Maximum Cross-correlation Technique Research Articles

Optimization of Sea Surface Current Retrieval Using a Maximum Cross-Correlation Technique on Modeled Sea Surface Temperature

AbstractUsing sea surface temperature from satellite images to retrieve sea surface currents is not a new idea, but so far its operational near-real-time implementation has not been possible. Validation studies are too region specific or uncertain, sometimes because of the satellite images themselves. Moreover, the sensitivity of the most common retrieval method, the maximum cross correlation, to the parameters that have to be set is unknown. Using model outputs instead of satellite images, biases induced by this method are assessed here, for four different seas of western Europe, and the best of nine settings and eight temporal resolutions are determined. The regions with strong currents return the most accurate results when tracking a 20-km pattern between two images separated by 6–9 h. The regions with weak currents favor a smaller pattern and a shorter time interval, although their main problem is not inaccurate results but missing results: where the velocity is too low to be picked by the retrieval. The results are not impaired by the restrictions imposed by ocean surface current dynamics and available satellite technology, indicating that automated sea surface current retrieval from sea surface temperature images is feasible, for pollution confinement, search and rescue, and even for more energy-efficient and comfortable ship navigation.

Application of multi-window maximum cross-correlation to the mediterranean sea circulation by using MODIS data

In a previous study an improved Maximum Cross-Correlation technique, called Multi-Window MaximumCross-Correlation (MW-MCC), was proposed, and applied to noise-free synthetic images in order to show its potentialand limits in oceanographic applications. In this work, instead, the application of MW-MCC to high resolution MODISimages, and its capability to provide useful and realistic results for ocean currents, is studied.When applied to real satellite images, the MW-MCC is subject to cloud cover and image quality problems. As a consequence the number of useful MODIS images is greatly reduced. However, for every MODIS image, multiple spectral bands are available, and it is possible to apply the MW-MCC algorithm to the same scene as many times as thenumber of these bands, increasing the possibility of finding valid current vectors.Moreover, the comparison among the results from different spectral bands allows to verify both the consistency ofthe computed current vectors and the validity of using a spectral band as a good tracer for the ocean circulation.Due to the lack of systematic current measurements in the area considered, it has been not possible to perform an extensive error analysis of the MW-MCC results, although a case study of a comparison between HF radar measurementsand MW-MCC data is shown. Moreover, some comparison between numerical ocean model simulations and MW-MCCresults are also shown. The coherence of the resulting circulation flow, the high number of current vectors found, theagreement among different spectral bands, and conformity with the currents measured by the HF radars or simulated byhydrodynamic models show the validity of the technique.

Application of multi-window maximum cross-correla-tion to the mediterranean sea circulation by using MODIS data

In a previous study an improved Maximum Cross-Correlation technique, called Multi-Window Maximum Cross-Correlation (MW-MCC), was proposed, and applied to noise-free synthetic images in order to show its potential and limits in oceanographic applications. In this work, instead, the application of MW-MCC to high resolution MODIS images, and its capability to provide useful and realistic results for ocean currents, is studied. When applied to real satellite images, the MW-MCC is subject to cloud cover and image quality problems. As a con-sequence the number of useful MODIS images is greatly reduced. However, for every MODIS image, multiple spec-tral bands are available, and it is possible to apply the MW-MCC algorithm to the same scene as many times as the number of these bands, increasing the possibility of finding valid current vectors. Moreover, the comparison among the results from different spectral bands allows to verify both the consistency of the computed current vectors and the validity of using a spectral band as a good tracer for the ocean circulation. Due to the lack of systematic current measurements in the area considered, it has been not possible to perform an ex-tensive error analysis of the MW-MCC results, although a case study of a comparison between HF radar measurements and MW-MCC data is shown. Moreover, some comparison between numerical ocean model simulations and MW-MCC results are also shown. The coherence of the resulting circulation flow, the high number of current vectors found, the agreement among different spectral bands, and conformity with the currents measured by the HF radars or simulated by hydrodynamic models show the validity of the technique.

