Macro Compressed Sensing Data Acquisition

Abstract

An invention is described that relates to streamers. The streamers include one or more streamer segments. The streamer segments may have sensors, channels, and / or simulated arrays of sensors arranged along their length. At least one of the streamer segments has a variable density of sensors, channels, and / or simulated arrays along its length.

Description

[0001] The present invention relates to a streamer device, in particular, a sensor device having different densities of sensors, channels or sensor arrays along a streamer or streamer section.

[0002] The main application of the invention is seismic data acquisition, but it can also be applied to other fields such as acoustics and ultrasound.

[0003] Towed ocean streamers are used to conduct seismic surveys of the seafloor to map surface and subsurface properties. Each streamer consists of a cable that is towed behind a vessel beneath the water's surface. The streamer has a series of sensors coupled to it, which typically measure pressure differences in the water. Sensors that measure velocity or acceleration (motion sensors) may be used instead, or both types of sensors may be used together to reduce noise in the output signal, as described below. The sensors are typically evenly distributed along the streamer, and the distance between adjacent sensors may be fixed at approximately 12.5 m. This configuration results in a density of sensors that does not vary along the length of the streamer (e.g., in this example, the density of sensors would be one sensor every 12.5 m, or 0.08 m). -1 , in all directions along the streamer).

[0004] One or more seismic sources are located on the ship and periodically emit seismic waves that travel through the water surface and reflect off the seafloor or bottom. These compression waves then return to the ocean surface and are measured by sensors as a series of compressions and decompressions of adjacent water molecules, providing information about the topography and structure of the ocean surface and the sea below the ship. A single streamer can be used, or multiple streamers can be used together and towed alongside each other behind the ship to form a sensor blanket (these are called 3D surveys because of the type of underwater images that can be made with such a configuration). False signals are generated by pressure waves that have passed through the sensors once close to the streamers and are reflected back to the streamers by the ocean surface. Typically, false signals do not provide additional useful information, and it is desirable to remove them from the final signal. This can be achieved by supplementing the hydrophone sensor readings with particle motion sensors, which typically achieve a better signal-to-noise ratio than particle motion sensors. Particle motion sensors provide directional measurements and can therefore be used together with pressure readings to distinguish between waves propagating from the sea surface and the bottom. The combination of readings from both types of sensors allows for the removal of false signals.

[0005] When a streamer is being towed, the received signal is subject to high levels of additive noise due to the natural motion of the sea water and the motion of the vessel itself as well as the dragging of the streamer itself through the water. Of course, it is desirable to reduce this noise as much as possible and increase the signal-to-noise ratio.

[0006] Traditionally, both sensors and output points are evenly spaced at constant spacing along the length of a streamer or streamer segment. In this case, the distance between adjacent output points or between adjacent sensors on a streamer does not vary along the streamer. The choice of distance between output points and the number of output points required on a streamer is driven by signal sampling requirements, and the choice of distance between sensors and the number of sensors used is driven by noise sampling and attenuation requirements.

[0007] The Nyquist sampling criterion determines the minimum signal sampling rate to retain a certain amount of information from the signal, and states that to properly sample the signal, an output is required approximately every 3 meters (3.125m), corresponding to a measurement every 2ms. However, this does increase the cost and complexity of the equipment, and outputs are typically provided every 12.5 meters on the streamer. This lower sampling rate can cause aliasing, especially at higher frequencies and when the propagation direction of the compression wave is close to horizontal.

[0008] According to a first aspect of the invention, there is provided a streamer comprising at least one streamer segment, the at least one streamer segment comprising sensors, wherein a density of the sensors varies along the at least one streamer segment.

[0009] In an embodiment, the streamer comprises a plurality of output points to which data from a set of sensors is fed and the sensor density is highest around the output points with a lower sensor density between the output points.

[0010] In an embodiment, the sensor is a particle motion sensor and / or a pressure sensor.

[0011] In an embodiment, the density of sensors measured within a 5m sliding window varies along the streamer by a factor of at least 1.3.

[0012] In an embodiment, the sensor density measured within a 6.25m sliding window varies by more than 1.5, 2, or 3 times, the sensor density measured within a 12.5m sliding window varies by more than 1.5, 2, or 3 times, the sensor density measured within a 25m sliding window varies by more than 1.5, 2, or 3 times, or the sensor density measured within a 50m sliding window varies by more than 1.5, 2, or 3 times.

[0013] In an embodiment, a sensor density function is defined that describes the density of sensors in a streamer cross section centered at a particular location as a function of that location, and wherein the sensor density function has a non-uniform distribution of local maxima along the streamer.

[0014] In an embodiment, the spacing between adjacent output points along the streamer is not constant.

[0015] In an embodiment, the one or more sensors comprise an analog array of sensors.

[0016] In an embodiment, each sensor is associated with a single channel for transmitting data from the sensor or simulated sensor array to an output point.

[0017] In an embodiment, each output point is located midway between the locations of the outermost sensors in the sensor group from which it receives data.

[0018] According to a second aspect of the present invention, there is provided a streamer according to any one of the aforementioned embodiments, wherein the sensors are grouped into: a plurality of first sensor groups, each first sensor group comprising 3 to 5 sensors and being configured to feed data to associated output points, wherein the output points fed by the first sensor groups are spaced apart on average between 5 and 15 meters along the streamer; a plurality of second sensor groups, each second sensor group comprising a smaller number of sensors than the sensors in each first sensor group, the second sensor groups being located between adjacent first sensor groups along the streamer and being configured to feed data to associated output points.

[0019] In an embodiment, the first sensor groups are spaced approximately 12.5 meters apart along the streamer.

[0020] In an embodiment, the second sensor group is located midway between adjacent first sensor groups along the streamer.

