SYSTEM AND METHOD FOR IMPLEMENTING A DATA-DRIVEN FRAMEWORK FOR OBSERVATION, DATA ASSIMILATION, AND PREDICTION OF OCEAN CURRENTS

§102 §103 §112

DETAILED ACTION Notice of Pre-AIA or AIA Status 1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . 2. Claims 1-20 have been presented for examination based on the application filed on 06/07/2023. 3. Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. 4. Claim(s) 1-3, 5-12, 14-20 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US Patent No. US20080167822A1 by Loukos. 5. Claim(s) 4 and 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over US Patent No. US20080167822A1 by Loukos further in view of NPL: "Assimilation of drifter observations for the reconstruction of the Eulerian circulation field" by Molcard (2003). 6. This action is made Non-Final. ---- This page is left blank after this line ---- Claim Rejections - 35 USC § 112 The following is a quotation of 35 U.S.C. 112(b): (b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention. The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph: The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention. Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. The term "selectively outputting" in claims 1, 10, and 19 renders the claim scope uncertain. This term implies a choice or condition for the output of ocean condition predictions, yet the claim fails to recite the specific criteria, quantitative metrics, or triggers (e.g., accuracy thresholds or variance limits) that govern this selection process. While the specification (e.g., paragraph [004] and FIG. 3, step 140) suggests that a determination is made, the absence of an objective standard within the claim language prevents a person of ordinary skill in the art from determining the metes and bounds of the limitation with reasonable certainty. Consequently, it is unclear whether a system that outputs predictions continuously, or only under specific unstated conditions, would infringe the claim. (See MPEP § 2173.05(b) and MPEP § 2173.05(g)) Claim Rejections - 35 USC § 102 In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action: A person shall be entitled to a patent unless – (a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention. (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. (g)(1) during the course of an interference conducted under section 135 or section 291, another inventor involved therein establishes, to the extent permitted in section 104, that before such person’s invention thereof the invention was made by such other inventor and not abandoned, suppressed, or concealed, or (2) before such person’s invention thereof, the invention was made in this country by another inventor who had not abandoned, suppressed, or concealed it. In determining priority of invention under this subsection, there shall be considered not only the respective dates of conception and reduction to practice of the invention, but also the reasonable diligence of one who was first to conceive and last to reduce to practice, from a time prior to conception by the other. A rejection on this statutory basis (35 U.S.C. 102(g) as in force on March 15, 2013) is appropriate in an application or patent that is examined under the first to file provisions of the AIA if it also contains or contained at any time (1) a claim to an invention having an effective filing date as defined in 35 U.S.C. 100(i) that is before March 16, 2013 or (2) a specific reference under 35 U.S.C. 120, 121, or 365(c) to any patent or application that contains or contained at any time such a claim. The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action. Claim(s) 1-3, 5-20 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US Patent No. US20080167822A1 by Loukos. Regarding Claim 1 Loukos teaches A system comprising: (Abstract]: “A climatic forecast system for generating an optimized forecast includes: a computer including a processor and a storage unit; means for remotely acquiring digital data via a digital network, said digital data including observations and at least one model digital forecast; and means for hybrid forecasting, including, means for inputting said observations, said at least one model digital forecast, and first forecast data, means for hybrid forecasting adapted to process said observations, said at least one model digital forecast, and to apply at least one of a statistical model, a simple general circulation model, and a multi-model synthesis algorithm to said first forecast data, and means for outputting at least one hybrid forecast.”) at least one processor; (Abstract]: “A climatic forecast system for generating an optimized forecast includes: a computer including a processor and a storage unit; means for remotely acquiring digital data via a digital network, said digital data including observations and at least one model digital forecast; and means for hybrid forecasting, including, means for inputting said observations, said at least one model digital forecast, and first forecast data, means for hybrid forecasting adapted to process said observations, said at least one model digital forecast, and to apply at least one of a statistical model, a simple general circulation model, and a multi-model synthesis algorithm to said first forecast data, and means for outputting at least one hybrid forecast.”) and, at least one memory having instructions stored thereon that, when executed by the at least one processor, cause the processor at least to: : (Abstract]: “A climatic forecast system for generating an optimized forecast