Mapping soil properties

Abstract

An apparatus for and a method (1) of mapping a property of a soil structure (113) is described. A seismic wave (115) is detected (3) after it has passed through the soil structure (113). The detected wave is used to map a property (5) of the soil structure (113). The seismic wave (113) has a frequency that is sufficiently high to resolve features that are less than 1 m apart from each other. Thus, the present disclosure uses seismic techniques that can resolve and map soil properties at scales below 1m.

Description

MAPPING SOIL PROPERTIES

[0001] The present disclosure concerns methods and apparatus for mapping a property of a soil structure, more particularly, methods and apparatus that use seismic waves to map a property of a soil structure.Background

[0002] Soil structure is important for agriculture, and especially for sustainable agriculture, supporting terrestrial ecosystems (including soil ecosystems), climate regulation and ground water recharge. Soil structures form as layers bounded by interfaces or surfaces, and the soil structures can be heterogeneous and multi-scale. As soils are complex and biologically active ecosystems, their properties can vary significantly on relatively short length scales and timescales. Understanding shallow soil and topsoil, which forms the uppermost layer or layers is particularly important for agricultural purposes. Degradation of soil structure can occur on short timescales and may lead to adverse environmental consequences such as increased greenhouse gas emissions, reduced crop productivity and landslides, whilst soil compaction, resulting for example from the use of vehicles or tillage on mechanically sensitive soil, adversely affects soil in terms of its habitability for plants and animals, reduced porosity and hydrological function, and increased mechanical impedance.

[0003] Comprehensive understanding of soil structures requires high resolution measurement and mapping methods, in order to determine the soil composition, moisture content, bulk density, porosity, pore connectivity, carbon content, the volume of soil in different layers or horizons, horizon depth, the locations and depth of compaction, amongst other properties.

[0004] Known methods of characterising soil structures often rely on combining point sampling measurements, which are often obtained using invasive methods, such as drilling. Point sampling methods can be extrapolated using estimated, statistical, or assumed bulk property values. The density of soil within a horizon will typically be estimated, for example using pedotransfer functions or extrapolated from point samples. However, as soil structures are typically complex and heterogeneous, extrapolated data may be inaccurate or give an incomplete picture. Given the importance of soil structures from an agricultural, nutritional and ecological perspective, there is a need to develop efficient, non-invasive and comprehensivetechniques for mapping and characterising soil structures at scales that can help agricultural and land management decision making.

[0005] The present disclosure seeks to provide an efficient, scalable, non-invasive and high-resolution method of mapping a property of a soil structure. Additionally, or alternatively, the present disclosure seeks to provide an improved method and apparatus for mapping a property of a soil structure.Summary

[0006] In a first aspect, the present disclosure provides a method of mapping a property of a soil structure according to claim 1. Optional but preferred features are set out in the dependent claims.

[0007] In a second aspect, the present disclosure provides a method of determining soil health according to claim 15.

[0008] In a third aspect, the present disclosure provides a method of improving soil health according to claim 17.

[0009] In a fourth aspect, the present disclosure provides apparatus for mapping a property of a soil structure, according to claim 18.

[0010] It will of course be appreciated that features described in relation to one aspect of the present disclosure may be incorporated into other aspects of the present disclosure. For example, methods of the disclosure may incorporate any of the features described with reference to the apparatus of the disclosure and vice versa.Description of the Drawings

[0011] Example embodiments will now be described, by way of example only, with reference to the accompanying schematic drawings, of which:FIG. 1 shows the steps of a method of mapping a property of a soil structure in accordance with an embodiment of the present disclosure;FIG. 2 shows the steps of a method of mapping a property of a soil structure in accordance with another embodiment of the present disclosure; andFIG. 3 shows apparatus for mapping properties of a soil structure, according to another embodiment of the present disclosureDetailed Description

[0012] According to a first aspect, the present disclosure provides a method of mapping a property of a soil structure. A seismic wave is detected after it has passed through the soil structure. The detected wave is used to map a property of the soil structure. The seismic wave has a frequency that is sufficiently high to resolve features that are less than 1 m apart from each other. Thus, the present disclosure uses seismic techniques that can resolve and map soil properties at scales below Im.

[0013] The skilled person would understand that, in the context of the present disclosure, mapping soil properties comprises mapping spatial, and optionally temporal, co-ordinates in the soil to values of one or more properties of the soil at those co-ordinates. The map of soil properties may be an abstract array. The array may be displayed as an image or other graphical representation.

[0014] Seismic waves having frequencies below 50 Hz are referred to herein as low- frequency seismic waves. Seismic waves having frequencies of between 50 Hz and 200 Hz are referred to herein as high frequency waves, and seismic waving having frequencies between 200 and 3000 Hz are referred to herein as ultra-high frequency waves. For seismic waves having a continuous spread of frequencies, for example broadband waves, references herein to a frequency of the waves is a reference to their centre frequency.

[0015] The use of seismic waves for mapping subsurface regions is widespread in geotechnical and engineering applications. These known applications make use of low frequency seismic waves to map large scale features down to depths of tens or hundreds of meters, to resolve features and objects with length scales of larger than or around Im to 10 m. Detected seismic waves may be used to form seismic images, mapping large scale geological features such as rock formations, groundwater tables, potential hydrocarbon deposits, subduction zones, fault regions, and mantle plumes.

[0016] If soil structure is considered in such methods, it is typically considered as a bulk conglomerate including particles of different sizes. In contrast, methods of the present disclosure determine the properties of the soil structure over significantly smaller length scales, or higher resolution, such that features that are less than 1 m apart can be distinguished (which is the scale of the living soil ecosystem). The present disclosure therefore provides higher resolution mapping methods than geotechnical applications.