[0021] In an embodiment, all output points on the streamer are configured to forward to process sensor data associated with signals having a frequency above a threshold frequency, and only output points associated with the first sensor group are configured to forward to process data associated with signals having a frequency equal to or below the threshold frequency.

[0022] In an embodiment, the threshold frequency is approximately 60 Hz.

[0023] According to a third aspect of the invention, there is provided a method for optimizing a plurality of sensors distributed along a streamer and a plurality of output points to which data from a group of sensors are fed, the method comprising: the location of the output points along the streamer; and using the determined locations of the output points, configuring sensor groups to feed data to each output point and positioning the groups along the streamer, wherein the locations of the sensor groups are determined so that the distribution of the sensors along the streamer is uneven.

[0024] In an embodiment, the location of each output point is at the center of gravity of the sensor group from which it is configured to receive data.

[0025] In an embodiment, the sensor locations are determined so that the density of sensors is highest around the output points and lowest between the output points.

[0026] In an embodiment, the step of determining the positions of the output points comprises ensuring that the output points are non-uniformly distributed along the streamer.

[0027] In an embodiment, the method comprises reconstructing a set of regularly spaced outputs along the streamer from signals at the output points using a compressed sensing algorithm, wherein the number of regularly spaced outputs is greater than the number of output points.

[0028] According to a fourth aspect of the invention, there is provided a streamer comprising one or more streamer segments, each streamer segment having a plurality of sensors mounted along it and a plurality of output points to which data from a set of sensors is input. The sensors comprise: a plurality of first sensor groups, each of which comprises 3 to 5 sensors and is configured to feed data to one of the output points, wherein the output points fed by the first sensor groups are spaced 5 to 15 meters apart on average along the streamer; a plurality of second sensor groups, each of which comprises a number of sensors less than the number of sensors in each of the first sensor groups, the second sensor groups being located between adjacent first sensor groups along the streamer, and each being configured to feed data to an output point, wherein the output points fed by the second sensor groups are also located between adjacent output points fed by the first sensors. Thus, each of the first (larger) sensor groups is larger than each of the second (smaller) sensor groups. The first sensor group may also be longer than the second sensor group and comprise a greater number of sensors, but this is not required.

[0029] In an embodiment, each output point is located midway between the locations of the outermost sensors in the sensor group from which it receives data.

[0030] In an embodiment, the first sensor groups are spaced approximately 12.5 meters apart along the streamer.

[0031] In an embodiment, the second sensor group is located midway between adjacent first sensor groups along the streamer.

[0032] In an embodiment, the sensors in the second group are a subset of the sensors in the first sensor group. This may be true for some but not all of the second sensor groups. This may be true for about half of the second sensor groups (because they are spaced closer together than the larger first sensor group).

[0033] In an embodiment, all output points on the streamer are configured to forward sensor data associated with signals having a frequency above a threshold frequency for processing, and only output points associated with the first sensor group are configured to forward data associated with signals having a frequency at or below the threshold frequency for processing.

[0034] In an embodiment, the threshold frequency is approximately 60 Hz.

[0035] In an embodiment, the density of sensors is highest around the output points and lowest between the output points.

[0036] According to a fifth aspect of the invention there is provided a streamer characterised in that it comprises one or more streamer sections comprising sensors, wherein the density of the sensors varies along at least one of the streamer sections.

[0037] In an embodiment, the sensor density is highest around the output points and the sensor density decreases between the output points.

[0038] In an embodiment, the sensor is a particle motion sensor and / or a pressure sensor.

[0039] In an embodiment, the density of at least one sensor streamer section has a sensor density that varies by a factor of at least 1.3 over at least a 5 m sliding window.

[0040] In an embodiment, the density of sensors varies by more than a factor of 1.5 or 2 or 3 over a 6.25 or 12.5 or 25 or 50 m sliding window.

[0041] In an embodiment, the local maximum of the sensor density function along at least one streamer segment has a non-uniform distribution.

[0042] In an embodiment, there is a non-constant spacing between density peaks.

[0043] According to a sixth aspect of the present invention there is provided a streamer characterised in that it comprises one or more streamer sections comprising channels, wherein the density of the channels varies along at least one of the streamer sections.

[0044] In an embodiment, the sensor density of a channel is highest around the output points, and the channel density decreases between the output points.

[0045] In an embodiment, the channel records particle motion sensors and / or pressure.

[0046] In an embodiment, the density of at least one sensor streamer section has a channel density that varies by a factor of at least 1.3 over a sliding window of at least 5 m.

[0047] In an embodiment, the density of channels varies by more than a factor of 1.5 or 2 or 3 over a 6.25 or 12.5 or 25 or 50 m sliding window.

[0048] In an embodiment, the local maxima of the channel density function along at least one streamer cross section have a non-uniform distribution.

[0049] In an embodiment, there is a non-constant spacing between density peaks.

[0050] According to a seventh aspect of the present invention, there is provided a streamer, characterised in that it comprises one or more streamer segments, said streamer segments comprising at least one sensor simulation array, wherein the simulation array centres are distributed along at least one non-uniform distribution of the sensors.

[0051] In an embodiment, at least one of the simulated array densities is highest around the output points and the density decreases between the output points.

[0052] In an embodiment, at least one of the analog array of sensors is recording a particle motion sensor and / or pressure.

[0053] In an embodiment, the density of at least one sensor streamer segment has a density that varies at least 1.3 times the density of the sensor simulation array over a sliding window of at least 5 m.

[0054] In an embodiment, the density of at least one of the sensor simulation arrays varies by more than a factor of 1.5 or 2 or 3 over a 6.25 or 12.5 or 25 or 50 m sliding window.