includes: a computer including a processor and a storage unit; means for remotely acquiring digital data via a digital network, said digital data including observations and at least one model digital forecast; and means for hybrid forecasting, including, means for inputting said observations, said at least one model digital forecast, and first forecast data, means for hybrid forecasting adapted to process said observations, said at least one model digital forecast, and to apply at least one of a statistical model, a simple general circulation model, and a multi-model synthesis algorithm to said first forecast data, and means for outputting at least one hybrid forecast.”) generate data ensembles (0014]: “generating a multimodel superensemble of climatic models based upon the performances of each of the models.”) using a stochastic data driven ([0032]: “In relation thereto, the present invention relates, in its more general meaning, to a climatic forecast system, using at least one computer comprising in particular a processor and a storage unit, means for remotely acquiring digital data via a computer network or a telecommunication network, said digital data being observations and single-model or multimodel digital forecasts, said system being characterised in that it further comprises hybrid forecasting means inputting said overall observations and forecasts and outputting at least one hybrid forecast in the form of probability density, said hybrid forecasting means including a combination and not just a simple juxtaposition of at least two models among statistical models, simple general circulation models and multimodel synthesis algorithms.” [0032]: “comparing the historical forecast data and the observations to determine the historical performance of the various models” The examiner interprets where "stochastic" and "data driven" is shown in the form of "probability density" based on "historical performance") ensembles prediction model (0068]: “Referring to FIG. 1, the simple-model or multimodel digital forecasts (11) are overall or deterministic forecasts, issued by global atmospheric circulation models coupled or not coupled to other models (ocean, biosphere, cryosphere, chemical, etc.) a plurality of forecasting types the main ones of which are weather forecasts provided by the ECMWF (European Centre for Medium range Weather Forecasts) or the NCEP (National Center for Environmental Prediction), seasonal forecasts from the ECMWF, NCEP or the MetOffice and climate change simulations.” [FIG. 1]: PNG media_image1.png 930 893 media_image1.png Greyscale ) trained on ocean simulations; ([0032]: “In relation thereto, the present invention relates, in its more general meaning, to a climatic forecast system, using at least one computer comprising in particular a processor and a storage unit, means for remotely acquiring digital data via a computer network or a telecommunication network, said digital data being observations and single-model or multimodel digital forecasts, said system being characterised in that it further comprises hybrid forecasting means inputting said overall observations and forecasts and outputting at least one hybrid forecast in the form of probability density, said hybrid forecasting means including a combination and not just a simple juxtaposition of at least two models among statistical models, simple general circulation models and multimodel synthesis algorithms.” receive observation data; (Abstract]: “A climatic forecast system for generating an optimized forecast includes: a computer including a processor and a storage unit; means for remotely acquiring digital data via a digital network, said digital data including observations and at least one model digital forecast; and means for hybrid forecasting, including, means for inputting said observations, said at least one model digital forecast, and first forecast data, means for hybrid forecasting adapted to process said observations, said at least one model digital forecast, and to apply at least one of a statistical model, a simple general circulation model, and a multi-model synthesis algorithm to said first forecast data, and means for outputting at least one hybrid forecast.”) estimate an analysis state using a multi-layer perceptron ([0015]: “Prior art also knows, from American patent U.S. Pat. No. 5,461,699 (IBM), a forecast system including a three-layer (or more generally multilayer) neural network. Said network is applied to climatic data and trained on such data. The first node of the input layer is connected to the statistical forecasts whereas the other nodes are linked to one historical datum from the set of historical data.”) that represents the analysis state as a non-linear conjunction of observations ([0015]: “According to one preferred embodiment, the system according to the invention uses neural networks for statistical analysis in the climatic field. Said neural networks firstly provide a non-linear approach of the analysis and rapid processing of a large data flow.”) and minimizing variance across the ensembles; (See [FIG. 1], [Abstract], [0032], [0040]: “According to one embodiment, said multimodel synthesis, statistical forecasting, hybrid forecasting, downscaling, benchmarking, or overall synthesis means use Bayesian statistical methods.” [0064]: “the combination of forecasts in order to provide the optimum forecast.” [0082]: “Said algorithm uses a synthesis based on digital forecasts alone which may, as described in the Kernels method (Silverman 1986), be a Probability Density Function (PDF) in one point from individuals of a single-model assembly. Alternatively, said method may be extended to 2D or 3D models, as in the S. Corti method. Preferably, said method is extended to multimodel assemblies.” The examiner interprets where The synthesis step involving Bayesian statistical methods to compile probability density functions (PDFs) inherently minimizes variance (uncertainty) to produce the "optimum forecast.”) and selectively output a prediction on ocean conditions. ; ([Abstract]: “A climatic forecast system for generating an optimized forecast includes: a computer including a processor and a storage unit; means for remotely acquiring digital data via a digital network, said digital data including observations and at least one model digital forecast; and means for hybrid forecasting, including, means for inputting said observations, said at least one model digital forecast, and first forecast data, means for hybrid forecasting adapted to process said observations, said at least one model digital forecast, and to apply at least one of a statistical model, a simple general circulation model, and a multi-model synthesis algorithm to said first forecast data, and means for outputting at least one hybrid forecast.” [0034]: “synthesis means of said overall forecasts outputting at least one multimodel forecast, to which is associated a probability density of at least one environmental variable, said synthesis means including multimodel synthesis algorithms” [0084]: “Simple digital forecasts (33) are forecasts of one specific variable, either deterministic, or a probability density function of the variable, and are obtained with the aid of simple general circulation models (22). Said models are particularly well suited to a specific problem and have the advantage of requiring few computing means. They are, usually, specific to a particular region. Numerous simple general circulation models already exist. For example, for forecasting weather regimes for mid-latitudes, a simple atmospheric model, such as one of those which have been developed by Marshall and Molteni, may be used. The Cane-Zebiak model, developed by two scientists with the same name, may be used for forecasting the sea surface temperature in the Equatorial Pacific.” The examiner interprets where ocean conditions are shown in "sea surface temperature" listed as target environmental variables. ) Regarding Claim 2 Loukos teaches The system as set forth in claim 1 (See Claim 1). Loukos teaches wherein the analysis state (See [FIG. 1] & [0080]: “As illustrated by FIG. 1, all of said observations (12) and said single-model or multimodel digital forecasts (11) provide via the present invention an optimised climatic forecast (60). For this, a plurality of algorithms and models (20) are used on the observations and single-model or multimodel forecasts. Forecasts (30) issued by said algorithms and models are, then, optimised via two successive steps: downscaling (40) and benchmarking (50), a step consisting in evaluating the reliability and robustness of the forecasts.” [0091]: “In order to optimise the forecasts provided by the system according to the present invention, feedback mechanisms (51) are used at the time of the final synthesis. Said feedback takes place on the models and algorithms (20) and/or on the downscaling (40) module. The objective is to select the most relevant forecasts according to one or more predetermined criteria and to optimise the parameters of the various modules (20 and 40) based on the observations (12).” The examiner interprets where forecasts optimized through "feedback mechanisms (51)" that take the result of the "final synthesis" (analysis state) and feed it back to the "models and algorithms (20)” and where FIG. 1 shows a feedback arrow (51) from the "Benchmarking and Overall Synthesis (50)" step back to the primary forecasting modules is shown as analysis state. ) is used as an initial condition to perform forecasting ([0086]: “Statistical models (23) generate statistical forecasts (34), forecasts based on statistical forecasts determined by the historical analysis of available data obtained via observations. Said models are based on notions such as persistence, auto regression or analogy and produce combinations of the latter. Persistence consists in maintaining anomalies from the previous month or months. Using auto regressive models allows, for example, the future month to be forecast based on the current month and/or past months. Various types of auto regressive models exist, the most known of which are the ARMA (Auto Regressive Moving Average) type auto regressive models and neural auto regressive models which are based on neural networks. Analogue models produce a statistical link between “predictor” variables at month M and a “predicting” variable at month M+X, where X is a whole number. The analogue models are, for example, used for forecasting precipitations in Australia from the surface temperatures of the Austral Ocean during previous months as described in White 2000. Statistical forecasts thus generated provide a probability density of one variable and other statistical moments.” The examiner interprets where "auto regressive models" which allow a "future month to be forecast based on the current month” is shown as using the current state to drive a future prediction is the standard definition of providing an "initial condition.") or generate the ensembles [FIG. 1], [0081]: “By using a multimodel synthesis algorithm (21), single-model or multimodel digital forecasts, only, provide