System, Method, and Apparatus for Reconciling Direct and Indirect Emission Measurements FIELD OF THE INVENTION [0001] The present invention relates to a system, method and apparatus for reconciling direct and indirect emission measurements. More particularly, the present invention pertains to a method for reconciling direct and indirect emission measurements having utility in the oil and gas industry. BACKGROUND [0002] Companies across the energy sector are becoming increasingly aware of the importance of measuring accurately the emissions associated with the production, transportation, processing and refining of oil and gas products. Of particular concern is the emission of methane (CH4), which has a global warming potential (GWP) of ~27, meaning that one tonne of methane is equivalent to 27 tonnes of carbon dioxide (CO ₂ ) in its ability to warm the planet. [0003] The amount of methane emitted from a piece of equipment can be estimated using engineering equations that take as input operational information and emissions factors (EF). Taking a sum of the emissions from each piece of equipment at a facility allows for an estimation of the total facility emissions; and this is the method used to report emissions to various regulatory bodies, such as the Environmental Protection Agency (EPA) in the United States of America. [0004] Increasingly, independent research studies are finding that estimations of methane relying solely on engineering estimates are inadequate [1]. As a result, many companies in the energy sector have installed methane measurement devices at their facilities. The measurement devices may be installed locally, such as point concentration sensor or an optical gas imaging (OGI) system; or the devices may be infrequent, such as a drone or satellite imaging system. These devices measure the concentrations of methane in the air, either at a point in space (point sensors), or over an integrated area (OGI and flyover/satellite images), and can be temporally continuous or discrete. This results in a disconnect between the measured emissions (concentration of methane at different points in space) and the calculated emissions (emissions rates of individual pieces of equipment). [0005] Concentrations, whether integrated spatially or locally, must be causally connected to an emissions rate using a plume model [2, 3]. A plume model calculates the concentration profile of an emitted plume given information about the emissions rate, the emissions location, and atmospheric data (such as wind speed and direction). However, in the case of several possible emissions sources, solving for the emissions rates of these sources given measured concentrations becomes intractable as there are an infinite number of solutions that will yield the measured concentration. In order for companies to estimate accurately the true methane emissions rate, it is necessary to reconcile the measured emissions concentrations with the emissions rates. [0006] One method of reconciling data sources is to leverage Bayes’ Theorem [4]. Bayes’ Theorem provides an approach to estimate the probability of an event, such as equipment having a particular emissions rate, given prior knowledge of conditions that are related to the event, such as global engineering estimates of the emissions rate for a particular piece of equipment. The knowledge of the probability of an event prior to an observation is referred to as a prior. Observations that are made, such as the direct measurement of an emissions rate, define a probability distribution for the observed measurements known as a likelihood. Using the prior and likelihood it is possible to construct a probability distribution known as a posterior. The posterior describes the probability of an event occurring given both prior assumptions and real observations of the event. Bayes Theorem describes the mathematical approach to constructing the posterior from the prior and likelihood, which enables estimation of the posterior probability distribution. [0007] There is a need for further systems, methods and apparatuses for reconciling direct and indirect emission measurements in the oil and gas industry. SUMMARY OF THE INVENTION [0008] Herein is provided an approach to reconcile emissions measurement data with calculated emissions rates. This approach leverages Bayes’ Theorem to enable a direct comparison between the calculated emissions rates that rely on engineering estimates, emissions factors, and operational data and direct measurement methods that measure in real time the emissions concentrations from a site or facility. The approach is explained in detail in subsequent sections: in summary a plume is constructed at each possible emissions source in which the emissions rate that generates the plume is sampled from a prior distribution of emissions rates. A likelihood distribution is constructed from the observed concentrations from the measurement devices, and from this likelihood and the predicted stochastic plumes, a posterior distribution on the emissions rates at each source of equipment is obtained. The parameters defining the