[0017] In the present methods, the seismic wave may have a frequency between 50 and 3000 Hz, i.e., the seismic wave may be a high-frequency or ultra-high frequency seismic wave. The seismic wave may have a frequency between 200 and 3000 Hz, i.e, the seismic wave may be an ultra-high frequency seismic wave. For example, the seismic wave may have a frequency of more than 400 Hz (which is the highest frequency typically recorded by conventional geophones). Using high or ultra-high frequency seismic waves enables properties of a soil structure to be mapped with a higher resolution, hence providing mapping of properties over smaller length scales. The resolution that can be achieved when using a seismic wave to map a property of a soil structure is typically comparable to the wavelength of that seismic wave. The wavelength of the seismic wave is inversely proportional to the frequency of the seismic wave. The wavelength is proportional to the seismic wave speed of the soil structure, for example a doubled P-wave speed results in a doubled wavelength for the P-wave. Methods of the present disclosure may comprise pre-determining a seismic frequency and wavelength that is comparable to the length scale of a soil structure property of interest, and using a seismic wave source to generate a seismic wave having that frequency and wavelength. Of course, the generated seismic wave may form part of a set of seismic waves of different frequencies and wavelengths that are generated simultaneously, i.e. the source may be a broadband source. Generated seismic waves often have a spectrum that is approximately Gaussian in shape. Methods may comprise estimating the size of a soil structure of interest, and determining a wavelength that is appropriate to map a property of that structure. The wavelength may be less than twice the size of the structure of interest. The method may comprise using an estimate of the seismic wave speed and the determined wavelength to calculate the frequency of the seismic wave required to map a property of the soil structure of interest.

[0018] Methods according to the present disclosure seek to map properties of soil structures on length scales of less than 1 m, to depths of around 1-2 m from a surface of the soil structure, and this typically requires high or ultra-high frequency seismic waves. Methods according to the present disclosure map properties of soil with a resolution of better than 1.0 m, i.e. points in the soil structure that are closer together than 1.0 m can have different values of the property, and spatially extended features that are less than 1.0 m apart can be distinguished, i.e. they do not appear to overlap inthe map. Methods according to the present disclosure may map properties of soil with a resolution of better than 0.5 m, better than 0.2 m or better than 0.1 m.

[0019] The detected seismic wave will be a component of a seismic wavefield. It may be a seismic wave related to acoustic waves, for example a body wave, a direct wave, a reflected wave, a transmitted wave, or a refracted wave; it may be a P wave (for example a direct P wave, i.e., a compressional wave arriving without reflection or refraction), an S wave, a converted wave (for example P-to-S converted), or a surface or other interface-related wave (e.g. a Rayleigh wave or a Love wave, or a Stoneley wave), free oscillations, trapped wave, environmental noise vibrations, a coda wave, or a wave that has propagated through a porous medium.

[0020] The seismic wave may be generated at or near to a surface of the soil structure. The surface may be an upper surface of the soil structure, for example at ground level. The seismic wave may be generated within the soil structure, for example from within a hole or an exposed structure.

[0021] The method may comprise measuring a characteristic of the detected seismic wave, for example travel time, wavespeed, phase velocity, waveform, wave amplitude, or a cross-correlation or autocorrelation of the wave, and using the characteristic to infer a property of the soil structure. The method may comprise using the detected seismic wave to map a first property of the soil structure, for example, to map an elastic modulus of the soil structure, and inferring a second property of the soil structure from the first property.

[0022] The method may comprise using an inference method to map a property of the soil structure from the detected wave. The method may comprise using an inference method to determine the property of the soil structure from the detected wave.

[0023] Inference methods generally make use of an initial model (such as soil structure model) and observed data (such as seismic data), and update the initial model to an updated model that is constrained by the observed data. Methods of updating the model include inversion methods, and forward modelling wherein synthetic data are compared to the observed data.

[0024] Inversion methods derive models from observed data by solving the equations of a physical system having a given parameter set, i.e. by inverting the forward problem. In many cases, closed form expressions of the inverse solutions do not exist, iterative and thus approximate solutions are used instead. Inverse problems are often non-uniqueand nonlinear. Seismic inversion may involve using travel times to infer on the seismic wavespeeds of substrate interfaces with their spatial undulations. Subclasses of inversion common in seismology include imaging, deconvolution, b ackprojection.

[0025] Forward modelling requires using either analytical, numerical, machinelearning or other mathematical tools to simulate synthetic data for a system model, given an initial model. For seismic waves, the initial model may be wavespeeds, density and source location, the system model may be a numerical solution for linear elastodynamic partial differential equations of motion, and the outcome synthetic seismic wavefields or waves. The forward modelling procedure seeks to mimic natural processes. Forward problems can be linear or nonlinear. Typically seismic wave propagation is assumed to obey linearity.

[0026] In the context of the present disclosure, inference methods may involve updating a spatial or temporal model of a soil structure, constrained by data relating to the detected seismic wave(s). This is in contrast to a point-sampling method, in which point samples of the soil are taken and the distribution of soil properties is estimated by interpolation or extrapolation, as discussed above.

[0027] The inference method may solve an inverse problem to determine the model or approximate a solution to the inverse problem using a numerical method or machine learning. The inverse problem may for example be inverting the continuous equation d(x_0)=G(m), with data vector d along observation points x_0, process G (e.g. wave propagation or ray tracing), and model m(x), depending on spatial sampling points x. Inverse problems are mostly solved in a discrete world, such that d i = G_ij mJ , where index i indicates the spatial points where data are collected (for example along the surface), and j are the discrete spatial points in the medium where the soil property is determined, and G ij is the operator that relates the discrete model vector m to the discrete data vector d, encapsulating the physical process. In a simplified example raytracing method, the data d i are times t i taken for a detected seismic wave to travel between the seismic wave source and the seismic wave detector, the model is the wavespeed distribution vj across the domain, the process is ray tracing, with each ray path path i satisfying vj(xj) = path i(xj) / t i, where path i is the path of the i-th ray passing through model sampling points xj. Other inverse problem solutions may be found by refraction imaging, surface wave methods, full-waveform inversion, adjoint tomography, 2D or 3D body wave tomography, noise tomography, seismic imaging,machine-learning-based inversion, or probabilistic solutions based on Bayes Theorem (Bayesian inference), or any combination of those.