a multimodel forecast (31) which is, for example, a probability density of values of one variable, most probably from other statistical moments.” The examiner maps ensembles to multimodel forecast and interprets where the feedback loop (51) in Loukos directs information back to the "multimodel synthesis algorithm (21)" and "Hybrid algorithms (24);” generating the "multimodel forecast" (the ensemble), and feeding back the synthesized analysis state is shown as generating/updating the ensembles for the subsequent prediction cycle.) for a next cycle. (See [0091], [FIG. 1], [0081]: “By using a multimodel synthesis algorithm (21), single-model or multimodel digital forecasts, only, provide a multimodel forecast (31) which is, for example, a probability density of values of one variable, most probably from other statistical moments.” [0087]: “Finally, hybrid models (24) generate hybrid forecasts (32) in the form of value probability density for a variable at a point, from single-model or multimodel digital forecasts (11) and/or observation data (12). Said models are a combination and not a simple juxtaposition of at least two models taken among statistical models (23), simple general circulation models (22) and multimodel synthesis algorithms (21). More specifically, hybrid models combine the methodological approaches and tools according to which the previously mentioned models are constructed. For example, a hybrid model will generate a probability density function from a specific value (for example, the median) issued by a probability density generated from one of the previously mentioned models (for example, multimodel synthesis algorithms) and the geometry of a probability density function issued by one other type of model (for example, a statistical model). The PDF of the hybrid models will therefore be generated from a combination of at least two models and not from the combination of results from at least two or more models. Said example is not limited, diverse combination techniques may be used in hybrid models (as for example, discriminant multivariate analysis or data assimilation techniques) and therefore do not constitute a limit to their definition.” The examiner interprets where "data assimilation techniques" is shown as a method used in its hybrid models, which by definition involve updating model states for the next cycle.) Regarding Claim 3 Loukos teaches The system as set forth in claim 1 (See Claim 1). Loukos teaches wherein the stochastic data drive prediction model comprises a fully data driven model (See [0086]: The examiner interprets where “statistical models (23)" that generate "statistical forecasts (34)" being "based on statistical forecasts determined by the historical analysis of available data obtained via observations;" relying on historical data analysis is shown as the model as "fully data driven.") which is a stochastic variational model (See [0040] & [0090]: “As regards the overall synthesis, it more specifically relates to an analysis and a multi-method forecast synthesis. For example, it may be produced by comparing the probability density functions of forecasts issued by the preceding steps, according to the Kolmogorov Smirnov test, then by compiling probability density functions according to their reliability score. Bayesian statistical methods may also be used in order to perform the synthesis of various PDF.” The examiner interprets where forecasting and synthesis means (e.g., multimodel synthesis, statistical forecasting) "use Bayesian statistical methods” and "Bayesian statistical methods may also be used in order to perform the synthesis of various PDF [Probability Density Functions]. ” The examiner interprets where in the field of computational statistics, a model that performs Bayesian synthesis of probability distributions to approximate an optimal state is shown as a "stochastic variational model.") or a variational autoencoder (See [0086], [0015]: The examiner interprets where "neural auto regressive models which are based on neural networks" and "multilayer neural networks” is shown as a variational autoencoder. While the specific 2013-era term "variational autoencoder" is not used, the reference discloses the alternative "stochastic variational model" (see 3c), which satisfies the "OR" condition of the claim.) configured to predict a large number of ensembles of states of the ocean. (See [0014], [0034], [0084] [0082]: “Said algorithm uses a synthesis based on digital forecasts alone which may, as described in the Kernels method (Silverman 1986), be a Probability Density Function (PDF) in one point from individuals of a single-model assembly. Alternatively, said method may be extended to 2D or 3D models, as in the S. Corti method. Preferably, said method is extended to multimodel assemblies.” [0088]: “The method according to the present invention produces a spatial as well as temporal downscaling step (40) on the forecasts obtained by the previously described models, i.e. the multimodel (31), hybrid (32), simple digital (33) and statistical (34) models. Prior art already knows methods allowing forecast downscaling to be carried out. The downscaling step allows in particular small scale scenarios to be calculated, i.e. specific to one site, by means of a global climatic model (on a large scale). The forecast obtained by this step is in the form of value probability or value probability density of one variable at a point where the geographical and/or temporal forecast is clearly greater than that of the initial data. Three types of