posterior distribution enable a more accurate estimation of the true emissions rates from each equipment source, and provide an estimation of the uncertainty associated with each source. Furthermore, the calculated posterior distribution can be extended to infer the distribution on the total sum of emissions rates across each equipment source. [0009] In one aspect, a method for reconciling direct and indirect emission measurements is provided, the method comprising: determining a calculated emissions rate for each emissions source of a plurality of emissions sources; modeling a stochastic plume at each emissions source utilizing the corresponding calculated emissions rate so as to estimate a concentration map of emissions; measuring observed emissions concentrations for the plurality of emissions sources via at least one device; determining a likelihood distribution from the observed emission concentrations for each emissions source; obtaining a posterior distribution of the emissions rates for each emissions source from the likelihood distribution of the emissions source and the stochastic plume of the emissions source. BRIEF DESCRIPTION OF THE FIGURES [0010] Fig. 1 depicts schematically a non-limiting example of a facility for which emissions reconciliation may be deployed. [0011] Fig.2 shows a non-limiting example of the processed output of different measurement sources. Upper panels show the output of a single point sensor, and lower panels show the output from an optical gas imaging (OGI) camera. [0012] Fig. 3 describes the workflow of the many-plume reconciliation method. [0013] Fig. 4 shows an exemplary output of the reconciliation process. The emissions rate for each piece of equipment is sampled from the posterior that is constructed from the prior emissions rate assumptions and the likelihood defined by the concentration observations. The sum of the emissions rate from each piece of equipment provides the estimate of the emissions rate for the entire facility. DETAILED DESCRIPTION [0014] According to at least one exemplary embodiment, a method, system, and apparatus for reconciling methane emissions measurements (referred to subsequently herein as “top-down”) with calculated methane emissions rates (referred to subsequently herein as “bottom-up”). Fig. 1 shows an example of a typical layout of a measurements reconciliation problem. Each possible emissions source, such as individual pieces of equipment, are assigned a calculated emissions rate. This emissions rate may be calculated individually for each source using a combination of global engineering estimates [1]; empirical formulae relying on operational data (i.e., temperature, pressure, etc.); and other approaches. For each emissions source, a plume is modeled such that a concentration map of methane is estimated over the entirety of the spatial domain. When considering multiple sources, the sum of the concentrations generated by each plume individually will result in a total concentration. [0015] The sum of the concentrations from each emissions source can be compared to the observed concentrations from the different measurement devices. In Fig. 1, the measurement devices are the two fixed point sensors, the mobile detector and the satellite image. Each of these measurements return a concentration at a fixed point, or an integrated concentration over a spatially confined region (e.g., over an area and/or along a line-of-sight). These data are measured with different temporal resolution and cadence. The plume model allows us to predict the observed concentration of methane resulting from k = 1 to N equipment locations (q _k , p _k , r _k ) and emissions rates E _k . The point concentration at the point (x, y, z) can be calculated following the formula where M _Plume is the chosen plume model that incorporates air dispersion at a given time t. Point concentrations are typically observed by fixed, mobile, and airborne sensors. It is also possible to calculate total line-of-sight concentrations from a point concentration map, e.g., where yo is the observer location. Line-of-sight concentrations are applied to spectroscopic measurements, which observe the sum total concentration along a particular line-of-sight. This applies to fixed, mobile, and airborne spectroscopic detectors. Fig. 2 depicts examples of the predicted observations for point sensors panels; note that in the upper right-hand panel, the line for Sensor 1 is the lower line on the graph) and integrated line-of-sight concentrations (lower panels) from an optical gas imaging (OGI) camera. [0016] The premise of reconciliation is to leverage the measurement data to refine the estimations of the emissions rates at each emissions source [Fig.3]. Priors are placed on either the function parameters relating operational data to emissions rates; or on the emissions rates themselves. These stochastic emissions rates generate a plume, which is itself stochastic. The observed concentrations are used to construct a likelihood distribution of concentrations as a function of time. This likelihood distribution allows us to generate posteriors