[0055] In an embodiment, the local maximum of the sensor array density function along at least one streamer segment has a non-uniform distribution

[0056] In an embodiment, there is a non-constant spacing between density peaks.

[0057] The present invention solves the noise and aliasing problems in this type of signal, which will be described in detail below. The present invention is applicable to all seismic data acquisition scenarios, including towed streamer, seabed, node and land seismic data acquisition, but we use the towed streamer scenario as a platform to explain these ideas in the following description. Therefore, when making arguments, towed streamer terminology (where each streamer may include multiple streamer segments, such as 100m streamer segments) will be used without affecting the generality of the final claims.

[0058] The following terms have broader meanings to those skilled in the art than will be presented. These terms are more concise, but no limitation to the standard meanings is implied or intended.

[0059] It is important to understand the concepts of input points and output points. Input points correspond to physical locations of sensors that sense or measure raw data, and output points correspond to physical locations along the streamer that provide signal estimates, which are typically provided by combining data from multiple sensors that form a group by hardware or software means. The positions of sensors within the group can be measured relative to the positions of linked output points, and the positions of the individual output points are used together with the received signals to reproduce the subsea structure by signal processing the output signals received from the known positions of the output signals along the streamer.

[0060] A streamer, which may be several kilometers long, is usually not a continuous cable, but is made by connecting several streamer segments together, for example each streamer segment may be 100m, 150m or 200m in length. A combination of streamers of different lengths may be used. This allows the length of the streamer to be changed as required and provides the ability to replace a streamer segment instead of the complete streamer in the event of a technical failure.

[0061] The term sensor as used herein refers to any device or devices that can be used to measure various properties of interest. Particle motion sensors (e.g., geophones, accelerometers, and / or rotation sensors) or pressure sensors (e.g., hydrophones) that measure seismic signals are the most common, and these are particularly suitable for use with the methods described below. However, any sensor suitable for measuring particle motion or pressure changes and suitable for mounting on a streamer may be used.

[0062] The particle motion sensor may be a sensor for measuring velocity or acceleration, such as a moving coil geophone, a MEMS accelerometer or a piezoelectric accelerometer. A hydrophone is a sensor for measuring pressure changes, for example by piezoelectric means.

[0063] An analog array is a group of hardwired sensors connected together to one electronic channel. These sensors can typically include 4 to 16 sensors, each 2m to 15m long (and can overlap). Sensor distribution can be uniform or non-uniform. The output point is usually located at the center of the array. The center of the array can be the physical center, the center of gravity, or both. For this type of array, at least partial noise attenuation is achieved by adding together the analog signals from the sensors within the array.

[0064] Where individual sensors are used rather than forming an analog array, the sensors can each be linked to its own channel and the channels of multiple sensors can provide input data to a particular output point.

[0065] Channels or electronic channels usually include amplifiers or preamplifiers and ADCs (analog-to-digital converters). They digitize the signal from the sensor or from the sensor array, so there is one channel each time the reading from the analog sensor or analog sensor array is digitized. If only one sensor is connected to a channel, this is called a single sensor recording. If multiple sensors are connected to the channel, this is called a sensor array recording. The number of channels present can correspond to the number of data points transmitted to the ship through the seismic cable. The output point itself is not equivalent to a channel. The output point corresponds to the data point recorded to the tape or provided as an output to the client. These data points can be linked to a channel that itself carries data collected from multiple sensors, or can be linked to multiple channels that each transmit measurement data collected from a single sensor. Alternatively, multiple channels, each of which is linked to more than one sensor, can be coupled to a single output point. Therefore, each output point does not have to be associated with one or only one analog sensor array or channel. Other channels carry data (which are usually digitized at this time) from the output point to the processing software. The processing software usually receives signals from many output points located at known physical locations along the streamer for reproducing the signals received from the seabed and the structure of the seabed itself.

[0066] For single sensor recordings (one channel for one sensor), the location of the channel or channel position is the same as the location of the sensor. For analog array recordings (multiple sensors combined into an array for one channel), the location of the channel is usually at the center or centroid of the array.

[0067] In some cases, noise events are locally coherent. Figure 1 An example of strong local coherent noise from a tow test is disclosed. In some data acquisition scenarios, the data may be affected by noise with relatively short coherence lengths. An example is shown in Figure 1 ,exist Figure 1 Accelerometer data acquired during a towed streamer experiment are shown in Figure 1. The left panel is in the time offset (spatial) domain and the right panel is in the frequency offset domain.

[0068] Particularly where local coherent noise is a problem, signal reproduction can be improved by more dense sensor sampling near the output point, as described below. To illustrate this, we again use the example of towed streamer seismic acquisition. Figure 2 Examples of sensors are disclosed where the sensors are spaced to provide an alias-free output of 125 Hz (or higher if filters include anti-aliasing). Triangles are used in the figures to represent the location of sensors along the streamer segment. Figure 2 , 3 , 5, and 7 use arrows to indicate the locations of output points along the streamer.

[0069] This application describes a new method for acquiring data with a higher signal-to-noise ratio (SNR) than is commonly done in the presence of locally coherent noise. The system described is particularly relevant where the number of sensors is limited, and describes how to place the sensors to reduce the adverse effects of noise and aliasing while minimizing the number of sensors required on the streamer. Part of how the invention does this is by using variable density functions of sensors, channels, and / or analog sensor arrays along the length of a streamer or streamer segment.

[0070] Two types of sensor layouts are considered, such as Figure 2 and Figure 3 The blue triangle indicates the relative position of the sensor and the red arrow indicates the output point. Figure 2 In the example, the desired output points are evenly spaced at 6.25m intervals. This is intended to provide alias-free output up to 120Hz, assuming an apparent velocity of 1500m / s for the slowest seismic signal energy in marine seismic acquisition. This shows a non-uniformly spaced sensor to avoid noise aliasing. If an anti-aliasing filter is used, alias-free output to higher frequencies can be achieved, but at the expense of some dipping events at higher frequencies.