[0028] Solving the inverse problem may include the step of constraining the model to take account of structures that have a dimension larger than Im.

[0029] Alternatively, the inference method may comprise a forward problem method. The forward problem approach may comprise constructing a plurality of models of the soil property, calculating the seismic wave results generated by each of those models, and selecting the model that produces the results that best fit the measured data being as representing the soil structure or evolution. The inverse problem may be solved, for example, using statistical inference (that is, a method in which statistical methods are used to infer the model of the soil property), for example Bayesian inference, or deterministic methods for example optimisation based on linear or nonlinear solutions to the inverse problem.

[0030] The method may include mapping the property using, in combination with the detected seismic wave, further data relating to the soil structure, selected from the following list: geophysical data (for example radar data, electromagnetic data, e.g. electrical resistivity), conventional soil-sampling data (for example penetrologger data, data from chemical analysis, eDNA data, data relating to above-ground biodiversity, data relating to below-ground biodiversity), remote sensing data (for example data from a satellite or drone), data relating to local knowledge (e.g. data received from a farmer of the soil structure, data relating to previous management or use of the soil structure).

[0031] The mapped property may for example be combined with the further data using geospatial statistical approaches and / or layers using a geographic information system (GIS). The further data may be used in an inference step, for example in an inversion step.

[0032] Different sources of data can provide different information relevant to the soil structure. Techniques such as soil sampling provide detailed information but only for the sample area, which will typically be relatively small. A satellite can provide a large- scale map related to soil moisture, but no depth resolution; ground-penetrating radar can provide data on deeper horizons but little data about depth of moisture. Combining sets of different kinds of data can augment coverage, provide ground-truth sampling (for example for machine-learning inference), and better constrain uncertainties. The high-resolution map of the present disclosure provides a backbone to which otherdatasets can be added, as the seismic data can be the most promising dataset for both spatial coverage and assessing a wide range of soil parameters.

[0033] The method may comprise quantifying the uncertainty associated with a model or solution. The method may comprise mapping an uncertainty of the property.

[0034] The method may comprise determining a spatial or temporal or spatiotemporal map of the soil structure.

[0035] The method may comprise displaying the mapped property as an image. A map of the property of the soil structure may be displayed as an image in 2D or 3D. A map of the property of the soil structure, or an image of the soil structure may be used to determine the volume of soil in a layer or horizon of the soil structure. A map of the property of the soil structure, or an image of the soil structure may be used to detect the location of an interface or boundary between layers or horizons in the soil structure, for example, by identifying a change in the mapped or imaged property. A map of the property of the soil structure, or an image of the soil structure may be used to detect a location of the uppermost layer of the soil structure. A map of the property of the soil structure, or an image of the soil structure may be used to detect a location of a boundary of the uppermost layer of the soil structure. A map of the property of the soil structure, or an image of the soil structure may be used to determine the volume of soil in the uppermost layer of the soil structure. A map of the property of the soil structure, or an image of the soil structure may be used to detect the depth of location of the boundary of the uppermost layer of the soil structure.

[0036] Detecting a location of a boundary or a depth of location of a boundary may be achieved by inspecting a map or an image of the mapped property. Alternatively, detection of a location of a boundary or a depth of location of a boundary may be automated, or for example, achieved using a machine learning method.

[0037] In methods according to the present disclosure, a plurality of the seismic waves may be detected after they have passed through the soil structure and used to determine the property of the soil structure. A plurality of the seismic waves may be measured after they have passed though the soil structure and used to map the property of the soil structure. Data representing the plurality of seismic waves may be generated. Arrival times of the detected seismic waves may be recorded, and wavespeeds of the detected seismic waves may be inferred from the arrival times. The arrival times of the seismic waves may be measured and a map of the wavespeeds may be generated. A seismicimage may be generated from the inferred seismic wavespeeds. The amplitudes of the detected seismic waves may be used in the determination of the property of the soil structure; for example, differing detected wave amplitudes may be caused by differing levels of attenuation in the soil, which may in turn result from, for example, differing moisture contents. The waveforms of the detected seismic waves may be used in the determination of the propriety of the soil structure, for example by using a machinelearning algorithm to classify the soil type according to the waveform. Data representing the plurality of seismic waves may be processed by the inference method. The method may further comprise generating a map of the seismic wave speeds.

[0038] It may be that the map of seismic wavespeeds is generated by a method including solving an inference problem, an inverse problem or approximating a solution for an inverse problem. The inference method may involve mapping of seismic wavespeeds across a soil structure by solving non-linear inverse problems. Seismic wave detectors may be used to measure the arrival time (t) of a seismic wave relative to its generation at a seismic source or relative to a wave detected at another point in space or time. The inverse problem may be solved iteratively or with statistical methods or machine learning methods. As a simple example, the arrival time (t) of the seismic wave is related to the distance between the seismic wave source and the seismic detector (x) along the path of the wave below ground, and the seismic wavespeed (v) along that path through which the seismic wave has passed, by the equation t=x / v, where path x depends on v (as waves bend according to wavespeed variations), and v varies in space, i.e. along x, but this forward problem can still be linear and easy to solve, for instance as a sum over piecewise constant, flat layers. A map of seismic wavespeeds (v) may be generated by solving an expression for v, which is a function of x. As v is dependent upon x, and x non-trivially upon v, solving the expression v=function(x, t) is an inverse problem, which may be highly non-unique and non-linear depending on the accuracy of initial wavespeed estimates and data coverage, in that the problem can become over- or under-determined, or both.