of the emissions rates at each emissions source, of which the mean is a better estimate of the true emissions rates. This enables a more accurate estimation of the emissions from each source and the total emissions from a particular facility or site. We use Bayes’ Theorem to relate the posterior distribution to the prior and likelihood distributions. First, we suppress the arguments in the variables such that the observed concentration at different times can be expressed as and the emissions rate at different times can be expressed as . The posterior distribution is proportional to the prior and likelihood distributions where is the prior for each unit of equipment (labeled from 1 to N), is the likelihood distribution, the probability of observing concentrations from each sensor (labeled from 1 to K) for a given set of emissions rates (e.g., the probability of the observation when the emissions rate is known), and is the posterior distribution, the probability rates concentrations. When the observational uncertainty is Gaussian, we express the likelihood using a Gaussian distribution, e.g., for independent identically distributed Gaussian uncertainty: However, this can be generalized to other types of observational uncertainty by replacing the likelihood distribution. The prior distribution the probability of emissions rates prior to observation. We assume the emissions rates from each unit of equipment to be independent, so that To correlate the emissions rates in time, we employ a statistical class of distributions called Gaussian Processes (GP; [4]). We leverage historical operations data to set the properties of the chosen GP; to ensure that the chosen GP is restricted to positive emissions rates, we transform the GP to be expressed in a logarithmic scale or to take the absolute value of the associated random variable. To estimate the posterior probability distribution, , we use Markov chain Monte Carlo (MCMC; [4]). The MCMC algorithm of choice efficiently samples from the likelihood and prior distributions to estimate the posterior distribution, e.g., MCMC, will simulate different plume realizations based on the prior emissions rate distributions and the observed concentration likelihoods to find an approximation to the emissions rate posterior distribution as a collection of samples [Fig. 4]. It is then straightforward to estimate the posterior distribution from variables that can be derived from the emissions rate. We use the MCMC samples of each individual emissions source to estimate the posterior probability distribution for the total emissions rate. [0017] Fig. 4 depicts an exemplary output of the many-plume model. In this figure, the emissions rate for each piece of equipment is sampled from the posterior that is constructed from the prior emissions rate assumptions and the likelihood defined by the concentration observations. The sum of the emissions rate from each piece of equipment provides the estimate of the emissions rate for the entire facility. In the panels shown in the top row of Fig. 4, the two darker circles represent the sensors (e.g. see the top left-hand side panel, where the sensors are denoted with labels “1” and “0” in white text to the right of same), and the remaining three lighter circles represent the sources. In the line graph showing Sensor Concentration (ppm) versus Emissions Rate 1, the line for Sensor 1 is the lower line in the graph. In the other plots, the Mean line closely follows the lighter Median line (the latter of which is more visible in the plots), and the darker line shown in these plots is the Ground Truth line. [0018] All publications, patents and patent applications mentioned in this Specification are indicative of the level of skill of those skilled in the art to which this invention pertains and are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. [0019] Although the present invention has been described with reference to the preferred embodiments, it is to be understood that modifications and variations may be resorted to without departing from the spirit and scope of the invention, as those skilled in the art readily understand. Such modifications and variations are considered to be within the purview and scope of the invention and the appended claims. References: 1. Rutherford, J.S., Sherwin, E.D., Ravikumar, A.P. et al. Closing the methane gap in US oil and natural gas production emissions inventories. Nat Commun 12, 4715 (2021). 2. Turner, D. B. (2020). Workbook of atmospheric dispersion estimates: an introduction to dispersion modeling. CRC press. 3. Cimorelli, A. J., S G. Perry, A. Venkatram, J. C. Weil, R. J. Paine, R. B. Wilson, R. F. Lee, W. D. Peters, AND R. W. Brode. AERMOD: A DISPERSION MODEL FOR INDUSTRIAL SOURCE APPLICATIONS PART I: General Model Formulation and Boundary Layer Characterization. Journal of Applied Meteorology and Climatology. American Meteorological Society, Boston, MA, 44(5):682-693, (2005). 4. Dunson, D. B., Rubin, D. B., Gelman, A., Vehtari, A., Stern, H. S., Carlin, J. B. (2014). Bayesian Data Analysis, Third Edition. United Kingdom: Taylor & Francis.