downscaling are distinguished. On the one hand, digital downscaling is based on regional digital models (MM5, HIRLAM, Méso NH), nested regional digital models (MM5) or CFD (Computational Fluid Dynamics) digital models (MésoNH, WRF). On the other hand, statistical downscaling produces a statistical relationship between either one model variable and one local variable, for example, via linear regression by means of one neural network or via a geostatistical interpolation technique (Kriging), or a predictor model variable and a predicting variable via analogue models which classify the structures (pattern) on a large scale and establishes links with local variables (SOM/Kohonen technique). Said structures may be atmospheric structures (pressure, temperature fields) or oceanic structures (surface temperature, sea current fields). Finally, hybrid downscaling is a statistical combination of at least one part of the fields having already undergone digital downscaling, via linear combination or not. A neural network combination technique (SOM Self Organizing Maps techniques), for example, may be used.” The examiner interprets where "multimodel superensemble" and "multimodel assemblies" configured to predict "environmental variables" such as "the sea temperature across specific oceanic regions" and "oceanic structures (surface temperature, sea current fields);" is shown as inherently involving a "large number" of ensemble members/model runs.) Regarding Claim 5 Loukos teaches The system as set forth in claim 3 (See claim 3). Loukos teaches wherein the fully data driven model is configured to add physical constraints (See [0087], [0067]: “In the field of climatic forecasting, a “digital model” means a mathematical equations resolution model for a system of physical (but sometimes also chemical and/or biological) laws on the initial conditions and at suitable limits via a digital method which allows integration of the system of equations in time and space.” The examiner interprets where "digital models" based on "physical... laws" and "hybrid models" that "combine the methodological approaches and tools according to which the [statistical and physical] models are constructed" is shown as the resulting model configured with physical constraints.) within its architecture (See [0087]: The examiner interprets where hybrid models are "a combination and not just a simple juxtaposition" of statistical and physical models is shown as constraints integrated into the model's structural design/architecture.) or through regularization. (See [0040] & [0090]: The examiner interprets where the use of "Bayesian statistical methods" to perform the synthesis of probability density functions is shown as the functional equivalent of regularization, as it penalizes deviations from the physically-consistent prior state.) Regarding Claim 6 Loukos teaches The system as set forth in claim 3 (See Claim 3). Loukos teaches wherein the fully data driven model is configured to perform robust uncertainty quantification (See [0032] & [0080]: The examiner interprets where forecasts outputting "in the form of probability density" and "evaluating the reliability and robustness of the forecasts" is shown as the fundamental method of performing uncertainty quantification.) through the statistics obtained from the large ensembles. (See [0014], [0082], [0081], [0032], [0034], [0084], [0086], [0087], [0090], : The examiner interprets where the use of "multimodel superensemble[s]" and "multimodel assemblies;” where these models provide a "multimodel forecast (31) which is, for example, a probability density of values of one variable, most probably from other statistical moments" and "compiling probability density functions according to their reliability score" using "Bayesian statistical methods" is shown as statistics obtained from the large ensembles.) Regarding Claim 7 Loukos teaches The system as set forth in claim 1 (See claim 1). Loukos teaches wherein the multi-layer perceptron is configured to perform a Lagrangian data assimilation method (See [FIG. 1], [0015], [0040], [0090], [0080], [0087]: The examiner interprets where its forecasting system utilizes “multilayer neural networks” to perform an “overall synthesis” and “benchmarking” of observations and forecasts and its hybrid models utilizing “data assimilation techniques” is shown as Lagrangian data assimilation.) capable of integrating distributed sensor observations from the ocean. (See [Abstract], [0084], [0001]: “The present invention relates to the field of climatic forecast computer systems using the observations provided by various measuring stations distributed across a geographical territory.” [0075]: “Thus, one-off measurements made at a specific geographical point are distinguished. For said measurements, surface sensors such as automatic weather stations, Synop or Metar, are used. Sounding balloons are also used for the air, or buoys for the sea environment.” [0077]: “As opposed to remote sensing on the ground, airborne remote sensing provides observations concerning the wind, for example, by means of an airborne LIDAR or an SAR (Synthetic Aperture Radar), but also sea surface temperature measurements or plant cover measurements, said measurement being made by satellites.” The examiner interprets where “measuring stations distributed across a geographical territory” and “means for remotely acquiring digital data... digital data including observations” and the identification of ocean-based sensors, including “buoys for the sea environment” and “satellites” measuring “sea surface temperature” is shown as integrating distributed sensor observations from the ocean.) Regarding Claim 8 Loukos teaches The system as set forth in claim 7 (See claim 7). Loukos teaches wherein the multi-layer perceptron is non-linear (See [0015], [0086], [0088], [0042]: “According to one variant, said forecasting and statistical processing means perform non-linear regression for each neural network.” The examiner interprets where "neural networks firstly provide a non-linear approach of the analysis" and the system performs "non-linear regression for each neural network” is shown as the multi-layer perceptron being non-linear.) and free of ad-hoc choices in de-correlation lengths. (See [0015], [0086], [0088], [0087]: The examiner interprets where its hybrid models utilize "data assimilation techniques" and "multilayer neural networks" to synthesize observations and forecasts is shown as the use of an MLP to learn the mapping from observations to an analysis state necessarily; removing the requirement for the ad-hoc de-correlation length scales mandated by traditional parametric DA methods.) Regarding Claim 9 Loukos teaches The system as set forth in claim 7 (See claim 7). Loukos teaches wherein the multi-layer perceptron is implemented on the same device ([0015], [0086], [0081] [0007]: “The present invention relates more specifically to a climatic forecast system, using at least one computer comprising in particular a processor and a storage unit, means for remotely acquiring digital data via a computer network or a telecommunication network. According to a first aspect, the present invention consists in particular in synthesizing the forecasts issued by operational centres and transposing them on the local scale by adding the information issued by specifically developed statistical models. The principle of statistical models is based on connecting land observation histories provided by national or private weather centres, with large scale data provided by private or public operational centres.” The examiner interprets where a climatic forecast system “using at least one computer comprising in particular a processor and a storage unit” is shown as the implementation site for the various modules, including the “statistical models (23)” and the “multilayer neural network” used for synthesis.) as the fully data driven model (See [FIG. 1, [Abstract], [0086], [0041] [0007]: The examiner interprets where the “statistical models (23)” (the fully data-driven models) and the “neural networks” (the MLP for DA) are as parts of the same “climatic forecast system” executed by the “processor” and FIG. 1 illustrating these components within a single logical and physical system boundary is shown as the fully data driven model.) allowing for on-the-fly data assimilation. (See [FIG. 1], [0091] [0065]: “According to one preferred embodiment, the system according to the invention uses neural networks for statistical analysis in the climatic field. Said neural networks firstly provide a non-linear approach of the analysis and rapid processing of a large data flow.” The examiner interprets where neural network-based system providing for “rapid processing of a large data flow” and “feedback mechanisms (51)” that operate at the time of synthesis to “optimise the parameters of the various modules” based on observations; where collocating these components on a single computer to perform rapid processing and recursive feedback is shown as the system inherently performs “on-the-fly” assimilation.) Regarding Claim 10 Loukos teaches A method comprising: generating data ensembles using a stochastic data driven prediction model trained on ocean simulations; receiving observation data; estimating an analysis state using a multi-layer perceptron that represents the analysis state as a non-linear conjunction of observations and minimizing variance across the ensembles; and selectively outputting a prediction on ocean conditions. (Method version of Claim 1. Similar rejection as Claim 1) Regarding Claim 11 Loukos teaches The method as set forth in claim 10, wherein the analysis state is used as an initial condition to perform forecasting or generate the ensembles for a next cycle. (Method version of Claim 2. Similar rejection as Claim 2) Regarding Claim 12 Loukos teaches The method as set forth in claim 10, (Method version of Claim 3. Similar Rejection to Claim 3) wherein the stochastic data drive prediction model comprises a fully data driven model (Method version of Claim 3. Similar Rejection to Claim 3) which is a stochastic variational model (Method version of Claim 3. Similar Rejection to Claim 3) or a variational autoencoder (Method version of Claim 3. Similar Rejection to Claim 3) configured to predict a large number of ensembles of states of the ocean (Method version of Claim 3. Similar rejection as Claim 3) at low computational cost. (See [0084] & [0029]: “One of the main aims of the present invention consists in producing the most reliable possible local forecast, and effectively, i.e. in particular with reduced computing times, and with reasonable technical means.” The examiner interprets where the main aim is to provide forecasts with “reduced computing times” and “reasonable technical means” and its statistical models “have the advantage of requiring few computing means” compared to high-power computers is shown as low computational cost. Regarding Claim 14 Loukos teaches The system as set forth in claim 11, wherein the fully data driven model is configured to add physical constraints within