resources in dealing with emerging subsidence matters in the post-UWIR period. Delaying the publication of some companion documents provides an opportunity to update additional work that OGIA is currently undertaking, including farm-scale modelling and LiDAR.

What was the helicopter survey about in mid-2023?

OGIA commissioned SkyTEM Australia to fly an airborne electromagnetic (AEM) survey in the western Condamine Alluvium in May 2023 (see factsheet PDF, 455KB). The survey was designed to gather new data to advance understanding of the shallow sub-surface geology and groundwater system in this area, with a particular focus on the Horrane Fault system. The data collected will support the continued evaluation of the connection between the Condamine Alluvium and underlying coal seams.

What is an AEM survey?

AEM is an airborne geophysical technique where an aircraft is fitted with specialised equipment to gather information about the shallow sub-surface, using electromagnetism. In the case of the May 2023 survey, flown by SkyTEM, this was done by helicopter with a frame suspended below. A helicopter-flown AEM survey uses an electric current generated in a loop suspended on the frame below the aircraft. The electric current produces a magnetic field, which penetrates the ground (up to around 200 metres) and creates secondary electric currents (eddy currents) in any conductive material that it passes through. These eddy currents then produce a secondary magnetic field, which is detected by receivers located on the frame. The received signal is recorded at regular time intervals along each flight line and generally equates to a spacing of around 30 m.