[0039] A map of seismic wavespeeds may be displayed as an image. A map of seismic wavespeeds may be used to generate a seismic image. The image may be used to infer properties of the soil structure.

[0040] A map of seismic wavespeeds, or an image of seismic wavespeeds may be used to determine the volume of soil in a layer of the soil structure. A map of seismicwavespeeds, or an image of seismic wavespeeds may be used to determine the location of a boundary or interface of the soil structure, for example, by identifying a change in the mapped or imaged seismic wavespeeds. A map of seismic wavespeeds, or an image of seismic wavespeeds may be used to detect the location of interfaces within the uppermost layers of the soil structure. A map of seismic wavespeeds, or an image of seismic wavespeeds may be used to determine the depth or location of a boundary of the uppermost layers of the soil structure. A map of seismic wavespeeds, or an image of seismic wavespeeds may be used to determine the volume of soil in the uppermost layers of the soil structure. A map of seismic wavespeeds may be used to determine spatial variations in properties of the soil, for example undulating interfaces and layer thicknesses and / or variations in structural properties within layers.

[0041] Generating the map of seismic wavespeeds may comprise generating at least one wavespeed model for the soil structure; modelling the transmission of at least one seismic wave through the at least one model; comparing the at least one modelled seismic wave for the at least one model to a detected seismic wave; and using the comparison to update the seismic wavespeed model or select the best-fitting seismic wavespeed model.

[0042] Methods according to the present disclosure may comprise generating and detecting a plurality of seismic waves that have passed through the soil structure. A plurality of seismic waves may be detected using a single seismic wave detector, or using a plurality of detectors. A plurality of seismic waves may be detected at the same time, or at different times. A plurality of seismic waves may be detected using an array of seismic wave detectors, which may be distributed across an area of the soil structure.

[0043] The method of the first aspect may be a method of determining a change in a property of a soil structure over time. The method may comprise, at a plurality of different times, for example at regularly or irregularly spaced intervals, measuring a seismic wave after it has passed through the soil structure a plurality of different times. The method may then comprise using the seismic waves so measured to determine a change in a property of the soil structure with respect to time. A map of the property of the soil structure may be generated at a plurality of different times to map variations in the property of the soil structure with time.

[0044] Inference methods may alternatively or additionally use cross-correlation, or autocorrelation, of detected seismic waves. Thus, the inference method may includethe step of autocorrelating a seismic wave arriving at a seismic wave detector at two different times. The inference method may include the step of autocorrelating a seismic wave arriving at a seismic wave detector at more than two different times. The inference method may include the step of cross-correlating at least two detected seismic waves, for instance at least two different detectors.

[0045] Autocorrelation techniques may be used to correlate a seismic wave arriving at a seismic wave detector at different times. Cross-correlation techniques may be used to correlate at least two detected seismic waves, which may arrive at the same detector, or at different seismic wave detectors, at the same time, or at different times. Crosscorrelation or autocorrelation techniques may be used to map and / or track timedependent variation in a property of the soil structure.

[0046] The method may comprise generating a seismic wave(s), for example using a seismic wave source. Alternatively, the method may comprise utilising seismic waves already being generated by a seismic wave source, for example biological sources of seismic waves in the soil, or wind, or tree roots. In this case, the method may comprise identifying the location and / or characteristics (e.g. strength, time evolution, spatial extent, or direction) of the seismic wave source, for example through methods of back projection, source localisation, source inversion, or beam forming.

[0047] The seismic wave source may be a vibrating source, i.e., a seismic vibrator. The seismic wave source may be a vehicle, for example a seismic vibrator vehicle such as a seismic vibrator truck or a tractor, or other agricultural machine or other vibrational device. Advantageously, the seismic wave source may be an inexpensive source and / or a readily portable source. For example, the seismic wave source may be a hammer, for example, an instrumented hammer. The seismic wave source may be human source, for example, the high or ultra-high frequency source may be a person, for example a person stamping or walking on a surface of the soil structure. The seismic wave source may be a biological source, which may be within the soil structure. For example, the seismic wave source may be an earthworm or plant roots. The seismic wave source may be environmental noise, for example random high or ultra-high frequency environmental noise vibrations. Data resulting from using random sources can be processed using cross-correlation or auto-correlation to produce seismic images. The seismic wave source may be an explosive, for example a bird-scarer. The seismic wave source may be a vibrating device.

[0048] The seismic wave source may generate seismic waves having a range of frequencies. The seismic wave source may be tuneable to generate a seismic wave having a pre-determined frequency, or having frequencies within a pre-determined range of frequencies. The method may comprise tuning the seismic wave source to generate a seismic wave having a pre-determined frequency or a frequency within a range of pre-determined frequencies.

[0049] The method may comprise filtering the detected seismic wave source to generate a seismic wave having a pre-determined frequency, or a frequency within a pre-determined range of frequencies.

[0050] The method may comprise measuring the frequency (or frequencies) of the seismic wave(s) generated by the high or ultra-high frequency source.

[0051] The seismic wave source may be located at the surface of a soil structure. For example, the high or ultra-high frequency source may be placed directly onto a surface of the soil structure. Alternatively, the high or ultra-high frequency source may be positioned within the soil structure, near to a surface of the soil structure. The seismic wave source may for example be positioned within 50 cm of a surface of the soil structure, preferably within 10 cm of a surface of the soil structure. Alternatively, in cases where a soil structure is exposed, the seismic wave source may be positioned within the soil structure, for example, alongside a digging task.

