Amplitude Variation With Offset Research Articles

Petrophysical inversion based on f-s-r amplitude-variation-with-offset linearization and canonical correlation analysis

The prediction of petrophysical properties, such as porosity and rock-fluid volumes, from partially stacked seismic data typically requires a rock-physics model that often is lithology dependent and difficult to calibrate. We adopt canonical correlation analysis (CCA) to infer the underlying relation between petrophysical properties and elastic attributes estimated from seismic data. We develop a two-step inversion approach: first, we predict elastic properties from partially stacked seismic data using a Bayesian linear inverse method based on an amplitude-variation-with-offset (AVO) linearization in terms of fluid, rigidity, and density factors, and then we predict petrophysical properties from the estimated AVO attributes using CCA. The novelty of our approach is the application of CCA to the fluid and rigidity factors, which avoids the calibration of an explicit rock-physics model by automatically deriving a linear relation in the lower dimensional space of the canonical variables. The parameterization of the linearization in terms of fluid, rigidity, and density factors maximizes the correlation with respect to the petrophysical properties of interest. Furthermore, the probabilistic approach is extended to the petrophysical inversion using Bayesian linear theory and the posterior distribution of petrophysical properties conditioned by seismic data is computed by combining the probability distributions obtained from seismic and petrophysical inversion to propagate the uncertainty from the seismic to the petrophysical domain. The inversion is validated on a synthetic case that finds high accuracy of our formulation. A case study with synthetic and real partially stacked seismic data also is presented and compared to a traditional inversion with an explicit rock-physics model.

Amplitude variation with incidence angle inversion of fluid parameters in porous media using an accurate estimation of the Jacobian matrix

Amplitude variation with incidence angle (AVA) or amplitude variation with offset (AVO) inversion provides remarkable information in discriminating lithology in hydrocarbon reservoirs. Fluid parameters (e.g., porosity, water, and oil saturations) are normally obtained through two steps and difficult to obtain from seismic data directly and accurately. Theoretically, the exact Zoeppritz equations and Biot-Gassmann equations can be combined together to invert fluid parameters using seismic data directly. We carry out AVA inversion to invert porosity, water, and oil saturations from seismic data using the exact Zoeppritz equations and accurate estimation of the Jacobian matrix. This matrix is composed of the partial derivatives of P- and S-wave reflection coefficients with respect to fluid parameters. Characteristics of these partial derivatives are analyzed. In addition, the iterative equations for AVA inversion are derived based on the Taylor series expansion. The numerical tests indicate that this algorithm can invert fluid parameters directly and accurately using seismic gathers. Unlike the conventional AVA using the approximations, the inversion using the exact Zoeppritz equations and accurate estimation of the Jacobian matrix is not limited by the selection of incidence angles and reflection interfaces. This method also is applicable to seismic data collected from formations with strong elastic contrast interfaces or large incidence angles.

Sparse Radon transform in the mixed frequency-time domain with ℓ1-2 minimization

Due to the finite acquisition aperture and sampling of seismic data, the Radon transform (RT) suffers from a smearing problem which reduces the resolution of the estimated model. In addition, inverting the RT is typically an ill-posed problem. To address these challenges, a sparse RT mixing the [Formula: see text] and [Formula: see text] norms of the RT coefficients in the mixed frequency-time domain is developed, and it is denoted as SRTL1-2. In most conventional sparse RTs, the sparse constraint term often is the [Formula: see text] norm of the Radon model. We prove that the sparsity effect of the [Formula: see text] minimization is better than that of the [Formula: see text] norm alone by comparing and analyzing their 2D distribution patterns and threshold functions. The difference of the convex functions algorithm and the alternating direction method of multipliers algorithm are modified by combining the forward and inverse Fourier transforms to solve the corresponding sparse inverse problem in the mixed frequency-time domain. Our method is compared with three RT methods, including a least-squares RT (LSRT), a frequency-domain sparse RT (FSRT), and a time-invariant RT in the mixed frequency-time domain based on an iterative 2D model shrinkage method (SRTIS). Furthermore, we modify the basis function in SRTL1-2 by including an orthogonal polynomial transform to fit the amplitude-variation-with-offset (AVO) signatures found in seismic data, and we denote this as high-order SRTL1-2. Compared to the SRTL1-2, the high-order SRTL1-2 performs better when processing seismic data with AVO signatures. Synthetic and real data examples indicate that our method has better performance than the LSRT, FSRT, and SRTIS in terms of attenuation of multiples, noise mitigation, and computational efficiency.