[0071] The sensor density at a point on a streamer is the number of sensors per unit length at that point. If the number of sensors per unit length is fixed, then the sensors can be distributed in a variety of ways. If there are enough sensors available, they can be placed at regular intervals according to the Nyquist criterion, but doing so usually means high cost, weight, and complexity of the equipment due to the large number of sensors required. If the sensor spacing is fixed, then the sensor density does not vary along the streamer. For example, for a sensor spacing of 5 meters, the sensor density would be 0.2 sensors per meter (or 1 / 5m) at all locations on the streamer. -1 ).

[0072] An alternative approach is to place sensors pseudo-randomly (at non-uniform intervals) to avoid hard aliasing. A possible pseudo-random sensor distribution is Figure 2 In order to provide clean output at all desired output points, the sensors must be relatively well distributed along the streamer.

[0073] Another approach that has proven to be particularly effective in reducing noise is to concentrate (cluster) sensors close to the output point to better attenuate short coherence length noise. A possible cluster sensor distribution is Figure 3 This discloses that the sensors are spaced apart to provide a cleaner signal at the output point (but only unaliased at 60Hz - or higher if the filter includes anti-aliasing). The figure shows evenly spaced output points, spaced 12.5 meters apart. Figure 3The sensor density of the streamer section shown is now non-uniform. The number of sensors per meter (sensor density) is higher near the output point and lower (dropping to zero) away from the output point.

[0074] Since sensors are expensive and therefore very valuable, denser sampling near the output points may require a sparser distribution of the output points (the distance between adjacent output points is larger on average along the streamer). The superior noise attenuation achieved by denser sampling near the output points is at least partially achieved by dedicating more sensors to each output point (twice in this example) compared to the pseudo-random sampling strategy described above. For example, if the total number of sensors per section is fixed due to cost implications, then this means that the number of output points must be smaller. Figure 3 This is shown for output points spaced at 12.5 m intervals, which is equal to Figure 2 Half of the interval is output in the example shown.

[0075] However, if the output points are uniformly spaced, sparser output points may result in unacceptable levels of signal aliasing. Output points can generally be positioned at uniform or non-uniform intervals. If they are positioned at uniform intervals, the Nyquist criterion places an upper limit on the length of the intervals required to achieve non-aliased output data at the desired frequency.

[0076] If the number of sensors (e.g. per streamer segment) is fixed, then this defines the maximum number of sensors that can be allocated per output group. If better noise attenuation is required, a higher number of sensors per output point is required, which would reduce the number of output points since the number of sensors is fixed. With uniform output spacing, this would mean longer spacing between output points, leading to aliasing.

[0077] Figure 4 Shows Figure 2 and Figure 3 A comparison of the noise attenuation performance of the two distributions shown in Figure 1 (where the output points are uniformly distributed along the streamer so that the distance between adjacent output points is fixed). The figure shows the power spectral density of the noise versus frequency. We see that compared to different pseudo-random configurations (such as Figure 2 Compared with the clustered sensor configuration (similar to the configuration shown in Figure 3 The noise attenuation achieved is approximately 10dB higher (except at very low frequencies) than that achieved by the 3D-scanning method described in [1].

[0078] If the output points are evenly spaced, sparser output points may result in unacceptable levels of noise / signal aliasing. For example, a conventional output spacing (the distance between adjacent output points along the streamer) of 12.5m will only provide alias-free output at 60Hz. Therefore, if the output points are evenly spaced, then Figure 4The noise attenuation shown in is only possible at the expense of allowing aliasing of the output signal.

[0079] As mentioned above, denser sampling around an output point has been shown to be beneficial for attenuating short coherent noise at that output point. This requires a given minimum number of sensors per output point (to achieve the desired level of noise attenuation), but there is usually also a maximum number of sensors achievable per seismic segment (this is usually dictated by cost considerations). The result is that there must be fewer output points than are required to adequately sample the signal above a certain wavenumber according to classical sampling theory. In particular, in the case of uniform output points, some signal wavenumbers will be aliased and cannot be properly recovered. Thus, there seems to be a dilemma in which we must allow for more noise or more aliased signal than expected.

[0080] The solution to this is to position the output points at uneven intervals to prevent hard aliasing of the signal. The non-uniformly sampled output signals can be regularized (interpolated onto a regular grid) using one of several compressed sensing methods. Otherwise, they can be used directly, for example for seismic imaging, using their known irregular (non-uniform) locations.

[0081] A sensor density function can be defined to describe the density of sensors in a streamer segment centered at a particular location as a function of the location of that streamer segment. The density function is highest where sensors are closely spaced (sensor density is high in streamer segments where sensors are closely spaced) and lowest where sensors are sparsely distributed. In the case where sensors are clustered in groups, each group will represent a local maximum in the density function. These local maxima do not have to be of the same size, but will be considered a local maximum if the density is higher than at neighboring locations. The global maximum of the density function for the streamer will represent the highest of these local maxima.

[0082] Using a non-uniform distribution of output points along the streamer and different densities of sensors along the streamer, particularly where this corresponds to a higher density of sensors at or near the output points, can reduce noise and aliasing, so a clean signal can be obtained without increasing the number of sensors required. Embodiments of the invention will now be described, by way of example only, with reference to the following drawings, in which:

[0083] Figure 1 The strong local coherent noise from the towing test is plotted.

[0084] Figure 2 An example of sensors spaced apart to provide an alias-free output of 125 Hz (or higher if the filter incorporates anti-aliasing) is shown.