[0052] The method may comprise providing a coupling medium between the seismic wave source and a surface of the soil structure. The coupling medium may be a metal shield or a metal plate. The coupling medium may be a stake or a support or holder for the seismic wave source. The coupling medium may filter the frequency of seismic waves generated by the seismic wave source and may reduce or eliminate the generation of harmonic frequencies or other unwanted signals.

[0053] The method may comprise detecting a seismic wave that has passed through the soil structure, using a seismic wave detector that is positioned at a location that may be remote from the seismic wave source. The seismic wave detector may be positioned at a surface of the soil structure. The seismic wave detector may be positioned for example more than 10 cm, more than 50 cm, more than 1 m, more than 5 m, more than 10 m, more than 15 m, or more than 20 m from the seismic wave source. The seismic wave detector may be positioned for example less than 100 m, less than 50 m, less than 20 m, less than 15 m, less than 10 m, less than 5 m, less than 1 m, less than 50 cm orless than 10 cm from the seismic wave source. The seismic wave detector may be positioned adjacent to the seismic source. A plurality of measurements may be made using a seismic wave source and a seismic wave detector that are at different distances from each other (this can be achieved for example by moving the source or the detector or both, or by using a plurality of sources and / or detectors at different locations). That is advantageous because the detected waves reflect from interface and other features at different locations and angles.

[0054] The seismic wave detector may be a geophone. Various designs of geophone are known by the skilled person. A geophone converts ground movement into voltage. For example, a basic form of geophone comprises a permanent magnet surrounding a wire coil, the wire coil being mounted on a spring. Seismic vibrations cause the wire coil to move within the field of the permanent magnet and hence to generate an electrical signal in the wire. This produces a voltage response dependent upon the speed of the ground movement. A geophone typically comes with a specific, unique conversion factor to translate output voltage back into the desired speed. More recent designs of geophone are implemented using a micromechanical system (MEMS). These typically produce a voltage response dependent upon the acceleration of the geophone and hence may be regarded as a form of accelerometer. Acceleration and speed outputs can be easily converted into one another. The seismic wave detector may be an accelerometer. Various designs of accelerometer are known by the skilled person. Traditional accelerometers measure the compression of a damped spring due to acceleration of the accelerometer. More recent designs of accelerometer are implemented using piezoelectric technology or a MEMS, for example a proof mass on a cantilevered beam. The seismic wave detector may be a seismic node receiver system. Advantageously, the seismic wave detector may be inexpensive and / or readily portable. For example, the seismic wave detector may be an accelerometer inside a mobile phone. Data related to the detected seismic wave may be stored remotely. For example, data relating to the detected seismic wave may be transmitted to a remote storage system such as a cloud-based system, for example via a 4g / 5g uplink, WiFi or Bluetooth.

[0055] The seismic wave detector may be arranged in use to touch the surface of the soil structure or to be within the soil structure. The seismic wave source may be arranged in use to touch the surface of the soil structure or to be wholly or partially within the soil structure.

[0056] The seismic wave detector may comprise an array of sensors. The array of sensors may be spaced apart over an area of a surface of the soil structure or within the soil structure.

[0057] The method may comprise using an array of seismic wave sensors to detect seismic waves across an area of the surface of the soil structure or within the soil structure.

[0058] The seismic wave detector may be mounted on a plate or surface, such as a metal plate. The seismic wave detector may be attached to the surface magnetically, or via a fixing means such as screws, or by gravity. The seismic wave detector may include a needle for penetrating into the ground, to improve the seismic coupling to the ground and / or to help ensure that the detector remains level.

[0059] The mapped property may be an elastic property (for example a seismic wave speed, for example a P-wave speed or an S-wave speed, a bulk modulus, a shear modulus, or Poisson’s ratio). The mapped property may be intrinsic attenuation of the structure. The mapped property may be anisotropy. The property may be a porosity. The property may be a composition. The property may be a pore connectivity. The property may be a moisture content of the soil. The property may be a carbon content of the soil. The property may be a bulk density of the soil. The property may be the depth of the uppermost boundaries or interfaces of the soil structure at a given location. The property may be the biodiversity of the soil, for example the biodiversity of species in the volume of the soil. The property may be a specific feature or localised structure inside the soil, such as at least one of a rock, an inclusion, an aquifer, a root, a cavity, a manufactured or otherwise non-indigenous material (for example metal, plastic, dead wood). The mapped property may be a volume of a soil layer. The mapped property may be undulations of soil horizons. The mapped property may be soil aggregate size distribution. It may be that the mapped property is not directly derived from the detected seismic wave, but rather is derived for example from a map of seismic wavespeeds or from a map of impedance contrasts.

[0060] The method may comprise mapping the carbon content or carbon concentration of the soil. The method may comprise mapping the composition of the soil. The property may be a density of the soil. The method may comprise using the mapped property to determine a volume of soil in a layer of the soil structure. The method may comprise mapping a volume of soil in a layer of the soil structure. The method maycomprise mapping the moisture content of the soil. The method may comprise using the mapped property to detect the location or depth of the uppermost boundary of the soil structure. The method may comprise mapping a time-dependent variation one or more of those properties over time. The mapping over time may be a continuous, or substantially continuous, mapping over a period of time, which may be a period of seconds, minutes, hours, days, weeks, months or years. The method may comprise mapping the location and / or abundance of biological sources inside the soil.

[0061] As the skilled person will understand, a “map” of the soil or soil structure is typically an array of data representing the values of at least one property of the soil structure as a function of spatial position (and optionally time) within the soil structure. The map may be displayed as an image on a display device, for example a computer monitor. The map may be a 2-dimensional map, for example a cross-section, or a collection of more than one 2D map, for example in different cross-sections, or different orientations. The map may be a 3 -dimensional map or a collection of 3D maps. In either case, the map may additionally vary over time. Current soil maps are 2D and based on geospatial statistics to estimate soil properties between sampled points. To understand soil structural properties and key characteristics such as spatial topsoil / A- horizon depth it is advantageous to map soils in 3D.