AVO study on a known geothermal reservoir located in the fractured carbonate formations of the pre-Cenozoic basement, Northwest Hungary

Amplitude Versus Offset (AVO) analysis has been successfully applied in the hydrocarbon industry for more than three decades. This effective tool of the joint seismic and well-log data processing enables to predict and analyze fluid saturated porous geological formations. AVO methodology is based on the anomalous behavior of the pre-stack reflected amplitudes observed from fluid bearing rocks. However, the potential of AVO methodology is still unexploited in geothermal exploration, although the lithology and rock physical properties are very similar. In this study, we summarize the theoretical backgrounds and calculate synthetic AVO responses of a known geothermal reservoir located in the fractured carbonates of the Mesozoic basement. We demonstrate that the AVO response of a deep geothermal reservoir can be quite different from the amplitude response observed from a hydrocarbon bearing clastic formation. AVO attributes of the investigated geothermal reservoir are presented, and the potential of its detection by seismic amplitude data are discussed.

Fluid discrimination incorporating amplitude variation with angle inversion and squirt flow of the fluid

Pre-stack seismic inversion is an important method for fluid identification and reservoir characterization in exploration geophysics. In this study, an effective fluid factor is initially established based on Biot poroelastic theory, and a pre-stack seismic inversion method based on Bayesian framework is used to implement the fluid identification. Compared with conventional elastic parameters, fluid factors are more sensitive to oil and gas. However, the coupling effect between rock porosity and fluid content is not considered in conventional fluid factors, which may lead to fuzzy fluid identification results. In addition, existing fluid factors do not adequately consider the physical mechanisms of fluid content, such as squirt flow between cracks and pores. Therefore, we propose a squirt fluid factor (SFF) that minimizes the fluid and pore mixing effects and takes into account the squirt flow. On this basis, a novel P-wave reflection coefficient equation is derived, and the squirt fluid factor is estimated by amplitude variation with offset (AVO) inversion method. The new reflection coefficient equation has sufficient accuracy and can be utilized to estimate the parameters. The effectiveness and superiority of the proposed method in fluid identification are verified by the synthetic and field examples.

Seismic inversion for density using a transdimensional approach

Commercial and low-saturation gas (also called paleoresidual gas [PRG]) show similar strong amplitude signatures on P-wave seismic data. This poses an exploration risk in gas reservoir regions. However, density correlates inversely with gas saturation and can differentiate a zone of full gas saturation from PRG. This can improve the chances of success in terms of predrill prediction of gas saturation. Amplitude-variation-with-offset (AVO) inversion using prestack seismic data is the most commonly used technique that can estimate elastic parameters such as P-wave velocity, S-wave velocity, and density. Out of these three parameters, extracting density from seismic data is the most challenging due to its weak sensitivity to seismic reflection amplitude and the lack of good quality seismic data at far offsets. However, with recent improvements in seismic data acquisition and processing technology, which produces reliable AVO gathers, density estimates have improved. This requires that strong density sensitivity to AVO exists. Note that multiple density models may fit the data equally well. Therefore, quantifying uncertainty is crucial for interpretation and risk assessment. We apply a recently developed stochastic approach based on the Bayesian framework to solve the problem in a transdimensional framework, where the number of model parameters is treated as a variable and estimated along with the elastic properties. We use the reversible jump Hamiltonian Monte Carlo (RJHMC) algorithm to sample models from a variable dimensional model space and obtain a globally optimum model and uncertainty estimates. We use a synthetic and good quality real data set from Columbus Basin in Trinidad, which has a proven gas reservoir, to demonstrate the algorithm. The RJHMC results calibrate well with the logs and show the areal extents of the density anomalies within the 3D volume.

GPR antennas geometry and its impact over detection of near-surface water contamination through AVO Data analysis at laboratory site

Polarization is an important technique for ground penetrating radar (GPR) surveys, particularly when determining the shape, orientation, size, and precise location of any buried material. A linearly-polarized incident GPR wave experiences polarization change upon scattering from the host medium, which highly depends upon the technique and antenna orientation used for the acquisition of the GPR data. This paper focuses on the polarization of GPR data acquired at a laboratory test site. We kept the GPR antenna orientation and geometry in mind to understand how they impact the results while detecting subsurface contamination. Because of its direct indication ability, amplitude variation with offset (AVO) has been widely used to detect subsurface contamination; however, a technique to evaluate the distribution of subsurface contamination is not yet available; thus, in this study, we will also discuss the distribution of contamination using polarization of GPR data. Two major phenomena have been studied: the polarity of the data acquired and slice view with and without contamination in the testing tank, as well as trace, frequency, and amplitude plot during data analysis to determine the difference of plot with and without contamination. Our findings suggest that using the polarization geometry technique to obtain an accurate view of the subsurface contamination will not only improve the GPR AVO analysis but will also provide a clear distribution of contamination in the subsurface.