[0085] Figure 3An example of sensors spaced apart to provide a cleaner signal at the output point (but unaliased only at 60Hz; higher if the filter incorporates anti-aliasing) is shown.

[0086] Figure 4 Test results for sensor placement and filter design for evenly spaced output points are shown.

[0087] Figure 5 An example of non-uniformly spaced output points is shown for avoiding aliasing while allowing good noise attenuation by including sensors clustered near the output points.

[0088] Figure 6 A flow chart showing a method for reproducing a signal whose output points are non-uniformly distributed.

[0089] Figure 7 Examples are disclosed where more than one output per local sensor cluster is used.

[0090] Figure 8 Illustrative examples of arrays that are non-uniformly distributed along a streamer are provided.

[0091] Fig. 9 Results for a synthetic example using SEAM finite-difference data are shown.

[0092] Fig.10 Schematic examples of sensors distributed in groups of different sizes along a streamer are shown. The top row evenly distributes sensor groups along the streamer segment, and the bottom row distributes the same groups non-evenly.

[0093] Fig.11 A method for reproducing a signal is illustrated in which different sets of sensor groups are used to account for different frequency ranges in the signal.

[0094] Fig.12 A method for reproducing a signal is illustrated in which different sets of sensor clusters are used to account for different frequency ranges in the signal and the clusters are non-uniformly distributed along the streamer segment.

[0095] In some embodiments, the present invention uses non-uniform spacing of input points (sensor locations) and output points to obtain high quality (high SNR) estimates of the signal from noise-affected measurements.

[0096] As mentioned previously, there is a trade-off between allowing more noise than desired or allowing more aliased signals. Non-uniform output sampling can address both the signal and noise sampling issues. The aliasing issue can be at least partially overcome by relaxing the requirement that the spacing of the output points should be regular, and by allowing non-uniformly spaced output points to enable compressed sensing (CS).

[0097] Figure 5One possible configuration of sensors (input points) and output points in an embodiment is shown. Non-uniformly spaced output points (average spacing greater than 12.5m) are included to avoid aliasing, while the device allows good noise attenuation by using sensors clustered near the output points. The sensors are clustered by output and are not uniformly distributed. Non-uniform distribution means that the distance between adjacent sensors or between adjacent output points is not fixed along the streamer. For example, for an output point located between two adjacent output points, the distance to the nearest output point in one direction along the streamer may be less than the distance to the nearest output point in another direction along the streamer. If the distribution of sensors is also selected to be non-uniform, then the same is true for sensors located between two other sensors on the streamer. The output points can be pseudo-randomly spaced along the streamer segment or streamer.

[0098] If 6.25m regular spacing is chosen, this example contains fewer output points per unit length of streamer, but produces cleaner signals (i.e., produces output data with higher SNR) than 16 output points per 100m streamer segment. Cleaner output data is more suitable for compressed sensing solutions, including regularization to a regular grid using state-of-the-art CS methods. Figure 6 A flow chart showing a method of reproducing a signal using non-uniformly distributed output points. A compressed sensing algorithm is used to account for the fact that the output points are non-uniformly distributed by reconstructing regularly spaced outputs along a streamer using the non-uniformly spaced outputs at the output points. The number of reconstructed outputs will typically be greater than if the output points were physically spaced apart along the streamer.

[0099] Thus, the above example allows for various noise attenuation methods to be used to produce cleaner output traces. As a non-limiting example, assume that each streamer segment (perhaps 100m long) has 56 sensors (input points). If there are 16 evenly sampled output points as shown in the above example, there will be 3.5 sensors per output point, and these sensors must be fairly well (non-uniformly) distributed along the streamer to ensure that each of the 16 output points gives a reasonably clean output. Conversely, if there are fewer output points that require attention (but the spacing is uneven), then noise attenuation methods can provide cleaner results because there will be more sensors available to provide data to each output point (or the device can use fewer sensors). For example, if we find that 8 non-uniformly sampled output points per streamer segment are sufficient, then there will be 7 sensors per output point, and if we only need 5 output points, then 11 sensors will be required per output point.

[0100] A compressed sensing recovery algorithm will be used to reconstruct the signal using cleaner output data. Examples include: IAA-MP and BPDN. Other examples will be apparent to those skilled in the art.

[0101] When implementing an embodiment using a non-uniform distribution of sensors and output points, it is recommended to first optimize the non-uniform output point positions. Assuming a clean (signal-only) output, one can search for the best non-uniform output positions that allow for optimal reconstruction. Next is the step of optimizing the input (sensor) positions to match the output points.

[0102] Adjusting any of our current sensor optimization techniques, we will then optimize the sensor locations to give us the cleanest output at the selected location. This may involve clustering sensor locations around an output point. The sensors within a swarm associated with a particular output point may be evenly distributed (e.g., a swarm of 5 sensors may be placed at 0.2 meter spacing, with the output point equidistant between the two outermost sensors). This will still result in an uneven or non-uniform distribution of sensors along the streamer, as the distance to the adjacent sensor swarm associated with the next output point along the streamer will be different than the spacing between sensors within the swarm (i.e., it will not be 0.2 meters in this example).

[0103] Of course, in some embodiments, sensors within a group coupled to a single output point may also be distributed non-uniformly. Non-uniform distribution of sensors or output points may, but does not always, mean that all distances between adjacent sensors or adjacent output points on a particular streamer segment are different from each other.

[0104] However, the distribution of sensors or output points may be considered non-uniform as long as at least some of the distances between adjacent sensors or adjacent output points on a streamer segment are different.

[0105] This two-stage optimization can also be coupled to a single stage, but this will require more expensive optimization work.

[0106] In some embodiments, it is possible (and advantageous) for each sensor cluster to produce more than one output.

