This PDF file contains the front matter associated with SPIE Proceedings Volume 10778, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

Seagrasses are distinct flowering plants which thrive underwater. They are part of a complex ecosystem that supports different forms of life. Recent studies found out that coastal wetlands – mangroves, saltmarshes, and seagrass, are far more proficient in sequestering and storing carbon than terrestrial ecosystems. Although seagrasses occupy only 0.2% of the area of the oceans, they sequester approximately 15% of total carbon storage in the ocean. Several remote sensing techniques are available to map and monitor seagrasses but most of them focus only on extent and area coverage. To estimate the carbon sequestration of seagrass beds, aside from extent, other parameters are needed such as leaf area index, percent cover, density, biomass etc., However, there are limits in mapping seagrass parameters using remote sensing. The reflectance measured by sensors is affected by other factors such as water absorption, turbidity, dissolved organic matter, depth and phytoplankton which affects the backscattering of energy. In this study, different remotely sensed datasets and field data were used to measure the parameters needed to estimate the carbon sequestration. Multispectral satellite images such as Sentinel-2 and PlanetScope were utilized to map the distribution and percent cover. High-resolution RGB images obtained by unmanned aerial vehicle (UAV) were also utilized to correlate field data gathered parameters. Field data such as species, percent cover, leaf area index, canopy height and above ground biomass were gathered in situ. Data extracted from different remote sensing technologies were put together to support the estimation of carbon sequestration of seagrass beds.

Water body extraction plays an important role in flood control and the utilization of water resources. With the launch of China’s first high-resolution (50m) geostationary optical GF-4 satellite at the end of December 2015, the wide-swath (400km) and high-frequency (up to minutes) imaging capabilities have been greatly improved, which provides new possibilities for rapid and accurate water body monitoring. To explore the potential of GF-4 satellite in water body monitoring, this paper proposes a water body extraction method based on the temporal variability of near infrared (NIR) spectral features. For a series of preprocessed and coregistered GF-4 images, one of them is chosen as the base image whose NIR band (B5) thresholding is firstly applied to eliminate most of the non-water regions. Then, for each pixel, the variance of B5 radiance values of all images is calculated to obtain a variogram, and pixels whose variogram values are larger than a certain threshold given by the OTSU algorithm are further eliminated. Finally, the final water body extraction result can be obtained after post-classification processing. To evaluate the efficacy of the proposed method, two groups of GF-4 datasets with complex water distribution are selected in the areas of the middle and lower reaches of Yangtze River in China. Experimental results demonstrate that thanks to the high-frequency and high-resolution characteristics of GF-4, the proposed method can extract more tiny waters and effectively remove built-up areas, and is superior to the extraction accuracy of water index way by about 4%.

Photosynthesis available radiation (PAR) that makes primary producers to compose carbon compounds is the energy source of carbon circulation at the ocean. In these days, global scale PAR is efficiently observed from satellite remotesensing with low cost and high resolution. Here, Geostationary Ocean Color Imager (GOCI) which is geostationary orbit sensor is used to estimate daily PAR at smaller scale area for decreasing influence of diurnal variation such as cloud. GOCI daily PAR is estimated using PAR model based on Plane-parallel theory and compared with in-situ data observed during year of 2015 at two stations that has turbid and clear ocean area, respectively. Each band image of GOCI L1B data and solar altitude data are input data for PAR model to estimate daily PAR. Correlation coefficient between GOCI daily PAR and in-situ daily PAR is 0.98 and root-mean-square error (RMSE) is 4.50 Ein/m² /day. To correct underestimated GOCI daily PAR, correction equation is developed from linear regression between GOCI daily PAR and in-situ daily PAR observed during clear sky condition days. RMSE of GOCI daily PAR which corrected with correction equation is decreased to 3.08 Ein/m² /day and seasonal bias between GOCI and in-situ daily PAR is decreased, too. Validation is carried out with in-situ daily PAR observed during year of 2016. Correlation coefficient is 0.98 and RMSE is 2.69 Ein/m² /day. Estimating GOCI daily PAR is expected to make accurate daily PAR by reducing meteorological element and regional error.