Anisotropic three-term amplitude-variation-with-offset projections: A case study

A seismic amplitude-variation-with-offset (AVO) analysis is an important element in seismic reservoir characterization and has been successfully applied in many fields. Traditionally, AVO studies are based on AVO intercept ([Formula: see text]) and gradient ([Formula: see text]) only; curvature ([Formula: see text]) is neglected because it is very sensitive to noise and avoiding noisy [Formula: see text] is very difficult in practice. The recently introduced concept of three-term (3T) AVO projections, combined with continuous improvements in seismic data acquisition and processing, allows interpreters to effectively use [Formula: see text] by optimizing projection angles based not only on rock physics but also on seismic noise. The effectiveness of the 3T AVO projection, in which [Formula: see text], [Formula: see text], and [Formula: see text] are used as the triplet of reflectivities that are summed in the projection, and the impact of random noise on the 3T AVO projection have been investigated using data from an offshore Western Australian field. The 3T AVO projection is used in the framework of the extended elastic impedance. Because the study area is known to have nonnegligible differences in anisotropy parameters between the various lithologies, its impact on [Formula: see text] and [Formula: see text] is taken into account. First, the 3T projection angle targeting lithology fraction logs is optimized using well-log data. Analyses using synthetic seismograms are then performed to investigate the impact of seismic noise. Finally, the method is applied to field data. The results find that (1) [Formula: see text] is an important component of seismic reservoir characterization even if it appears unusable and (2) seismic noise must be taken into account when optimizing the projection angle.

AVO Detuning Effect Analysis Based on Sparse Inversion

The wave field characteristics of thin reservoirs are extremely complex due to the tuning and interference between the top and bottom interfaces of the reservoirs, which leads to large uncertainty in thin layer AVO (Amplitude Versus Offset) analysis. In order to reduce the uncertainty of thin layer AVO analysis, we study the uncertainty dominant factors of the effect of thin layer on the AVO response characteristics from the aspects of theoretical derivation and forward simulation. Based on the research results, we use the AVO fitting forward method with offset and tuning utility as the joint inversion operator to establish an AVO detuning effect method, based on the sparse fitting inversion strategy, and study the objective function of the fitting inversion method. We optimize the sparsity constraints and the sparsity method to reduce the non-independence of multiparameter variables and seismic data, and the noise of inversion. Through the verification analysis of the model using actual data, the AVO detuning effect method studied in this paper has a correct and reasonable technical theory and obvious application effect.

Quantitative prediction of porosity and gas saturation based on a new dual-porosity rock-physics model and Shuey’s Poisson ratio for tight sandstone reservoirs

The prediction accuracy of physical parameters from seismic data has strong dependence on the quality of the used elastic parameters and rock-physics models especially for tight sandstone reservoirs. In this study, a new dual-porosity model with stiff and compliant porosities is firstly deduced from classical differential effective medium model according to the characteristics of both pores and fractures development in tight sandstones, it is the product of power function of both compliant and stiff porosities, and is independent of the addition order of pores. Then, the Shuey's AVO (Amplitude Variation with Offset) approximation is expanded, and a new Shuey Poisson's ratio which is the reciprocal of the difference between 1 and Poisson's ratio is defined and inverted from Shuey Poisson's ratio reflection seismogram to ensure the inversion accuracy of both P-wave impedance and ShueyPR. Next, a new two-dimension rock-physics template on P-wave impedance versus Shuey Poisson's ratio is built by the new dual-porosity model and Gassmann equations to characterize the quantitative relationship between P-wave impedance, Shuey Poisson's ratio and physical properties and is used to invert the porosity and saturation through global optimization algorithm. The new method has been applied to quantitative reservoir prediction of tight sandstone of Shaximiao Formation of QL gas field in Northwest Sichuan, China. Real application indicates that the distribution of high-porosity reservoirs with porosity more than 8% and their gas saturation can be accurately predicted.

