Xiao-Hai Yan

A Neural Network Model for Estimating Sea Surface Chlorophyll and Sediments from Thematic Mapper Imagery

Abstract Two important parameters used for monitoring coastal water quality are the concentrations of chlorophyll and suspended sediment in surface waters. Ocean color remote sensing provides a convenient method of determining these concentrations from upwelling radiances. In the open ocean, it is not difficult to derive empirical algorithms relating the received radiances to surface concentrations of chlorophyll. In turbid coastal waters, however, this is much more difficult due to the presence of high concentrations of suspended sediments and dissolved organic material, which overwhelm the spectral signal of chlorophyll. Neural networks have been proven successful in modeling a variety of geophysical transfer functions. Here, a neural network is employed to model the transfer function between the chlorophyll and sediment concentrations and the satellite-received radiances. It was found that a neural network with two hidden nodes, using the three visible Landsat Thematic Mapper bands as inputs, was able to model the transfer function to a much higher accuracy than multiple regression analysis. The RMS errors for the neural network were 25%.

Temperature and size variabilities of the Western Pacific Warm Pool

Variabilities in sea-surface temperature and size of the Western Pacific Warm Pool were tracked with 10 years of satellite multichannel sea-surface temperature observations from 1982 to 1991. The results show that both annual mean sea-surface temperature and the size of the warm pool increased from 1983 to 1987 and fluctuated after 1987. Possible causes of these variations include solar irradiance variabilities, EI Ni�o—Southern Oscillation events, volcanic activities, and global warming.

Theoretical expression for an ocean internal soliton synthetic aperture radar image and determination of the soliton characteristic half width

This paper deals with the development of techniques for satellite synthetic aperture radar (SAR) ocean image interpretation. We derived a theoretical model of a radar image for a Korteweg-de Vries type ocean internal soliton and validated the model using ocean internal wave signals taken from ERS-1 SAR and RADARSAT SAR images. The results indicate that the model perfectly simulates ocean internal soliton signatures with double-sign variations of radar backscatter. On the basis of the model, we developed the curve fitting method and the peak-to-peak method for determining the internal soliton characteristic half widths, which then were used to calculate the internal soliton amplitudes. The test results indicate that ocean internal soliton amplitudes derived by the two methods agree with in situ data acquired on the Portuguese Continental Shelf and in the South China Sea with reasonable accuracy. The role that wind fields play in ocean radar remote sensing was also analyzed. Finally, the modulation ratio of ocean internal waves on radar images was quantitatively estimated.

The Probability Density Function of Ocean Surface Slopes and Its Effects on Radar Backscatter

Abstract Based on Longuet-Higgins’s theory of the probability distribution of wave amplitude and wave period and on some observations, a new probability density function (PDF) of ocean surface slopes is derived. It is where ζx and ζy are the slope components in upwind and crosswind directions, respectively; σ2u and σ2c are the corresponding mean-square slopes. The peakedness of slopes is generated by nonlinear wave–wave interactions in the range of gravity waves. The skewness of slopes is generated by nonlinear coupling between the short waves and the underlying long waves. The peakedness coefficient n of the detectable surface slopes is determined by both the spectral width of the gravity waves, and the ratio between the gravity wave mean-square slope and the detectable short wave mean-square slope. When n equals 10, the proposed PDF fits the Gram Charlier distribution, given by Cox and Munk, very well in the range of small slopes. When n → ∞, it is very close to the Gaussian distribution. Radar backscat...

