System and method for correcting a model-derived vertical structure of ocean temperature and ocean salinity based on velocity observations

Patent metadata

Field	Value
Publication number	US-9811614-B2
Application number	US-201314023030-A
Country	US
Kind code	B2
Filing date	Sep 10, 2013
Priority date	Mar 13, 2013
Publication date	Nov 7, 2017
Grant date	Nov 7, 2017

Abstract

System and method for correcting the vertical structure of the ocean temperature and salinity based on velocity observations. Three relations that can be precomputed are exploited: (1) the relation between temperature and salinity throughout a water column, (2) the relation between temperature/salinity and geopotential, and (3) the relation between geopotential and velocity. The relations are stored in a form that allows efficient application through a cross-correlation matrix.

First claim

What is claimed is:

1. A method for predicting eddies comprising:
computing a historical covariance matrix of a vertical structure of ocean temperature and ocean salinity at each grid point using in situ temperature and in situ salinity, wherein the in situ temperature and the in situ salinity are obtained from a plurality of sensors deployed at a geographic location;
computing geopotential heights along a water column at the geographic location based on the vertical structure;
including the geopotential height anomalies in the historical covariance matrix to determine a historical geopotential;
computing horizontal relationships for the historical geopotential between the geopotential height anomalies and velocity observations from the geographic location;
obtaining a predicted temperature and a predicted salinity in the water column from a database;
correcting the predicted temperature and the predicted salinity in the water column based on a covariance relationship of the ocean temperature and ocean salinity between the historical geopotential and the historical covariance matrix;
using the corrected predicted temperature and the corrected predicted salinity to generate a numerical model forecast for the geographic location; and
predicting the eddies at the geographic location based on the numerical model forecast.

2. The method as in claim 1 further comprising:
producing a matrix of eigenvectors based on decomposing the historical covariance matrix through eigenvalue decomposition;
storing a set of most significant eigenvectors of the matrix of eigenvectors, the set being sized up to a pre-selected value;
computing a TS standard deviation matrix based on the in situ temperature and the in situ salinity;
computing a geopotential standard deviation of the geopotential height anomalies; and
reconstructing the historical covariance matrix based on the set of most significant eigenvectors, the TS standard deviation matrix, and the geopotential standard deviation.

3. A system for predicting eddies comprising:
a matrix processor computing a historical covariance matrix of a vertical structure of ocean temperature and ocean salinity at each grid point using in situ temperature and in situ salinity, wherein the in situ temperature and the in situ salinity are obtained from a plurality of sensors deployed at a geographic location;
a geopotential processor computing geopotential heights along a water column at the geographic location based on the vertical structure, the geopotential processor including the geopotential height anomalies in the historical covariance matrix to determine a historical geopotential;
a correction processor computing horizontal relationships for the historical geopotential between the geopotential height anomalies and velocity observations from the geographic location, the correction processor obtaining a predicted temperature and a predicted salinity in the water column from a database, the correction processor correcting the predicted temperature and the predicted salinity in the water column based on a covariance relationship of the ocean temperature and ocean salinity between the historical geopotential and historical covariance matrix, and
a numerical model processor using the corrected predicted temperature and the corrected predicted salinity to generate a numerical model forecast for the geographic location, the numerical ocean prediction model predicting the eddies at the geographic location based on the numerical model forecast.

4. The system as in claim 3 wherein the geopotential processor comprises producing a matrix of eigenvectors based on decomposing the historical covariance matrix through eigenvalue decomposition, storing a set of most significant eigenvectors of the matrix of eignevectors, the set being sized up to a pre-selected value, computing a TS standard deviation matrix of the in situ temperature and the in situ salinity, computing a geopotential standard deviation of the vertical geopotential structure, and reconstructing the historical covariance matrix based on the set of most significant eigenvectors, the TS standard deviation matrix, and the geopotential standard deviation.

5. The method as in claim 2 wherein the pre-selected value is 6.

6. The system as in claim 4 wherein the pre-selected value is 6.

7. A computer system for predicting eddies comprising computer instructions stored on non-transitory computer readable media to:
compute a historical covariance matrix of a vertical structure of ocean temperature and ocean salinity at each grid point using in situ temperature and in situ salinity, wherein the in situ temperature and the in situ salinity are obtained from a plurality of sensors deployed at a geographic location;
compute geopotential heights along a water column at the geographic location based on the vertical structure;
include the geopotential height anomalies in the historical covariance matrix to determine a historical geopotential;
compute horizontal relationships for the historical geopotential between the geopotential height anomalies and velocity observations from the geographic location;
obtain a predicted temperature and a predicted salinity in the water column from a database;
correct the predicted temperature and the predicted salinity in the water column based on a covariance relationship of the ocean temperature and ocean salinity between the historical geopotential and the historical covariance matrix;
use the corrected predicted temperature and the corrected predicted salinity to generate a numerical model forecast for the geographic location; and
predict the eddies at the geographic location based on the numerical model forecast.

8. The computer system as in claim 7 further comprising computer instructions stored on non-transitory computer readable media to:
produce a matrix of eigenvectors based on decomposing the historical covariance matrix through eigenvalue decomposition;
store the most significant eigenvectors of the matrix of eigenvectors, the set being sized up to a pre-selected value;
compute a TS standard deviation matrix of the in situ temperature and the in situ salinity observations;
compute a geopotential standard deviation of the vertical geopotential structure; and
reconstruct the historical covariance matrix based on the most significant eigenvectors, the TS standard deviation matrix, and the geopotential standard deviation.

9. The computer system as in claim 8 wherein the pre-selected value is 6.

Assignees

• Jacobs Gregg Arthur
• Smith Scott R
• Helber Robert W
• Us Navy

Inventors

• Jacobs Gregg Arthur
• Smith Scott R
• Helber Robert W

Classifications

• G01V11/00

  Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00 · CPC title

• G06F2111/10

  Numerical modelling · CPC title

• G06F30/20 Primary

  Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) · CPC title

• G01W1/10

  Devices for predicting weather conditions (computers per se G06; display devices G09) · CPC title

• G06F17/16

  Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} · CPC title