Nonlinear petrophysical amplitude variation with offset inversion with spatially variable pore aspect ratio

Petrophysical amplitude variation with offset (AVO) inversion, which combines a rock-physics model and prestack inversion, aims at effectively quantifying reservoir-property parameters from prestack seismic data. However, for a reservoir with complex pore structures, this approach may suffer from difficulties in rock-physics modeling, which hinders solving the relevant inverse problem. We have proposed a petrophysical AVO inversion method that considers the complex pore structures within reservoir rocks. The method integrates a differential effective medium model, the Gassmann equation, and the exact Zoeppritz equation as the forward operator. In particular, the pore complexity is addressed by favoring the spatial variation of pore type. Differing from conventional sequential inversion methods, the pore-type variation is introduced by updating the aspect ratio together with the reservoir parameters during the inversion process. To enhance the stability of the results, the estimation of aspect ratio and reservoir parameters is jointly conditioned to seismic and elastic data under a Bayesian framework, which leads to the objective function with joint data misfit terms. The objective function is optimized with a simulated annealing algorithm that includes the full nonlinearity of the forward operator. The proposed method is demonstrated by testing and applying on a carbonate reservoir. It generally outperforms methods that are either purely conditioned on the seismic data or based on a constant aspect ratio.

AVO-Friendly Velocity Analysis Based on the High-Resolution PCA-Weighted Semblance

Velocity analysis using the semblance spectrum can provide an effective velocity model for advanced seismic imaging technology, in which the picking accuracy of velocity analysis is significantly affected by the resolution of the semblance spectrum. However, the peak broadening of the conventional semblance spectrum leads to picking uncertainty, and it cannot deal with the amplitude-variation-with-offset (AVO) phenomenon. The well-known AB semblance can process the AVO anomalies, but it has a lower resolution compared with conventional semblance. To improve the resolution of the AB semblance spectrum, we propose a new weighted AB semblance based on principal component analysis (PCA). The principal components or eigenvalues of seismic events are highly sensitive to the components with spatial coherence. Thus, we utilized the principal components of the normal moveout (NMO)-corrected seismic events with different scanning velocities to construct a weighting function. The new function not only has a high resolution for velocity scanning, but it is also a friendly method for the AVO phenomenon. Numerical experiments with the synthetic and field seismic data sets proved that the new method significantly improves resolution and can provide more accurate picked velocities compared with conventional methods.

Fluid Accumulation Zone by Seismic Attributes and Amplitude Versus Offset Analysis at Solfatara Volcano, Campi Flegrei, Italy

The imaging of volcanic systems is a challenging topic that attracts the scientific community’s attention. The characterization of structures and rock properties by means of seismic active methods is becoming fundamental for providing ultra-high-resolution images of the structures of interest. The Solfatara Volcano is a quiescent volcano in the Campi Flegrei resurgent nested caldera that is continuously under investigation and monitoring for its shallow activity, such as fumaroles. The purpose of this work is to characterize the fluid accumulation zone in the first 150 m depth in the middle of the crater, using several post-stack seismic attributes and Amplitude Versus Offset (AVO) analysis to characterize the contact between the CO2 and condensed water in the shallower accumulation zone. The two 400 m-long profiles to which we refer in this work have been acquired during the active Repeated InduCed Earthquakes and Noise experiment. The profiles were deployed along with the NNE-SSW and WNW-ESE directions across the whole surface of the crater including the main surface anomalies of the fumaroles, in the eastern area, and the mud-pool of Fangaia, located in the western area. The seismic pre-processing, pre-stack processing, and post-stack analysis previously applied on the NNE-SSW profile are here performed for the first time on the WNW-ESE profile, while partial-stack AVO analysis is performed for both profiles. The post-stack attributes including time gain, envelope, energy, and root mean square have been computed and extracted for determining the maximum and minimum values of amplitude zones on the migrated post-stack seismic profiles. Such anomalies are provided by complex and geometrical attributes embedding information on faults and chaotic zones. The AVO technique has also been used as a direct gas indicator to enhance fluid discrimination and identification. Finally, the analysis of the profile, seismic attributes, and near-surface structural interpretation related to the Solfatara Volcano has been incorporated into the proposed analysis. The multi-2D image depicts fluids trapped in the Solfatara Volcano at depths ranging from 10 to 50 m below the crater’s surface, as well as their migration paths up to 150 m deep: this evidenced contact between the fluids has been probably due to the solfataric alteration of the minerals, caused by the arising plume and the abovecondensed water which decreases the permeability of the rocks and forms an argillic phase working as cap-rock and trapping the gases. The application of the AVO analysis, coupled with the seismic attribute’s investigation, provides a very detailed multi-2D image of the shallower Solfatara Volcano, which outperforms in terms of accuracy the ones obtained with different tools in previous works, and that evidences the presence and the position of the liquid and the gases in the north-east area of the Solfatara Volcano.