Extensive analysis of potentialities and limitations of a maximum cross-correlation technique for surface circulation by using realistic ocean model simulations

As shown in the literature, ocean surface circulation can be estimated from sequential satellite imagery by using the maximum cross-correlation (MCC) technique. This approach is very promising since it offers the potential to acquire synoptic-scale coverage of the surface currents on a quasi-continuous temporal basis. However, MCC has also many limits due, for example, to cloud cover or the assumption that Sea Surface Temperature (SST) or other surface parameters from satellite imagery are considered as conservative passive tracers. Also, since MCC can detect only advective flows, it might not work properly in shallow water, where local heating and cooling, upwelling and other small-scale processes have a strong influence. Another limitation of the MCC technique is the impossibility of detecting currents moving along surface temperature fronts. The accuracy and reliability of MCC can be analysed by comparing the estimated velocities with those measured by in situ instrumentation, but the low number of experimental measurements does not allow a systematic statistical study of the potentials and limitations of the method. Instead, an extensive analysis of these features can be done by applying the MCC to synthetic imagery obtained from a realistic numerical ocean model that takes into account most physical phenomena. In this paper a multi-window (MW-) MCC technique is proposed, and its application to synthetic imagery obtained by a regional high-resolution implementation of the Regional Ocean Modeling System (ROMS) is discussed. An application of the MW-MCC algorithm to a real case and a comparison with experimental measurements are then shown.

Retrieval of coastal ocean currents using MCC technique on satellite imagery for supplementing altimeter derived currents

Abstract. Maximum Cross Correlation (MCC) technique have been used to derive coastal currents from sequential OCM derived suspended sediment concentration (SSC) maps and the MCC currents are integrated with the Ocean surface current analysis-Real time (OSCAR) ocean currents along the coastal regions of the Southern India. OSCAR currents are mainly depended on the oceanographic measurements from the satellite altimeter which have limitations in the coastal region and in representing the mesoscale properties dominating the coastal ocean. Sequential pair of atmospherically corrected and georeferenced OCM image is used to derive SSC, along the southern coast of India. Maximum Cross-Correlation (MCC) method is applied on the sequential time lapsed images of OCM for matching the suspended sediment dispersion patterns. The MCC technique involves computing of cross-correlation coefficients and identifying correlation peaks. The spatial change of the sediment pattern occurs due to the advection by the currents and the extent of the change corresponds to the speed and direction of the current. From the first image, template window is selected, which is searched in the second image within large region defined as the search window. The movement of the SSC pattern is calculated based on the maximum cross correlation in successive images. Error statistics are calculated between the spatially coherent MCC vectors and OSCAR currents. Based on statistical analysis, an error threshold of 30 degree for current direction and 0.06m/sec for current speed was given to MCC currents and MCC currents beyond the error limit are rejected, while the MCC current within the error limit is merged with the OSCAR ocean currents. The merging of MCC currents has spatially enhanced the OSCAR currents. MCC method is significant in calculating the coastal currents and its mesoscale properties, which dominate the coastal surface flow field and can be used in supplementing the OSCAR ocean currents. Merging of both the currents, have effectively resulted in preparing contiguous maps of coastal ocean currents.

Winter and spring surface velocity fields in the Cape Blanc region as deduced with the maximum cross-correlation technique

The ocean surface velocity field in the Cape Blanc region, off Northwest Africa, is investigated with the maximum cross-correlation (MCC) method applied to channel-4 Advanced Very High Resolution Radiometer satellite images. An initial sensitivity analysis allows us to select the four parameters that provide maximum area coverage and the best velocity resolution, while limiting the standard deviation for each velocity component within reasonable values. These are (m, n, MV, CT) = (22, 32, 50, 0.6), where m and n are the number of pixels of the search window (SW) and reference window (RW), respectively, MV is the maximum possible velocity (in cm s−1), and CT is a correlation threshold for a feature to be tracked. A total of 489 images, for years 2005 and 2006, are analysed, and 106 velocity maps are generated with good coverage of the coastal transition zone (CTZ), most of them for the winter (34) and spring (59) seasons. We remove spurious data using the method's own filters (MV, CT, and a neighbour-vector comparison), requesting the velocity components to have Gaussian distributions and smoothing the resulting velocity fields with a median-vector filter. The instantaneous velocity maps illustrate the response of the alongshore coastal jet north of Cape Blanc (and its extension along the Cape Verde frontal region) to wind forcing, as well as the presence of numerous mesoscalar features (100–300 km wide) superposed on a westward offshore transport south of Cape Blanc. We also produce mean and standard deviation winter and spring velocity maps, which are compared with the corresponding mean sea surface temperature fields. The along-shore and offshore flow is better defined and is more intense in spring than in winter, in concordance with cross-slope sharper temperature gradients during this season, and brings about a cooling of the whole region. We identify five different ubiquitous currents: a southwestward jet north of Cape Blanc, a northwestward jet off Banc d'Argin, an offshore convergent jet, a spring jet-like feature at 18° N, and a southward flow in the southwestern CTZ.