[0062] The method may use data from at least one experiment from at least one field site. The method may include the step of combining data from a plurality of different sites. The data that are combined and / or the combined data may be stored in one or more databases.

[0063] According to a second aspect, the present disclosure provides a method of determining soil health, comprising using a method according to the first aspect to map at least one property of the soil structure.

[0064] Soil health is strongly linked with soil porosity (see for example Neal, A.L., Bacq-Labreuil, A., Zhang, X. et al. Soil as an extended composite phenotype of the microbial metagenome. Sci Rep 10, 10649 (2020). https: / / doi.org / 10.1038 / s41598-020- 67631-0). Currently, there are no means of exploring soil porosity in the field; current methods require removal of samples for analysis in the laboratory. Methods according to the present disclosure may be used to characterise and quantify the porosity of soil, as well as soil organic matter content in soil (which also has strong associations with soil health), without the need to take soil samples for laboratory analysis.

[0065] The method of determining soil health may include the step of combining the mapped at least one property with further data, relating to the soil structure, selected from the following list: geophysical data (for example radar data, for example groundpenetrating radar data, or electromagnetic data), conventional soil-sampling data (for example penetrologger data, data from chemical analysis, eDNA data, data relating to above-ground diversity, data relating to below-ground diversity), remote sensing data (for example data from a satellite or drone), data relating to local knowledge (e.g. data or other information received from a farmer, for example a farmer of a specific region, for example a farmer of the soil structure, data relating to previous management or use of the soil structure), weather data. The method of determining soil health may include the step of combining the mapped at least one property with further soil-related data from at least one site remote from the mapped soil structure. The remote site may be a site that has a different ecosystem from that of the mapped soil structure. The remote site may be a site that has been farmed using different agricultural practices from that of the mapped soil structure.

[0066] Soil structural degradation reduces soil health through reduced gas exchange and water flow, exacerbating soil erosion and limiting plant growth and crop productivity. Currently it is not possible to identify areas of degraded soil structure including subsoil compaction, in the field, in a spatially explicit manner. Methods according to the present disclosure may be used to identify these areas, spatially or, in the case of subsoil compaction for example, the depth where the issues exist. Methods according to the present disclosure may therefore help farmers and land managers to identify where and how they need to take remedial action.

[0067] According to a third aspect, the present disclosure provides a method of improving soil health, the method including the steps of (i) using a method according to the first aspect to map at least one property of the soil structure and (ii) selecting a soil treatment plan based at least in part on the mapped property or properties.

[0068] The soil treatment plan may comprise subsoiling of the soil or an area of the soil. The soil treatment plan may comprise remedial tillage of the soil or an area of the soil. The soil treatment plan may comprise adding organic matter to the soil or an area of the soil.

[0069] Selecting the soil treatment plan may for example comprise the soil treatment plan being derived, developed or updated from the mapped property or properties. Thesoil treatment plan may be derived or developed using data science techniques, for example machine learning. The data science techniques may be applied to a database that includes the mapped property or properties. The database may include further data relating to other properties of the soil. For example, the further data may be one or more selected from the following list: radar data, ground-penetrating radar, electromagnetic data, penetrologger data, data from chemical analysis, eDNA data, data relating to above-ground biodiversity, weather data, data relating to below-ground biodiversity, remote sensing data, data, information received from a farmer, information received from a farmer about a specific region.

[0070] In embodiments according to the second or third aspects, the at least one property may be any of the properties of soil structures set out herein. A plurality of the properties of the soil structure may be mapped. The mapped plurality of properties may be combined to produce a combined property map.

[0071] According to a fourth aspect, the present disclosure may provide apparatus for mapping a property of a soil structure, the apparatus being arranged to receive a data signal representing a detected seismic wave from a seismic wave detector. The apparatus may comprise a computing device comprising a processor and memory storing a computer program for execution by the processor, the computer program comprising instructions that, when the program is executed by the processor, cause the computing device to use the generated data signal to map a property of the soil structure. The map has a resolution that resolves features that are less than Im apart.

[0072] The apparatus may further comprise a storage database, for example for storing the map, and / or unprocessed or processed data, and / or metadata.

[0073] The computing device may be for example a mobile phone, a tablet, a personal computer, a mainframe computer, a distributed computing system, or a cloud computing facility, or a bespoke computing device, for example attached to the apparatus.

[0074] The apparatus may include any of the features described in relation to the methods above.

[0075] The computer program may comprise instructions that, when the program is executed by the processor, cause the computing device to transmit the map to a remote database. The computer program may comprise instructions that, when the program is executed by the processor, cause the computing device to merge the mapped propertywith other data, for example data relating to soil, geography, or data gathered by remote sensing or from a farmer.

[0076] The apparatus may further comprise a seismic wave source.

[0077] The apparatus may comprise a seismic wave detector for detecting a seismic wave after it has passed through the soil or soil structure and for generating a data signal representing the measured seismic wave.

[0078] The seismic wave detector may be provided together with a seismic wave source in a common housing. The seismic wave detector may be remote from a seismic wave source. The apparatus may include a data connection from the seismic wave detector and / or the seismic wave source.

[0079] The seismic wave source may be a source that is capable of generating or configured to generate high or ultra-high frequency seismic waves. The seismic wave source may be configured to generate seismic waves having a frequency (which may be a centre frequency in the case of a broadband wave source) of between 50 and 3000 Hz. The seismic wave source may be configured to generate seismic waves having a frequency (which may be a centre frequency in the case of a broadband wave source) of between 200 and 3000 Hz.