Detection of hydrocarbon- saturated reservoirs in a challenging geological setting using AVO attributes: A case study from Poseidon field, Offshore Northwest region of Australia

Amplitude Versus Offset (AVO) methods relate seismic amplitude variations to subsurface pore fluid and lithology changes. Over the past decades, numerous AVO attributes have been extracted from seismic data and combined to detect seismic expressions associated with hydrocarbon-charged sediments. In this paper, we use an AVO attribute combination composed of gradient and scaled-Poisson reflectivity (SPR) to detect hydrocarbon expressions. Irrespective of sand-shale impedance contrast, the SPR and the gradient G produce negative anomalies for shale over gas saturated reservoir. We demonstrate that SPR-G product is a good alternative to the Intercept-Gradient product which works only for unconsolidated sands. The dataset used in this study is from Poseidon field, North Western Australia. The gas reservoirs belong to the Middle Jurassic-aged Plover Formation. The Plover Formation comprises sandstone, mudstone, and coal that deposited in a fluvial-deltaic environment. AVO analysis at a well location indicates the reservoirs belong to AVO Class II, which is characterized by small zero offset reflectivity and an anomalously large G. Due to the low impedance contrast between the reservoirs and the shale intervals, the I × G product failed to detect the fluid expressions of the reservoirs. The SPR × G product and SPR-G crossplot showed better fluid detection and lithology discrimination. Furthermore, the computation of the product SPR × G volume helped highlight the fluid response and spatial distribution of the reservoir units. New prospective undrilled areas could be identified using the SPR-G product.

Research on Prediction Method of Volcanic Rock Shear Wave Velocity Based on Improved Xu–White Model

Volcanic rock reservoirs have received extensive attention from scholars all over the world because of their geothermal, mineral, and oil and gas resources. Shear wave velocity is the essential information for AVO (amplitude variation with offset) analysis and the reservoir description of volcanic rocks. However, due to factors such as cost, technical reasons, and so on, shear wave velocity is not provided in many logging data. This paper proposes a shear wave velocity prediction method suitable for the conventional logging of volcanic rocks. Firstly, the Xu–White model is improved. The probability distributions formed by the prior information of the logging area are used to initialize the key petrophysical parameters in the model instead of the fixed parameter value to establish the statistical petrophysical model between the logging curve and shear wave velocity. Then, based on the Bayesian inversion method, the simulated P-wave velocity is matched with the actual P-wave logging data to calculate the key petrophysical parameters, and is then used for S-wave velocity prediction. The method is applied to the actual logging data of the No. 5 structure in Nanpu Sag, eastern China. The prediction effect of shear wave velocity is better than that of the conventional method, indicating the feasibility and effectiveness of this method. This study will provide more accurate shear wave velocity data for the exploration and development of volcanic reservoirs.

Application of AVO to thickness prediction of thin sand

Oil and gas are the resource buried in deep underground. The purpose of petroleum exploration is to delineate the sand body saturated with the oil or gas. The thickness of sand is an important property of sand layer in the reservoir research. In the oilfield of Eastern China, the thickness of sand body is usually very thin and below ten meters, so it is very hard for us to estimate the real thickness of sand body because its thickness is far less than the resolution limit of seismic wavelet in exploration. For further discussion of the problem, we probed into how to use the method of amplitude versus offset (AVO) to predict the thickness of sand body in this paper. Firstly, a set of thin sand models with different thickness are made; then, the prestack gathers of the models are simulated with the reflectivity method; after that, AVO attributes of prestack gathers are analyzed; next, the relationship between AVO attributes and the thickness of thin sand are established; finally, the relationship is used for the prestack gathers of the entire seismic data, and the thickness of thin sand of study area are estimated.