Velocity observations of the California Current derived from satellite imagery

This research examines the potential of ocean velocity observations generated from the maximum cross‐correlation technique (MCC) applied to thermal infrared imagery to capture the total current at a high sampling rate, especially in the nearshore region, allowing for a novel view of the California Current. Comparison of 12 years of weekly ocean velocity fields derived from the MCC method with velocity observations from drifting buoys reveals strong correspondence between the data sets. The MCC method, however, is able to produce over ∼10 times the number of observations. Comparison of MCC velocities with geostrophic velocities from altimetry demonstrates differences that suggest ageostrophic currents in the MCC observations that are likely wind driven. The time‐averaged mean velocity field from MCC observations reveals strong offshore jet‐like features that extend off of coastline promontories. Eddy statistics from the MCC velocity observations indicate dramatically different dynamics in the nearshore and offshore regions. Differences in spatial statistics derived from the MCC observations compared to previous studies of the region are attributed to wind‐driven currents in the observations that can dominate the variability of the current in the nearshore region.

Applying Satellite-Estimated Storm Rotation Speed to Improve Typhoon Rainfall Potential Technique

Abstract Heavy rainfall from typhoons or tropical cyclones often causes inland flooding and mudslides that threaten lives and property. In this study, Special Sensor Microwave Imager (SSM/I) satellite data observed from 2000 to 2004 were used to calculate the rainfall rates of different typhoons in the northwestern Pacific. Geostationary weather satellite infrared images were also applied to estimate the typhoon rotation speed via the maximum cross-correlation technique. By including such information in the tropical rainfall potential (TRaP) technique, an improved typhoon rainfall potential technique can be constructed. Considering the fact that a typhoon’s spiral rainbands move constantly, half-hourly or hourly infrared data observed from geostationary weather satellites were used to calculate the revolving speed, which was subsequently used to predict the rainband movement over the next hour. After comparing the predicted rainfall potential with the rain gauge data of Taiwan’s small offshore islands, it was found that this new method can improve the typhoon’s accumulated rainfall by approximately 40% over the original TRaP method. Therefore, to produce a more accurate short-term typhoon rainfall forecast, it is very important to factor in the satellite-estimated storm rotation speed.

The Potential of the Maximum Cross-Correlation Technique to Estimate Surface Currents From Thermal AVHRR Global Area Coverage Data

Having already shown its potential of deriving the vector fields representing the ocean-surface advection from sequential 1.1-km-resolution local area coverage (LAC) Advanced Very High Resolution Radiometer (AVHRR) images, the maximum cross-correlation (MCC) technique here is applied to four 4.4-km-resolution global area coverage (GAC) AVHRR images. The resulting three vector fields are compared to the vector fields obtained from the LAC imagery corresponding to the same satellite passages. To quantify the reduction in accuracy inevitable when applying the method to the lower resolution imagery, the LAC vector fields were assumed to be error free. The deviation of the GAC vectors from the LAC vectors is expressed as percentage errors of the signal variance of meridional u and zonal v velocity components, and they are 16%/30%, respectively, for the best case and 62%/117% and 92%/111% for the other two cases. These results indicate that, in its present state, the GAC data do not allow the MCC technique to extract reliable current-vector information from it

Extracting Multiyear Surface Currents from Sequential Thermal Imagery Using the Maximum Cross-Correlation Technique

Ocean surface circulation can be estimated by automated tracking of thermal infrared features in pairs of sequential satellite imagery. A 7-yr time series of velocity, extracted from thermal imagery of the East Australian Current using the maximum cross-correlation (MCC) technique, provides enough measurements for a more statistical evaluation of the method than has previously been possible. Excluding 1 yr with extensive cloud cover, the method produces about 8000 velocity estimates per month with some seasonal variation. Method precision is estimated to be between 0.08 and 0.2 m s21 rms, the lower value with more restrictive compositing. Mean flow, time-dependent flow, and eddy kinetic energy from the time series are compared with values derived from a dynamic height climatology, altimeter analyses, and drifter datasets in the region. The observations reproduce similar features in the flow. The differences between the observations are discussed in relation to noise in the methods and differences in the types of velocities they measure.