[0107] We can design noise attenuation methods to provide more than one point of output for a given sensor cluster. This can facilitate reconstruction; for example, by defining a complementary finite difference signal. An example of such a configuration is shown in Figure 7 In this example the average interval is greater than or equal to 12.5.

[0108] Figure 8 A schematic example of an array that is non-uniformly distributed along a streamer is disclosed. Each triangle represents a sensor, and a group of sensors connected together represents a sensor array. Note that the sensor array itself is distributed in a non-uniform manner along the streamer segment. A single channel will carry data from all sensors within the array to an output point. Although Figure 8The sensor arrays are shown as being equally spaced along or within each sensor array, but this is for illustration purposes only. The sensors may also be non-uniformly distributed. The output point locations may correspond to the physical center and / or center of gravity of each array.

[0109] Fig. 9 Numerical examples are shown. The data represented by the curves in the figure were created using a finite difference code based on the SEAM model. The red box in the top panel shows the portion of the data used for the test. The left panel of the middle row shows the fk transform of the test data; the right panel shows the aliasing that would appear if the data were sampled regularly at a fixed interval of 12.5. The bottom row shows the test results, comparing a finely sampled reference with a reconstruction performed using the IAA-MP algorithm in a single tx-window.

[0110] If the sampling is uneven, the clean signal can be reconstructed (regularized) onto a regular grid. This signal has frequency components up to 125Hz, so if sampled regularly at 12.5m, there is obvious aliasing. Figure 7 The second panel in the bottom row shows an example where 8 output points are non-uniformly distributed over 100m with an average spacing of 12.5m. The data is then regularized to a fine grid for comparison with the input. The reconstruction error using the IAA-MP algorithm is less than 5% RMS.

[0111] The density of at least one streamer segment has a density of sensors, channels, and / or sensor arrays that varies by at least a factor of 1.3 over a sliding window of at least 5 m in length. For example, the sensor density within a 6.25 m sliding window may vary by more than a factor of 1.5, 2, or 3 along the streamer. The sensor density within a 12.5 m sliding window may vary by more than a factor of 1.5, 2, or 3 along the streamer. The sensor density within a 25 m sliding window may vary by more than a factor of 1.5, 2, or 3 along the streamer, or the sensor density within a 50 m sliding window may vary by more than a factor of 1.5, 2, or 3 along the streamer.

[0112] A sliding window refers to a movable streamer segment or portion in which the number of sensors can be counted to determine a measurement of sensor density. The number of sensors counted will of course depend on the size of this window. The measurement of sensor density at the streamer segment within the window can then be determined by dividing the counted number of sensors by the window size. By sliding the window along a segment of streamer (moving the position of the streamer segment being inspected), multiple measurements representing the sensor density at the location of each streamer segment can be determined. These locations may or may not overlap. For example, a statement that the sensor density within a 5m sliding window varies by more than 1.5 times indicates that there will be locations on the streamer where the window can be positioned to provide a minimum density measurement and locations where the window can be positioned to provide a maximum density measurement, and that the maximum density measurement will be more than 1.5 times the minimum density measurement.

[0113] In some cases, the density measured at certain locations in the window may be zero. In one example of a streamer, the sensors are distributed in groups, each covering a length of about 3m on the streamer. These 3m long sensor groups are typically each connected to an output point located in the center of the group (about 1.5m from each outermost sensor), but may also be connected to multiple output points. The output points are spaced about 12.5m apart along the streamer segment.

[0114] The output points may not be spaced uniformly, so that the output points are spaced an average of 12.5m apart along the streamer segment. In this case, the average gap between sensor clusters is about 9.5m. When the 5m window is above this gap, there will be no sensors within the window and the density will be measured as zero. In this case, the ratio of the highest density measured in a sliding window on the streamer (which can be a 1m, 2.5m, 5m, 6.25m, 12.5m, or 25m sliding window) to the lowest non-zero density measured in a sliding window of the same size (e.g., the window contains 1 sensor) will exceed 1.5, 2, or 3.

[0115] If a sliding window at a point on the streamer measures zero density, then the ratio between the highest density measured in a 5m sliding window and the lowest non-zero density measured in a 1m, 2m, 5m, 6.35m, 12.5m, or 25m sliding window will exceed 1.5, 2, 3, 5, or 10. Any combination of these window sizes and ratios is intended to be covered herein.

[0116] For a given length of streamer, how many output points need to be present to achieve a given signal quality will depend on the frequency of the signal being measured. For higher frequency signals, the output points will need to be spaced closer together (the density of output points will be higher) in order to achieve a good reproduction of the signal at the processing stage. This is because too low a sampling rate may not be able to distinguish between two signals of different frequencies. For higher frequency signals, a higher sampling rate is generally required. However, in the case of seismic reflections at the streamer sensors, the noise is higher at lower frequencies (see Figure 4 ). Therefore, for lower frequencies, a lower density of output points is required, but the number of sensors required to feed the signal to each output point needs to be greater to offset the effects of the higher noise level in the signal.

[0117] To accommodate both high and low frequencies in the signal while minimizing the number of sensors required for a particular streamer segment, the sensors can be Fig.10 The output points (and sensor clusters) are evenly distributed along the streamer in the upper row and non-evenly distributed in the second row. In the example shown, the sensors are positioned in clusters, each cluster is centered around an output point, and the sensors are clustered around their associated output points as described above.

[0118] Clusters of 5 sensors (green / black triangles) are placed every 12.5 meters along the streamer. Smaller clusters of 3 sensors (pink / light triangles) are placed every 6.25 meters. Some (3 in this case) of the sensors in each larger cluster also form one of the smaller clusters of sensors. Some or all of the sensors in the cluster of 5 sensors are thus used to provide output for the lower frequency signal as well as the higher frequency signal. Both frequencies meet the requirements in terms of noise reduction and sampling rate, but fewer sensors are used than in the case of clusters of 5 sensors placed every 6.25 meters along the streamer.