[0080] The seismic wave source may be a vibrating source, i.e., a seismic vibrator. The seismic wave source may be a vehicle, for example a seismic vibrator vehicle such as a seismic vibrator truck, or a tractor, or another agricultural vehicle or vibrational device. The seismic wave source may be a hammer, for example, an instrumented hammer. The seismic wave source may be a person, for example, the seismic wave source may be a person stamping or walking on a surface of the soil structure. The seismic wave source may be a biological or an environmental source, which may be within the soil structure, for example within the soil, for example, the seismic wave source may be one or more earthworms or one or more tree roots (for example via which vibrations caused by the wind on the tree are transmitted to the soil).

[0081] The apparatus may include a filter for filtering the frequencies of seismic waves generated by the seismic wave source.

[0082] The apparatus may include a measuring or monitoring device for measuring and / or monitoring the frequency of seismic waves generated by the seismic wave source.

[0083] The apparatus may include a coupling medium provided or positioned between the seismic wave source and a surface of the soil structure. The apparatus may include a coupling medium positioned between the seismic wave source and a surface of the soil structure. The coupling medium may be a metal plate or metal sheet, rock, wood or an engineered material. The coupling medium may limit or reduce the generation of harmonic seismic wave frequencies from the seismic wave source.

[0084] The seismic wave detector may be a detector that is configured to detect relatively high frequency seismic waves. The seismic wave detector may be configured to detect seismic waves having a frequency of between 50 and 3000 Hz. The seismic wave detector may be configured to generate seismic waves having a frequency of between 200 and 3000 Hz.

[0085] The seismic wave detector may be a geophone. The use of geophones as seismic wave detectors for low frequency seismic waves is known in geotechnical applications. Apparatus of the present disclosure may make use of a geophone detector, or an array of geophone detectors, that can detect high or ultra-high frequency seismic waves, for example having frequencies up to 3000 Hz.

[0086] The seismic wave detector may be an accelerometer. The use of accelerometers as seismic wave detectors is known in engineering applications, and accelerometers are used in smart phone technology, for example as movement trackers or structural monitors. Apparatus of the present disclosure may make use of an accelerometer detector, or an array of accelerometer detectors that can detect high or ultra-high frequency seismic waves, for example having frequencies up to 3000 Hz. Environmental noise levels are typically lower at higher frequencies and therefore, advantageously, using an accelerometer to detect higher frequency seismic waves may result in less noise in detected seismic waves, compared to other known seismic wave detectors. The seismic wave detector may be an accelerometer inside a mobile phone. In this case, data related to the detected seismic wave may be stored remotely. For example, data relating to the detected seismic wave may be transmitted to a remote storage system such as a cloud-based system via a 4g / 5g uplink.

[0087] The seismic wave detector may be a seismic node receiver system. The seismic node receiver system may be used to detect seismic waves having frequencies of up to 200 Hz, for example up to 125 Hz. The apparatus may comprise an array of seismic nodes for detecting seismic waves. The array may include several tens, several hundredor several thousand seismic nodes or more. The seismic node receiver system may include an integrated battery. The seismic node receiver system may detect and record seismic data autonomously, and may include automated data-logging capabilities. Using an array of nodes to detect seismic waves may be particularly advantageous when measuring time dependent properties of a soil structure.

[0088] The seismic wave detector may comprise an array of seismic wave detectors distributed across an area of the soil structure.

[0089] The apparatus comprises a processor for processing data attributed to the detected seismic wave in order to map a property of the soil structure. The apparatus may include seismic imaging apparatus. The apparatus may include a display for displaying a seismic image. The apparatus may be a mobile phone. The method may be carried out by an app on a mobile phone.

[0090] FIG. 1 shows an example method 1 of mapping a property, in this example the composition, of a soil structure according to an embodiment of the present disclosure. A seismic wave is detected after it has passed through the soil structure (step 3). The detected wave is then used to map a property of the soil structure (step 5).

[0091] In another example method 10 (FIG. 2), a seismic wave is detected by generating a seismic wave (step 30) from a seismic wave source at or near to a surface of a soil structure, so that the wave passes through the soil structure. The seismic wave is detected using a seismic wave detector after transmission through the soil structure (step 50). The seismic wave detector produces a signal that comprises data representing the detected wave (step 70). A property of the soil structure, in this example the density of the soil structure, is determined by processing the data using an inference method, in this example, the inverse method of Bayesian inference (step 90). (In alternative embodiments, the composition of the soil structure may be calculated using method of solving the forward problem). The processed data is used to generate a map of a soil property, in this example the density of the soil structure (step 100). In step 102, the soil property map of the soil structure is displayed as an image. At a boundary between soil layers, a change in density may be observed. In step 104, density variations in the displayed image are used to detect the location of a boundary in the soil structure.

[0092] FIG. 3 shows example apparatus for mapping a property, in this example the compaction, of a soil structure according to an embodiment of the present disclosure. A vibrating seismic wave generator 110 is positioned on the surface of a soil layer 113.(In alternative embodiments, the source of seismic waves may be, for example, a person stamping on the soil surface, or the movement of earthworms within the soil.) A metal sheet 112 is provided between the seismic wave generator and the soil surface to act as a coupling medium, improving transmission of the generated wave into the soil whilst reducing artefacts from spurious signals such as an uneven surface. The seismic wave generator 110 generates seismic waves 115 at a central frequency of around 1000 Hz. The seismic waves 115a, 115b, 115c are transmitted through the soil structure 113, and detected by an array of accelerometers 117a, 117b, 117c, which are positioned at the surface of the soil structure 113, around 2 m from the seismic wave generator 110. The array of seismic wave accelerometers 117a, 117b, 117c are connected to a processor 119. The processor 119 collates the arrival times (t) of the seismic waves 115a, 115b, 115c, and uses these arrival times, along with the distance (x) between the seismic wave source 110 and each seismic wave detector 117a, 117b, 117c, to solve the non-linear equation system seismic wavespeed (v) = function(x, t) for the multiple measurements t and corresponding distances x to invert for a best-fitting subsurface seismic wavespeed map of values v. The processor 119 generates the map of wavespeeds, and uses this map to generate an image of the soil structure 113. The image generated by the processor is used to infer the density of the soil structure 113. In some embodiments, the image is generated at different points in time, for example over several months, to infer how the density of the soil surface changes with time.