The Earth Polychromatic Imaging Camera (EPIC) onboard the Deep Space Climate Observatory (DSCOVR) in Lagrange-1 (L1) orbit provides observations of the Earth’s surface lit by the Sun at a cadence of 13 to 22 images/day and optical resolution of 16 km in 10 spectral bands from 317 to 780 nm. The EPIC data collected in the bands centered on 443, 551, and 680 nm are used to estimate daily mean photosynthetically available radiation (PAR) reaching the surface of the global, ice-free oceans. The solar irradiance reaching the surface is obtained by subtracting from the extraterrestrial irradiance (known), the irradiance reflected to space (estimated from the EPIC measurements), while taking into account atmospheric transmission (modeled). Clear and cloudy regions within a pixel do not need to be distinguished, i.e., the methodology is adapted to the relatively large EPIC pixels. A first daily mean EPIC PAR imagery is generated. Comparison with estimates from sensors in polar and geostationary orbits, namely MODIS and AHI, shows good agreement, with coefficients of determination of 0.79 and 0.92 and RMS differences of 8.2 and 5.7 E/m²/d, respectively, but overestimation by 1.08 E/m²/d (MODIS) and 3.44 E/m²/d (AHI). The advantages of using observations from L1 orbit are: 1) the daily cycle of cloudiness is well described (unlike from polar orbit) and 2) spatial resolution is not significantly degraded at high latitudes (unlike from geostationary orbit). The methodology can be easily extended to estimate ultraviolet (UV) surface irradiance using the spectral bands centered on 317, 325, 340, and 388 nm, all the more as ozone content, a key variable controlling atmospheric transmittance, is retrieved from the measurements.

Algorithms to retrieve ocean color from space, deterministic or statistical, often use a simplified water reflectance model, specified by a few parameters (e.g., chlorophyll concentration, backscattering and absorption coefficients at a given wavelength). The model, however, may not be representative of the worldwide ocean conditions, since many variables affecting reflectance are fixed at some average values. In this context, the semi-analytical model of Park and Ruddick (2005), PR05, used in the spectral matching POLYMER algorithm (Steinmetz et al., 2011), is examined in terms of its ability to represent properly water reflectance. The PR05 model depends on chlorophyll-a concentration, a parameter specifying the contribution of algal and non-algal particles to the backscattering coefficient, and a parameter allowing different absorption coefficients for dissolved organic matter. Model estimates at MODIS wavelengths, obtained for a representative set of Case 1 and Case 2 waters, are compared with Hydrolight calculations that include fluorescence and Raman scattering and AERONET-OC measurements. The accuracy of retrieving inherent optical properties (IOPs) using the reconstructed reflectance is also evaluated. The model parameters that give the best fit with the simulated data are determined. Agreement is generally good between the two- or three-parameter model results and Hydrolight/AERONETOC values, even in optically complex waters, with discrepancies much smaller than typical atmospheric correction errors. Significant differences exist in some cases, but having a more intricate model (i.e., using more parameters) might not guarantee convergence of the inversion scheme. The trade-off is between efficiency/robustness and accuracy. Significant errors are observed when using the model estimates to retrieve IOPs. Importantly, the model parameters that best fit the input data, in particular chlorophyll-a concentration, may not represent adequately actual values. The reconstructed water reflectance, not the retrieved model parameters, should be used in bio-optical algorithms.

Physical deterministic sea surface temperature (PDSST) retrieval scheme is built on radiative transfer forward model and a mathematically deterministic approach to the solution for inverse problem. This requires atmospheric profiles information from Numerical Weather Prediction (NWP), which offers the prospect to account for local retrieval conditions and yields a more uniform product with superior accuracy. One of the unprecedented capabilities of the PDSST scheme is that it can use aerosol profiles in addition to atmospheric profiles information for the forward modeling, and also allows for adjustment of the aerosol burden by including it as a retrieved element. Cloud detection is a vital part of SST retrieval processing. An innovative cloud and error masking (CEM) algorithm has been developed, combining the functional spectral differences and radiative transfer based cloud detection tests, especially the functional double difference tests are unique. These advancements have led to substantial improvements in information retrieval from expensive satellite measurement. This improvement refers to a dual benefit of increased data coverage (reduced false alarms) and detection of actual cloud contamination (improved detection rate). The PDSST retrieval suite, is combining the PDSST retrieval scheme and CEM, demonstrates the superiority of this approach with an overall ~3-4 times information gain when implemented on data from MODIS-Aqua and GOES Imager. For example, RMSE reduction from 0.52 K to 0.35 K and data coverage enhanced from ~9% to ~19% as compared to NASA operational MODIS-AQUA SST products.