[0119] Of course, the exact distances given are examples only, and the distances between groups and the number of sensors in each group may be optimized for specific situations or where specific types of signals are to be detected. Groups of 3 to 12, preferably 4 to 10, more preferably 4 to 6, most preferably 5 sensors may be provided along the streamer every 10 to 25 meters, preferably every 10 to 20 meters, more preferably every 11 to 15 meters, most preferably every 12.5 meters, and smaller groups of 1 to 5, preferably 2 to 4, most preferably 3 sensors may be provided at shorter intervals, for example every 2 to 15 meters, preferably every 4 to 8 meters, most preferably every 6.25 meters.

[0120] The distances between clusters may be averaged so that the distribution of output points along the streamer segment and the distribution of sensors along the streamer segment are non-uniform to reduce the effects of noise and aliasing as described above (e.g. Fig.10If a non-uniform distribution of output points is used, then fewer points may be needed along the streamer and the average distance between larger and smaller groups may be able to be increased. A particular pattern of groups may be repeated multiple times along each streamer segment, and multiple such streamer segments may be connected to form a longer streamer.

[0121] To account for higher frequencies, additional single sensors may be included between clusters so that an output point is placed every 3.125 meters, or, if a non-uniform distribution of clusters and / or output points is used, so that an output point is placed every 3.125 meters on average. These sensors would be located between the clusters of 5 sensors and the clusters of 3 sensors. If a non-uniform cluster distribution is used, each pair of adjacent large and small sensor clusters would have a single sensor halfway or halfway on average.

[0122] The following table summarizes the estimated values ​​required for each frequency range. This table is particularly relevant when using particle motion sensors, but may be relevant in other situations as well.

[0123] Although it is preferred that the distribution of sensors and output points is non-uniform along the streamer segment, this is not necessary to cater the frequency to the use of different groups. For example, sensors can be located at fixed spacing distances along the length of the streamer, and the output points for different frequency measurements are simply linked to specific groups of these sensors, thereby meeting the requirements for the number of sensors and sampling frequency.

[0124]

[0125] Fig.11 A flow chart of a method for reproducing a signal is shown, which includes two sets of sensor groups to account for low and high frequencies in the signal. Fig.11 In the method shown, the groups are evenly distributed along the streamer (which generally means that the output points are also evenly distributed, as these are generally located at the center of each group of sensors). The first group of sensors is specifically designed to attenuate noise in the first frequency band, while the second group of sensors is specifically designed to attenuate noise in the higher second frequency band. The sensors are spaced along the streamer so that the second group is located between the first groups. Some sensors will belong to both groups at the same time. The noise-attenuated output in the first frequency band is obtained from the sensors in the first group and spatially interpolated to a smaller spacing to be related to the spacing of the second group output points. The noise-attenuated output in the second frequency band can be collected for the output points at the smaller spacing of the two groups (or sensors that only form part of the second group). These outputs for the two frequency bands are combined, resulting in outputs for the two frequency bands at the smaller spacing of the second group. When obtaining the signal of the first frequency band, only the signal from the larger first group can be used (because this can avoid using the noise signal from the second group with fewer sensors). In some cases, the signals from both groups can be used simultaneously.

[0126] Fig.12 is a flow chart of a similar method for signal reproduction. Fig.12 The method shown involves non-uniformly spaced sensor clusters along the streamer. In this case, the output points linked to the first cluster and the second cluster are non-uniformly spaced. This method is similar to Fig.11 method, but it must be used when processing signals from two groups of sensors Fig.10 Method to create a set of noise-attenuated regularly spaced outputs from non-uniformly distributed output points.

[0127] So far, we have only discussed operations in the common source setting. We can also consider the possibility of including non-uniformly sampled source points into the high-dimensional reconstruction problem. It is well known that signal representations in high dimensions are sparser than those in low dimensions, thus improving the quality of reconstruction using CS techniques. This in turn can allow for a sparser sampling of the output points in the first place.

[0128] The non-uniform sampling source points mentioned here can be considered as the coaxial direction (the general direction of the streamer) and the cross direction. Therefore, the coaxial (in-line) direction refers to the direction along the longitudinal direction or parallel to the longest side of the streamer. The cross-line direction refers to the direction perpendicular to the coaxial direction, or the direction generally parallel to the shortest side of the streamer. Sensors and / or output points may be non-uniformly distributed along the coaxial direction of the streamer and along the cross direction of the streamer, or may have a varying density. In addition, streamers with uneven spacing in the cross direction can also be considered, which increases the sparsity of the signal in the high-dimensional signal reconstruction method using the compressed sensing method. The same non-uniform distribution of output points and sensors can be applied to the cross direction and or instead of the coaxial direction. In order to extend the sensor coverage in the cross direction and along each streamer, several streamers need to be towed side by side away from the rear of the ship, or the streamer section needs to have a certain thickness to allow the sensor to be installed on a specific streamer in a 2D rather than a 1D configuration.

[0129] In one embodiment, the sensor density is highest around the output point, and the sensor density decreases between the output points. In another embodiment, there is no clustering around the output point. In these examples, the distribution of sensors is uneven.

[0130] Fewer output points (tracks) will enable cheaper compressed sensing algorithms. Although these methods are usually computationally expensive; their use will be more affordable given the small number of output tracks that need to be processed compared to the number of input tracks.

[0131] We use the word cluster in a general sense, since there can be sensors between these locations. This approach can find applications in streamer, seabed, node, borehole, and land seismic data acquisition scenarios.