[0093] Whilst the present disclosure has been described and illustrated with reference to particular embodiments, it will be appreciated by those of ordinary skill in the art that the disclosure lends itself to many different variations not specifically illustrated herein.

[0094] In embodiments of the invention, the inverse problem is solved, either for full wavefields or any measurements of the wavefield.

[0095] Where in the foregoing description, integers or elements are mentioned which have known, obvious or foreseeable equivalents, then such equivalents are herein incorporated as if individually set forth. Reference should be made to the claims for determining the true scope of the present disclosure, which should be construed so as to encompass any such equivalents. It will also be appreciated by the reader that integers or features of the disclosure that are described as preferable, advantageous, convenient or the like are optional and do not limit the scope of the independent claims. Moreover,it is to be understood that such optional integers or features, whilst of possible benefit in some embodiments of the disclosure, may not be desirable, and may therefore be absent, in other embodiments.

Claims

Claims1. A method of mapping a property of a soil structure, the method comprising: detecting a seismic wave after it has passed through the soil structure; and- using the detected wave to map a property of the soil structure, wherein the seismic wave has a frequency that is sufficiently high to resolve features that are less than 1 m apart from each other.

2. A method according to claim 1, wherein the seismic wave has a frequency of between 50 Hz and 3000 Hz.

3. A method according to claim 1 or claim 2, further comprising generating the seismic wave at or near to a surface of the soil structure.

4. A method according to any preceding claim, wherein an inference method is used to determine the property of the soil structure from the detected wave.

5. A method according to claim 4, wherein the inference method includes the step of autocorrelating a seismic wave arriving at a seismic wave detector at two or more different times.

6. A method according to claim 4, wherein the inference method includes the step of cross-correlating at least two detected seismic waves.

7. A method according to any preceding claim, wherein the property is at least one of a porosity of the soil, pore connectivity, composition of the soil, moisture content of the soil, density of the soil, volume of soil layers, undulations of soil horizons, soil aggregate size distribution, carbon content of the soil or an elastic property.

8. A method according to any preceding claim, further comprising displaying the mapped property as an image.

9. A method according to any preceding claim comprising using the mapped property to determine a volume of soil in a layer of the soil structure, or to detect a location or depth of an interface or boundary of the soil structure.

10. A method according to any preceding claim, wherein a plurality of the seismic waves are measured, after the waves have passed though the soil structure, and used to map the property of the soil structure.

11. A method according to claim 10, wherein the arrival times of the seismic waves are measured and a map of the wavespeeds is generated.

12. A method according to claim 11, wherein the map of seismic wavespeeds is generated by a method including solving an inference problem, an inverse problem or approximating a solution for an inverse problem.

13. A method according to claim 11, wherein the map of seismic wavespeeds is generated by a method comprising generating at least one wavespeed model for the soil structure; modelling the transmission of at least one seismic wave through the at least one model; comparing the at least one modelled seismic wave for the at least one model to a detected seismic wave; and using the comparison to update the seismic wavespeed model or select the best-fitting seismic wavespeed model.

14. A method according to any preceding claim wherein a map of the property of the soil structure is generated at a plurality of different times to map variations in the property of the soil structure with time.

15. A method of determining soil health, comprising using a method according to any proceeding claim to map at least one property of the soil structure.

16. A method according to claim 15, wherein the method includes the step of combining the mapped at least one property with further data, relating to the soil structure, selected from the following list: radar data, ground-penetrating radar, electromagnetic data, penetrologger data, data from chemical analysis, eDNA data, data relating to above-ground biodiversity, weather data, data relating to below-ground biodiversity, remote sensing data, data, information received from a farmer, information received from a farmer about a specific region.

17. A method of improving soil health, the method including the steps of (i) using a method according to any preceding claim to map at least one property of the soil structure and (ii) selecting a soil treatment plan based at least in part on the mapped property or properties.

18. Apparatus for mapping a property of a soil structure, the apparatus being arranged to receive a data signal representing a detected seismic wave from a seismic wave detector, the apparatus comprising:a computing device comprising a processor and memory storing a computer program for execution by the processor, the computer program comprising instructions that, when the program is executed by the processor, cause the computing device to use the generated data signal to map a property of the soil structure, the map having a resolution that resolves features that are less than Im apart.

19. Apparatus according to claim 18, further comprising a seismic wave detector for detecting a seismic wave after it has passed through the soil structure and for generating a data signal representing the detected seismic wave.

20. Apparatus according to claim 18 or claim 19, further comprising a seismic wave source.

21. Apparatus according to claim 20, wherein the seismic wave source is configured to generate a seismic wave having a frequency of between 50 Hz and 3000 Hz.

22. Apparatus according to claim 20 or claim 21, wherein the seismic wave source is a hammer, a vehicle, a person, or a biological source within the soil structure, an explosive, a vibrating device or environmental noise.

23. Apparatus according to any of claims 20-22, further comprising a coupling medium positioned between the seismic wave source and a surface of the soil structure.

24. Apparatus according to any of claims 18-22, wherein the seismic wave detector is a geophone.

25. Apparatus according to any of claims 18-22, wherein the seismic wave detector is an accelerometer.

26. Apparatus according to any of claims 18-22, wherein the seismic wave detector is a seismic node receiver system.

27. Apparatus according to any of claims 18-26 comprising an array of seismic wave detectors distributed across an area of the soil structure.

Images