Statistical and dynamical analysis of internal waves on the continental shelf of the Middle Atlantic Bight from space shuttle photographs

By interpreting two space shuttle photographs taken with a Linhof camera on June 8, 1991, a total of 34 internal soliton packets on the continental shelf of the Middle Atlantic Bight are recognized. The internal soliton field has a three-level structure: packet groups with average wavelength of 17.5 km, packets with average wavelength of 7.9 km, and solitons with average wavelength of 0.6 km. Using the finite-depth theory we derive that the maximum amplitude of solitons is 5.6 m, the phase speed is 0.42 m/s, and the period is 23.8 min. The frequency distribution of solitons is triple-peaked at 1.9 × 10−4 Hz, 3.0 × 10−4, and 6.9 × 10−4 Hz. Substituting statistical results of number of solitons in a packet into the fission law, we find that the upper and the lower edges of the shelf break are the primary and the secondary generation sources of internal solitons, respectively. This reveals that the sharp change in the bottom topography is a key condition for the soliton fission or disintegration. Calculations show that the group period of solitons is 12.5 hours coinciding with that of local semidiurnal tides. This fact confirms that the tides are a dominant generation force for internal solitons on the continental shelf.

Coastal lee waves on ERS-1 SAR images

Alternative dark-bright patterns on ERS-1 synthetic aperture radar (SAR) images of the west side of the Taiwan Strait taken on December 8, 1994, were recognized to be the sea surface signature of a coastal lee wave. Such waves are called coastal lee waves because they occur along the lee side of the coast. The coastal lee waves appeared in the form of a wave packet distributed within an offshore band 20–40 km wide. The first packet, which occurred in the northern portion of the observed area, contained six waves with variable wavelengths (defined as the spatial separation between two waves) from 1.7 to 2.7 km. The second packet, in the middle, contained 10 waves with a relatively uniform wavelength of 4.2 km. The third packet, in the southern portion, contained 17 waves with an average wavelength of 2.0 km. The crest lengths were from 20 to 80 km. Local meteorologic parameters observed simultaneously at Fuzhou, China, close to the imaged area, showed an offshore wind of 1.5–3.5 m/s and a land surface air temperature of 19°C, which was 4°C lower than the sea surface temperature (SST). Thus the lower atmospheric boundary conditions at imaging time were very favorable both for generating the land breeze circulation and small wind waves on the sea surface, which are in the Bragg-scattering wavelength band of the C band ERS-1 SAR. A physical model of a three-layer atmosphere was developed in order to explain how the land breeze circulation can generate the coastal lee waves. The results showed that the vertical velocity disturbance caused by the wind convergence at the land breeze frontal zone is of vital importance for the generation of coastal lee waves, and the model gave very good estimates of the processes observed. The SAR imaging mechanisms of the waves were analyzed in detail. The differences between coastal lee waves and ocean internal waves, which appear as similar alternative dark-bright patterns on SAR images, were also discussed.

Satellite observations of upper-layer variabilities in the western Pacific warm pool

The variabilities of the upper layer of the western Pacific warm pool (WPWP) were observed using satellite infrared data from 1982 to 1991 and altimeter data from November 1986 to September 1989. The warm pool was defined as the area where the sea surface temperatures are above 28 C. The eastern boundary oscillation, the centroid movement, and the upper-layer volume variation of the WPWP were intensively studied. Spectral analysis revealed that the eastern boundary oscillation of the WPWP was related to the El Nino event and the annual cycle. The centroid of the WPWP traced an ellipselike trajectory during a year and moved counterclockwise in most years. However, in 1982 and 1986, the years of the onset of El Nino events, the movements were clockwise. The upper-layer volume of the WPWP was divided latitudinally into three sections. The annual cycles in the northern (from 3 deg to 30 deg N) and southern (from 3 deg to 30 deg S) sections were dominant. No annual cycle was found in the equatorial section (from 3 deg S to 3 deg N), but the volume of warm water in the equatorial Pacific increased during the 1986/87 El Nino event. The equatorial section was further divided into the eastern and western sectors along 165 deg W. During the 1986/87 El Nino event, the volume of warm water increased in the eastern sector, but the variation was smaller in the western sector than that in the eastern sector. During the 1988 La Nina event, the warm water volumes decreased in both sectors.

The global warming hiatus: Slowdown or redistribution?