[0132] Typically, a streamer will include multiple streamer segments. In the case of only one streamer segment, the streamer segment can be referred to as a streamer.

[0133] The arrays (if used) may all be of the same type or vary by location on the streamer.

[0134] The non-uniform distribution of sensors discussed above can be applied to channels and simulated arrays of sensors. Channels, sensors, simulated arrays of sensors can be collectively referred to as measurement volumes. For the present invention, any measurement volume of description can be used.

Claims

1. A streamer comprising one or more streamer segments, each streamer segment having a plurality of sensors mounted along it and a plurality of output points, data from a set of sensors being input to one output point, the plurality of sensors comprising: a plurality of first sensor groups, each first sensor group configured to feed data to one of the plurality of output points, wherein the output points fed by the plurality of first sensor groups are spaced apart along the streamer segment; a plurality of second sensor groups, each second sensor group having fewer sensors than each first sensor group, the plurality of second sensor groups being located between adjacent first sensor groups along the streamer segment and each second sensor group being configured to feed data to an output point, wherein the output points for the first sensor group and the second sensor group are configured to forward to process sensor data associated with signals having a frequency higher than a threshold frequency and only the output points associated with the first sensor group are configured to forward to process data associated with signals having a frequency equal to or less than the threshold frequency, the first sensor group being designed to attenuate noise in a first frequency band, and the second sensor group being designed to attenuate noise in a second frequency band higher than the first frequency band.

2. The streamer according to claim 1, wherein: Forwarding the data for processing includes forwarding sensor data containing signal and noise.

3. The streamer according to claim 1 or 2, wherein: The output points for the first sensor groups are evenly spaced apart along the streamer by a distance d, and the output points for the second sensor groups are positioned midway along the streamer between adjacent first sensor groups.

4. The streamer according to claim 1 or 2, wherein: The output points for each of the first sensor group and the second sensor group are unevenly spaced along the streamer.

5. The streamer according to claim 1 or 2, wherein: The first sensor groups are spaced 10 to 15 meters apart along the streamer.

6. The streamer according to claim 1 or 2, wherein: The first sensor groups each include 3 to 7 sensors.

7. The streamer according to claim 1 or 2, wherein: Each output point is located midway between the locations of the outermost sensors in the sensor group to which it transmits data.

8. The streamer according to claim 1 or 2, wherein: The threshold frequency is approximately 60 Hz.

9. The streamer according to claim 1 or 2, wherein: The first sensor group and the second sensor group are formed by hardware or software means.

10. The streamer according to claim 1 or 2, wherein: The sensors in the second sensor group are a subset of the sensors in the first sensor group.

11. The streamer according to claim 1 or 2, wherein: The sensors in the first sensor group and the second sensor group are particle motion sensors or pressure sensors.

12. The streamer according to claim 1 or 2, wherein: The sensors within the first sensor group and the second sensor group are evenly or unevenly spaced.

13. The streamer according to claim 1 or 2, wherein: The streamer segment includes an additional individual sensor located midway between the centers of the first sensor group and the second sensor group.

14. A method of performing seismic surveys of the seafloor using towed ocean streamers, each streamer comprising: One or more streamer segments, each streamer segment having a plurality of sensors mounted along the streamer segment and a plurality of output points, data from a set of sensors being input to one output point, the plurality of sensors comprising: a plurality of first sensor groups, each first sensor group configured to feed data to one of the plurality of output points, wherein the output points fed by the plurality of first sensor groups are spaced apart along the streamer; a plurality of second sensor groups, each second sensor group having fewer sensors than each first sensor group, the plurality of second sensor groups being located between adjacent first sensor groups along the streamer and each second sensor group being configured to feed data to an output point, wherein the output points for the first sensor group and the second sensor group are configured to forward to process sensor data associated with signals having a frequency above a threshold frequency and the output points associated with only the first sensor group are configured to forward to process data associated with signals having a frequency equal to or less than the threshold frequency, Wherein, the method comprises: acquiring data in a first frequency band only from output points of the first sensor group; obtaining data in a second frequency band higher than the first frequency band from output points of the first sensor group and the second sensor group, The first sensor group is designed to attenuate noise in the first frequency band, and the second sensor group is designed to attenuate noise in the second frequency band which is higher than the first frequency band.

15. The method according to claim 14, wherein: Noise attenuated outputs in the first frequency band are obtained from sensors in the first sensor group, and noise attenuated outputs in the second frequency band are collected by sensors in the second sensor group.

16. The method according to claim 14 or 15, wherein: The output points of the first sensor groups are evenly spaced apart along the streamer by a distance d, and the output points of the second sensor groups are positioned midway along the streamer between adjacent first sensor groups.

17. The method according to claim 15, wherein: The noise attenuated outputs in the first frequency band are spatially interpolated to a smaller spacing to correlate with the spacing of the output points of the second sensor group.

18. The method according to claim 17, wherein: The outputs for the first frequency band and the second frequency band are combined, resulting in outputs at smaller spacings for the second sensor group.

19. The method according to claim 15, wherein: Output points associated with the first sensor group and the second sensor group are unevenly spaced along the streamer.

20. The method according to claim 19, comprising: A set of regularly spaced outputs is reconstructed along the streamer from the signals at the output points using a compressed sensing algorithm, wherein the number of the set of regularly spaced outputs is greater than the number of output points.

21. The method according to claim 20, comprising: Using interpolation to produce the same spatial sampling in the uniformly spaced first set of outputs as in the uniformly spaced second set of outputs; and The evenly spaced first set of outputs and the evenly spaced second set of outputs are combined.

22. A method as claimed in claim 14 or 15, comprising using an additional individual sensor located between the centres of the first and second sensor groups to attenuate noise in a third higher frequency band.

Images