Continuity preserving modified maximum cross‐correlation technique

The maximum cross‐correlation (MCC) method reconstructs the surface advective velocity fields from the displacements of spatial patterns in pairs of sequential satellite (normally infrared) images. However, the performance of the conventional MCC method is not always satisfactory. One of the main reasons for this is the fact that the method can correctly estimate only the velocity component parallel to the gradient of the property depicted in the images, while any small displacement perpendicular to the gradient (i.e., directed along the isolines) essentially maps the spatial pattern onto itself and therefore can not be detected using the conventional MCC technique. In the present work we propose a modification of the MCC method that allows circumventing this basic deficiency and improving the performance of the MCC technique. In this approach, the “cross‐isoline” components of the velocity field are obtained as in the conventional MCC scheme; however, the “along‐isoline” components derived from the MCC are disregarded as unreliable. Instead, the “true” along‐isoline components are then reconstructed from the given cross‐isoline velocity field based on the continuity requirement and on the condition of no normal flow at solid boundaries. This inverse problem is solved by constructing the two‐dimensional stream function in the curvilinear coordinate frame associated with the image isolines. The method is illustrated using AVHRR images from the southwestern Atlantic Ocean and the Black Sea. The results are compared with some direct drifter and current meter measurements and geostrophic estimates.

Retrieval of sea surface velocities using sequential Ocean Colour Monitor (OCM) data

The Indian remote sensing satellite, IRS-P4 (Oceansat-I) launched on May 26th, 1999 carried two sensors on board, i.e., the Ocean Colour Monitor (OCM) and the Multi-frequency Scanning Microwave Radiometer (MSMR) dedicated for oceanographic research. Sequential data of IRS-P4 OCM has been analysed over parts of both east and west coast of India and a methodology to retrieve sea surface current velocities has been applied. The method is based on matching suspended sediment dispersion patterns, in sequential two time lapsed images. The pattern matching is performed on a pair of atmospherically corrected and geo-referenced sequential images by Maximum Cross-Correlation (MCC) technique. The MCC technique involves computing matrices of cross-correlation coefficients and identifying correlation peaks. The movement of the pattern can be calculated knowing the displacement of windows required to match patterns in successive images. The technique provides actual flow during a specified period by integrating both tidal and wind influences. The current velocities retrieved were compared with synchronous data collected along the east coast during the GSI cruise ST-133 of R.V. Samudra Kaustubh in January 2000. The current data were measured using the ocean current meter supplied by the Environmental Measurement and CONtrol (EMCON), Kochi available with the Geological Survey of India, Marine Wing. This current meter can measure direction and magnitude with an accuracy of ±5‡ and 2% respectively. The measurement accuracies with coefficient of determination (R 2) of 0.99, for both magnitude (cm.s-1) and direction (deg.) were achieved.

Short time‐span ice tracking using sequential AVHRR imagery

As part of an international study of the North Water Polynya, the Royal Military College of Canada (BMC) installed a Quorum Advanced Very High Resolution Radiometer (AVHRR) ground station at Canadian Forces Station Alert (82°31'N 62°17'W), situated on the northern tip of Ellesmere Island. The extreme northern location of the station permitted image reception every 1.7 hours. A total of 1440 images were recorded and stored between March and August 1998. Of particular interest was a rapidly sampled sequence on 16 and 17 June that documented the rapid dissolution of ice in Nares Strait, extending from the Lincoln Sea to Smith Sound. This allowed a systematic investigation of the effect of the time increment in calculating surface flow speeds in an energetic regime. A maximum cross‐correlation (MCC) ice‐tracking algorithm was applied to 10 images to study the southward movement of ice through the strait. The short time span between sequential images, ranging from 1.7 to 6.7 hours, provided a unique dataset for the MCC technique. The results revealed a relationship between higher cross‐correlation coefficients (ρ) and decreasing time spans. The best results were obtained for a 5‐hour time span with ρ > 0.5. Additionally, it was found that lower values of p can be expected in areas of increased ice congestion. Assessment of the ice‐tracking algorithm was carried out by visually comparing the movement of selected floes to tracker results.