Abstract Global mean surface temperatures (GMST) exhibited a smaller rate of warming during 1998–2013, compared to the warming in the latter half of the 20th Century. Although, not a “true” hiatus in the strict definition of the word, this has been termed the “global warming hiatus” by IPCC (2013). There have been other periods that have also been defined as the “hiatus” depending on the analysis. There are a number of uncertainties and knowledge gaps regarding the “hiatus.” This report reviews these issues and also posits insights from a collective set of diverse information that helps us understand what we do and do not know. One salient insight is that the GMST phenomenon is a surface characteristic that does not represent a slowdown in warming of the climate system but rather is an energy redistribution within the oceans. Improved understanding of the ocean distribution and redistribution of heat will help better monitor Earths energy budget and its consequences. A review of recent scientific publications on the “hiatus” shows the difficulty and complexities in pinpointing the oceanic sink of the “missing heat” from the atmosphere and the upper layer of the oceans, which defines the “hiatus.” Advances in “hiatus” research and outlooks (recommendations) are given in this report.

Complex Singular Value Decomposition Analysis of Equatorial Waves in the Pacific Observed by TOPEX/Poseidon Altimeter

Abstract The mean of the sea level deviation data derived from the TOPEX/Poseidon altimeter in the equatorial Pacific, between 10°S and 10°N, and between 120°E and 78°W, from cycles 2 to 136 (3 October 1992–2 June 1996), are extracted using a maximum–minimum average method. Then, two-dimensional (2D) sea level deviation time series are developed to visualize the dynamics of equatorial waves. The complex singular value decomposition (CSVD) method is applied to decompose these 2D time series into empirical orthogonal modes. Using this method, zonal and meridional structures, propagation directions, periods, and propagation speeds of these empirical modes are obtained. The first empirical mode is propagating westward, and its structure is asymmetric to the equator. It has an average phase speed c = −0.6 m s−1 within 4°–6°N and c = −0.4 m s−1 within 6°–8°S, respectively, and a period of 15 months, which is associated with an interannual Rossby wave. The second empirical mode is propagating eastward along the ...

Interpretation of scatterometer ocean surface wind vector EOFs over the Northwestern Pacific

Abstract Satellite scatterometer winds over the northwestern Pacific were analyzed with the vector empirical orthogonal function (VEOF) method. The Hilbert–Huang transform (HHT), a newly developed non-linear and non-stationary time series data processing method, was also employed in the analysis. A combination of European Remote Sensing Satellite (ERS) −1/2 scatterometer, NASA Scatterometer (NSCAT) and NASAs Quick Scatterometer (QuikSCAT) winds covering the period from January 1992 to April 2000 and the area of 0–50°N, 100–148°E constitutes the baseline for this study. The results indicate that annual cycles dominate the two leading VEOF modes. The first VEOF shows the East Asian monsoon features and the second represents a spring–autumn oscillation. We removed the annual signal from the data set and calculated the interannual VEOFs. The first interannual VEOF represents the interannual variability existing in the spring–autumn oscillation. The temporal mode is correlated with the Southern Oscillation Index (SOI), but has a half-year lag with respect to the SOI. The spatial mode of the first interannual VEOF reflects the response of the tropical and extratropical winds to ENSO events. The second interannual VEOF is another ENSO related mode, and the temporal VEOF mode is correlated with the SOI with a correlation coefficient of 0.78, revealing the wind variability over mid-latitudes, which is associated with ENSO events. Further analysis indicated that the wind variability over the coast of East Asia represents anomalies of a Hadley cell. The quasi-biennial oscillation (QBO) was found in the temporal mode, indicating and verifying that the QBO in the wind fields is related to ENSO events. The third VEOF shows the interannaul variability in the winter–summer mode and displays the interannual variability of the East Asian monsoon. The three leading interannual VEOFs are statistically meaningful as confirmed by a significance test.