Robust data-driven AVO inversion algorithm based on generalized nonconvex dictionary learning

In recent years, a dictionary learning and sparse representation based data-driven amplitude variation with offset (AVO) inversion algorithm has achieved desirable performance. However, in this data-driven inversion (DDI) method, sparse coefficients sequence generated by K-SVD may not always be convergent. Furthermore, the convergence of the data-driven AVO inversion results cannot be guaranteed. On the other hand, using L2 norm as a loss function in the DDI algorithm will lead to a robustness problem when seismic data contains outlier noise. In this paper, we propose a robust AVO inversion algorithm based on generalized nonconvex dictionary learning. The generalized nonconvex dictionary learning algorithm utilizes a family of nonconvex functions as the sparsity-inducing function for accurate estimation and a convergent sparse coefficient sequence. To deal with outlier noise, a smoothed L1 norm is utilized as the loss function. Meanwhile, a new spectral Polak–Ribière–Polyak (PRP) conjugate gradient algorithm is used to optimize the entire robust AVO inversion problem. Furthermore, the convergence analysis of the proposed algorithm is provided. In comparison with the conventional DDI algorithm, the proposed algorithm is robust, convergent, and computationally efficient. Results of synthetic data and field data experiments verify the superior performance of the proposed algorithm.

Frequency-dependent Amplitude Versus Offset analysis of a Cenozoic mass-transport deposit on the Namibian slope

Abstract A partially tuned, ‘soft’ seismic reflection, overlying a regional limestone marker on the west African slope, is shown to represent a sequence, up to 30 m thick, within which various individual stratigraphic units are identified, some of which display characteristics of a mass-transport deposit (MTD). The seismic reflection is described in terms of its two-way time structure, reflection amplitude pattern, complex seismic attributes and AVO response. Frequency-dependent aspects of the AVO response, resulting from velocity attenuation and dispersion of seismic signals, are shown to indicate variations in pore-fluid gas concentration. These are exploited to assess lithological, petrophysical and pore-fluid characteristics of the various sub-units within the sequence. A relative acoustic impedance inversion of the seismic data is combined with a low-frequency model, constructed from acoustic velocity data, to approximate absolute acoustic impedance characteristics of the sequence and to model behaviour of the various sub-units across the area. These pseudo-absolute impedance data are combined with evidence from AVO analysis, to construct a 2D model of the MTD unit and to generate a synthetic seismic record, which confirms the interpretation of stratigraphic, petrophysical and lithological properties of the MTD complex.

Application of Far-Gather Seismic Attributes in Suppressing the Interference of Coal Beds in Reservoir Prediction

The sandstone reservoir of the Pinghu Formation in the Xihu Depression, East China Sea is characterized by great depth, small thickness, radical facies change and a widespread coal bed. It is difficult to describe the reservoir accurately using conventional reservoir prediction methods. In order to analyze the influence of coal-bearing strata on the prediction of the mid-low thickness sandstone reservoir, the seismic response of different sandstone–coal stratigraphic assemblages was simulated by seismic forward modeling. The modeling result indicates that the post-stack seismic response is dominated by coal bed, whereas the response of sandstone can hardly be recognized. In contrast, the difference between the pre-stack AVO (amplitude versus offset) response characteristics of coal seams and gas-bearing sandstones has been clarified based on the statistics pertaining to AVO characteristics of drilled wells. Therefore, we propose a method to reduce the interference of coal beds in sandstone reservoir prediction using far-gather seismic information. This method has significantly improved the accuracy of reservoir prediction and sand description in sand–coal coupled environments and has been applied successfully in the exploration of coal-rich strata in the Pingbei slope belt, Xihu Depression.

Inversion of low- to medium-frequency velocities and densities from AVO data using invertible neural networks

Amplitude-variation-with-offset (AVO) inversion and neural networks (NN) are widely used to invert elastic parameters. With more constraints from well-log data, neural network-based inversion may estimate elastic parameters with greater precision and resolution than traditional AVO inversion; however, neural network approaches necessitate a massive number of reliable training samples. Furthermore, because the lack of low-frequency information in seismic gathers leads to multiple solutions of the inverse problem, both inversions rely heavily on proper low-frequency initial models. To mitigate the dependence of inversions on accurate training samples and initial models, we have adopted solving inverse problems with the recently developed invertible neural networks (INNs). Unlike conventional neural networks, which address the ambiguous inverse issues directly, INNs learn definite forward modeling and use additional latent variables to increase the uniqueness of solutions. Motivated by the newly developed neural networks, we adopt an INN-based AVO inversion method, which can reliably invert low- to medium-frequency velocities and densities with randomly generated easy-to-access data sets rather than trustworthy training samples or well-prepared initial models. Tests on synthetic and field data indicate that our method is feasible, antinoise capable, and practicable.