Estimation of velocity fields at the estuary-coastal interface through statistical analysis of successive airborne remotely sensed images

The study of surface water velocity fields through in situ sampling is intrinsically difficult because they are highly variable in time and space. With airborne remote sensing, however, it is possible to determine changes in velocity fields because spatially and temporally comprehensive data may be obtained. This letter shows how changes in the statistical properties of successive remotely sensed images may be used to estimate velocity vectors associated with chlorophyll-a and sea surface temperature (SST). The study area is Kirkcudbright Bay, a small estuary in south-west Scotland. Multi-temporal imagery of the study area were acquired by the NERC Daedalus ADDS-1268 Airborne Thematic Mapper (ATM) and processed to show chlorophyll and thermal indices as substitutes for chlorophyll-a and SST. Velocity fields were estimated by the Maximum Cross Correlation technique. Complex patterns were found, confirming that the comprehensive coverage provided by airborne remote sensing is required for their analysis. The chlorophyll-a velocity field differed from the SST velocity field, suggesting that these fields are relevant to the water quality variable in question, and not necessarily the water body itself.

Advective Surface Velocities Derived from Sequential Infrared Images in the Southwestern Atlantic Ocean

Sequential infrared satellite images between November 1995 and August 1996 have been used to derive sea surface advective velocities in the southwestern Atlantic Ocean using the Maximum Cross-Correlation (MCC) technique. The infrared images, geometrically corrected and accurately registered to a map grid, are masked for land and cloud cover. The linear displacements of the sea surface thermal features are determined by locating the maximum cross-correlation in windowed portions of successive images. The advective velocities are estimated for each pair of images based on regular grids. Incoherent vectors are eliminated from these fields using a statistical method. After this, a visual inspection by the operator is completed to remove the remaining fictitious vectors. The region studied includes the southern Brazilian, Uruguayan, and northern Argentinian shelves and the neighboring oceanic area known as the Brazil-Malvinas Confluence. Despite the rather complex circulation of this region, the advective surface currents calculated by the MCC method are consistent with the shapes of the sea surface temperature features. The general patterns of the surface circulation estimated by the MCC agree well with previous studies that used direct measurements, hydrographic data, and/or modeling efforts. In the shallower areas of the study region, the MCC technique is less applicable due to the predominance, in violation of the assumptions of the method, of nonadvective processes.

Observations of coastal filaments in the Adriatic Sea

At the end of August 1990, 1991 and 1992, temperature readings from satellite AVHRR images provided a description of cold filaments departing from the Croatian coast directed offshore in the Adriatic Sea. The main filament was observed in the middle Adriatic extending from the Croatian to the Italian side along a topographic escarpment, giving rise to a double recirculation cell. Surface velocities associated to the offshore flow, estimated with the maximum cross-correlation technique, exhibited values on the order of 10–20 cm/s. ECMWF wind data indicate the Adriatic was characterized by winds with intensity of 5–7 m/s (breeze superposed to the climatological signal) excluding the eventuality of a wind driven filament. The filament formation process was also analyzed in terms of residual currents associated to tidal rectification processes. Comparisons between velocity values predicted by this theory and observations indicate the dynamics associated to the observed structures cannot be explained in terms of the theory of tidal residual flows. Although much more experimental evidence is required to finally prove the following hypothesis, it is shown that the baroclinic extension of the theory of the Rossby adjustment over a bottom escarpment can ascribe the filament formation in the middle Adriatic to interaction of steady coastal currents with bathymetry.

Computing advective velocities from satellite images of sea surface temperature

A method of employing sequential satellite images of sea surface temperature (SST) patterns to estimate surface advective current velocities is described. SST images are obtained by processing data received twice daily from the Advanced Very High Resolution Radiometer. The current velocities are estimated by applying a maximum cross correlation (MCC) technique on two time-lapsed images. The MCC technique involves computing matrices of cross correlation coefficients and identifying correlation peaks. A two-stage multilevel statistical test over two-dimensional correlations is developed for determining the relative significance levels of velocity estimates. The test also identifies areas where the MCC technique cannot be effective. Aspects of implementation and limitations of the MCC technique for computing current velocity are also discussed. Advective velocity fields computed through MCC for the Chatham Rise area in New Zealand extending from (40.5 degrees S, 173.5 degrees E) to (49 degrees S, 178 degrees W) show good agreement with the known geostrophic flow patterns in this area.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>