its architecture or through regularization. (Method version of Claim 5. Similar rejection as Claim 5) Regarding Claim 15 Loukos teaches The system as set forth in claim 11, wherein the fully data driven model is configured to perform robust uncertainty quantification through the statistics obtained from the large ensembles. (Method version of Claim 6. Similar rejection as Claim 6) Regarding Claim 16 Loukos teaches The system as set forth in claim 10, wherein the multi-layer perceptron is configured to perform a Lagrangian data assimilation method capable of integrating distributed sensor observations from the ocean. (Method version of Claim 7. Similar rejection as Claim 7) Regarding Claim 17 Loukos teaches The system as set forth in claim 16, wherein the multi-layer perceptron is non-linear and free of ad-hoc choices in de-correlation lengths. (Method version of Claim 8. Similar rejection as Claim 8) Regarding Claim 18 Loukos teaches The system as set forth in claim 16, wherein the multi-layer perceptron is implemented on the same device as the fully data driven model allowing for on-the-fly data assimilation. (Method version of Claim 9. Similar rejection as Claim 9) Regarding Claim 19 Loukos teaches A non-transitory computer readable medium having stored thereon instructions that, when executed by a processor, cause a system to: generate data ensembles using a stochastic data driven prediction model trained on ocean simulations; receive observation data; estimate an analysis state using a multi-layer perceptron that represents the analysis state as a non-linear conjunction of observations and minimizing variance across the ensembles; and selectively output a prediction on ocean conditions. (Articles of Manufacture version of Claim 1. Similar rejection as Claim 1) Regarding Claim 20 Loukos teaches The non-transitory computer readable medium as set forth in claim 19, wherein the analysis state is used as an initial condition to perform forecasting or generate the ensembles for a next cycle. (Articles of Manufacture version of Claim 2. Similar rejection as Claim 2) Claim Rejections - 35 USC § 103 In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action. The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention. Claim(s) 4 and 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over US Patent No. US20080167822A1 by Loukos further in view of NPL: "Assimilation of drifter observations for the reconstruction of the Eulerian circulation field" by Molcard (2003). Regarding Claim 4 Loukos teaches The system as set forth in claim 3 (See claim 3). Loukos fails to teach wherein the fully data driven model comprises a 4th order Runge Kutta (RK4) based time-integrator for accurate estimation of future time steps with low error growths. However; Molcard teaches wherein the fully data driven model comprises a 4th order Runge Kutta (RK4) based time-integrator ([P. 6 §3.1 ¶1]: “[27] In order to exactly quantify the performance of the data assimilation scheme, the identical twin experiment approach is used (Figure 1). First, the model is integrated for 20 years until the Rossby waves cross the basin, enhance the western boundary current and the midlatitude jet, and the model energetics reach a statistically steady state. Synthetic surface drifters are then launched in the model and advected with the model velocity field using a fourth- order Runge–Kutta scheme with the same time step of 1.6 hours used for the model integration.” [FIG. 1]: PNG media_image2.png 640 740 media_image2.png Greyscale The examiner interprets where state variables in an oceanographic model are “advanced with the model velocity field using a fourth order Runge–Kutta scheme” is shown as Fully data driven model comprises a 4th order Runge Kutta (RK4) based time-integrator) for accurate estimation of future time steps with low error growths. (See [FIG. 1]], [P. 6 §3.1 ¶1], [FIG. 2] : PNG media_image3.png 694 866 media_image3.png Greyscale The examiner interprets where RK4 in an “identical twin experiment” framework to provide a benchmark “control” run is shown as illustrating its recognized role in providing accurate estimation of paths and future states.). It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to implement the temporal integration described in Loukos using the RK4 scheme taught by Molcard. The motivation to do so would be to improve the numerical accuracy of the forecast and minimize error accumulation over time steps, as RK4 is a well-known, higher-order method that yields the predictable result of increased precision in dynamical simulations compared to lower-order integration methods. This represents the application of a known technique (RK4) to a known system (oceanic forecasting) ready for improvement to yield predictable results [MPEP 2143.01(D)]. Regarding Claim 13 Loukos teaches The method as set forth in claim 10, wherein the fully data driven model comprises a 4th order Runge Kutta (RK4) based time-integrator for accurate estimation of future time steps with low error growths. (Method version of Claim 3. Similar rejection as Claim 3) Motivation to combine similar to Claim 4. See claim 4. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to AARIC RAYJEE MARKS whose telephone number is (571)467-6372. The examiner can normally be reached Monday-Friday 8am-5pm. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Ryan Pitaro can be reached at (571) 272-4071. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000. /AARIC R MARKS/Examiner, Art Unit 2188 /AKASH SAXENA/Primary Examiner, Art Unit 2188