Wavelet analysis of satellite images for coastal watch

Liu, A.K.; Peng, C. Y.; Chang, Seokcheol

doi:10.1109/48.557535

Cited by 85 publications

(24 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent investigations in physical oceanography have demonstrated the efficiency and utility of wavelet transforms to analyze nonlinear dynamical ocean systems. For example, a two-dimensional Gaussian wavelet (often referred to as a "Mexican-hat") transform has been applied to satellite images (SAR, AVHRR, and ocean color) to separate various scale processes including relative phase/location information for coastal watch applications [40] and for ice edge and ice floe tracking [41]. Wavelet transforms of satellite images can be used for near real-time "quick look" analyses of satellite data for feature detection, for data reduction using a binary image, and image enhancement by edge linking.…”

Section: Sea Ice Motion Trackingmentioning

confidence: 99%

Cryosphere applications of NSCAT data

Long

Drinkwater

1999

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Abstract-Though designed to measure vector winds over the ocean, new imaging techniques facilitate the use of NASA scatterometer data (NSCAT) in cryosphere studies. NSCAT provides data of unprecedented coverage, resolution, and quality which, when coupled with the scatterometer image reconstruction with filtering (SIRF) algorithm, enables images of at resolutions approaching 8 km over stationary targets. Such images are useful in ice mapping and classification, and multidecadal studies are possible by comparison with Seasat Scatterometer (SASS) data. The utility of NSCAT data in polar ice studies is illustrated through a review of two cryosphere applications of NSCAT data: 1) sea-ice mapping and tracking and 2) ice-sheet change in Greenland and Antarctica. The wide swath and frequent revisit, coupled with incidence and azimuth angle diversity makes NSCAT data very effective in mapping the extent of sea-ice. In Greenland, snow and ice "facies" are clearly delineated on the basis of the seasonally dependent radar backscattering cross section, due to sensitivity of radar backscatter to diagenetic changes occurring at and beneath the surface. Comparison of NSCAT and SASS data enables study of change in Greenland between 1978 and 1996.

show abstract

Section: Sea Ice Motion Trackingmentioning

confidence: 99%

Cryosphere applications of NSCAT data

Long

Drinkwater

1999

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The two-dimensional Gaussian wavelet (often referred to as Mexican hat) has been used as an edge detector in SAR imagery for small-scale features [Canny, 1986]. In a marginal ice zone study by Liu et al [1994], the ice edge in each SAR image was delineated by using the two-dimensional wavelet transform, which has also been applied to satellite images (SAR, AVHRR, and ocean color) to separate various scale processes including relative phase/location information for coastal watch applications [Liu et al, 1997a]. The wavelet transforms of satellite images can be used for near-real-time "quick look" analyses of satellite data for feature detection, for data reduction using a binary image, and for image enhancement by edge linking.…”

Section: Introductionmentioning

confidence: 99%

Arctic sea ice drift from wavelet analysis of NSCAT and special sensor microwave imager data

Liu¹,

Zhao²,

Wu³

1999

J. Geophys. Res.

View full text Add to dashboard Cite

Abstract. Wavelet analysis of NASA scatterometer (NSCAT) backscatter and DMSP special sensor microwave imager (SSM/I) radiance data can be used to obtain daily sea ice drift information for the Arctic region. This technique provides improved spatial coverage over the existing array of Arctic Ocean buoys and better temporal resolution over techniques utilizing data from satellite synthetic aperture radars. Comparisons with ice motion derived from ocean buoys give good quantitative agreement. Both comparison results from NSCAT and SSM/I are compatible, and the results from NSCAT can definitely complement that from SSM/I when there are cloud or surface effects. Then, three sea ice drift daily results from NSCAT, SSM/I, and buoy data can be merged as a composite map by some data fusion techniques. The ice flow streamlines are highly correlated with surface air pressure contours. Examples of derived ice drift maps in December 1996 illustrate large-scale circulation reversals over a period of 4 days. These calibrated/validated results indicate that NSCAT, SSM/I merged daily ice motions are suitably accurate to identify and closely locate sea ice processes and to improve our current knowledge of sea ice drift and related processes through the data assimilation of ocean-ice numerical model. IntroductionIn 1996 the NASA scatterometer (NSCAT) rode into orbit on the Japanese satellite Advanced Earth Observing System (ADEOS) and gathered 8.5 months of valuable wind data.

show abstract

“…A two-dimensional wavelet transform is a highly efficient band-pass data filter, which can be used to separate various scales of processes [15,16]. For effective identification and tracking of the common features in a pair of chosen subscenes, a two-dimensional Mexican-hat wavelet transform is applied to the subscenes with several spatial scales corresponding to extracted features and filtering out noise in the data [17].…”

Section: Image Processing By Wavelet Analysismentioning

confidence: 99%

Deriving Ocean Surface Drift Using Multiple SAR Sensors

Liu

Hsu²

2009

Remote Sensing

View full text Add to dashboard Cite

Tracking and monitoring ocean features which have short coherent time periods from sequential satellite images requires that the images have both very high spatial resolutions and short temporal sampling intervals (i.e., repeated cycles). Satellite images from a single sensor in a polar-orbiting satellite usually cannot meet these requirements since high spatial resolution of the sensor may result in relatively long temporal sampling interval and vice versa, such as the case of Synthetic Aperture Radar (SAR). This paper presents a new multi-sensor approach to overcome the long temporal sampling interval associated with a single SAR sensor while taking advantage of high spatial resolution of SAR images for the application of ocean feature tracking. Currently, there are two SAR sensors on different satellites, the European Remote Sensing Satellite-2 (ERS-2) and the ENVIronment SATellite (ENVISAT), having acquisition time offset around 28 minutes with almost exactly the same path. That is, ERS-2 is following ENVISAT with a 28-minutes delay, which is a good time-scale for ocean mesoscale feature tracking. A pair of SAR images from ERS-2 and ENVISAT collected on April 27, 2005 has been chosen to track ocean surface features by using wavelet analysis. As demonstrated in the case studies, this technique is robust and capable to derive ocean surface drift near an oil slick and around a big eddy in the South China Sea (SCS).

show abstract

Wavelet analysis of satellite images for coastal watch

Cited by 85 publications

References 9 publications

Cryosphere applications of NSCAT data

Cryosphere applications of NSCAT data

Arctic sea ice drift from wavelet analysis of NSCAT and special sensor microwave imager data

Deriving Ocean Surface Drift Using Multiple SAR Sensors

Contact Info

Product

Resources

About