JAXA polar-orbit satellite, Global Change Observation Mission for Climate (GCOM-C) which carries Secondgeneration Global Imager (SGLI) has been launched on 23 Dec. 2018. SGLI has 19 channels from near-UV (380 nm) to thermal infrared (12 μm) wavelengths with swath-width of 1150-km (for visible and near infrared, VNIR, channels) or 1400-km (for short-wave infrared, SWIR, and thermal infrared, TIR, channels). The SGLI 250-m spatial resolution of 11 VNIR channels in 380-867 nm, one SWIR channel of 1.6 μm, and two TIR channels of 11 μm and 12 μm can be advantage in monitoring fine structures of coastal areas through the retrieved products such as ocean color and seasurface temperature. After starting the test observation since 1 Jan. 2018, we are investigating calibration performance including on-board calibration, vicarious/cross calibration and image quality (random or stripe noise). The initial product evaluation indicates that the SGLI data has capability for the coastal observations with the 250-m spatial resolution. The SGLI standard products will be released to the public after the initial evaluation phase until Dec. 2018.

The Copernicus programme brings a wealth of ocean colour data at medium and high spatial resolution with a full, free and open data access policy, allowing for unprecedented monitoring capabilities of the open ocean and coastal and inland waters. The POLYMER atmospheric correction algorithm, with its genericity and robustness to most atmospheric and surface perturbations (aerosols, sun glint, thin clouds, adjacency effect), allows to maximize these observation capabilities, in particular for Sentinel-2 MSI and Sentinel-3 OLCI. The algorithm is fully consistent between these sensors, which gives access to a unique product in terms of potential applications. The evolution of the POLYMER algorithm will be presented, with examples of applications and validation results for Sentinel-2 and Sentinel-3.

Coral reef live percent cover (LPC) mapping has always been a challenging application for remote-sensing. The adoption of machine-learning algorithm in remote-sensing has opened-up the possibility of mapping coral reef at higher accuracy. This paper presents the application of machine-learning regression in the empirical modeling of coral reef LPC mapping. Stepwise regression, Support Vector Machine (SVM) regression, and Random Forest (RF) regression were used model the percentage of live coral cover in optically shallow water of Parang Island, Central Java, Indonesia using field photo-transect data to train the PlanetScope image. PlanetScope multispectral bands were transformed into water column corrected bands, Principle Component bands, and Cooccurrence texture analysis bands to be used as predictors in the regression process. The results indicate that the accuracy of machine learning algorithm to map coral reef LPC is relatively low due to the radiometric quality issue in the PlanetScope image (RMSE = 15.43%). We could not yet fairly justify the performance of machine learning algorithm until we applied the algorithms in other images.

The Indonesian Throughflow (ITF) is one of the key ocean current in considering global climate change because ITF transports substantial heat content from the Western Pacific tropical zone to the Eastern Indian Ocean tropical zone. However, the field observation system at the site is very few, and the detail of ITF is not clear. Meanwhile, Japanese geostationary weather satellite, "Himawari-8", which has been in full operation since July 2015, has the ability to observe the hemisphere including the ITF area from visible to infrared radiation at 10 minute intervals. In this research, techniques to estimate the flow distribution of ITF were discussed taking advantage of the features of such high frequency observation. Specifically, surface velocity was estimated by MCC (Maximum Cross Correlation) method using multiple SST (2 km spatial resolution, 1 hour average interval, Level 3) data of P-tree System provided by JAXA. As a result, the flow velocity of 0.5 m/s or more was estimated using the data of August 2017 in the Major Straits of the ITF such as the Lombok Strait. The estimate is consistent with the field observation value measured by ADCP. By using such high-frequency latest satellite data, the possibility of quantifying short-term coastal environment change was shown.

The Northwest Pacific region, which includes parts of northeast China, Japan, Korea and southeast Russia, is one of the most densely populated areas of the world. Eutrophication is an emerging environmental problem in this region, where a significant number of red tides and hypoxic conditions have been reported in coastal waters - possibly due to anthropogenic influences such as extensive chemical fertilizer use and sewage effluent. To assess this problem, NOWPAP CEARAC, the Special Monitoring and Coastal Environment Assessment Regional Activity Centre of the Action Plan for the Protection, Management and Development of the Marine and Coastal Environment of the Northwest Pacific Region of the United Nations Environmental Programme, has developed "Procedures for assessment of eutrophication status including the evaluation of land-based sources of nutrients for the NOWPAP region" (NOWPAP Common Procedures). The NOWPAP Common Procedures include the screening procedure to detect symptoms of eutrophication with selected parameters. One of the selected parameters is remotely sensed chlorophyll-a concentration (satellite Chl-a). To prepare a long-term consistent satellite Chl-a from 1998 to 2016, regression analysis was conducted by pixel to pixel using the daily composites of SeaWiFS and MODIS Remote Sensing Reflectance for overlapping period (July 2002 to December 2004). Two different empirical in-water algorithms, a NASA standard and a regionally developed one for turbid water, were applied to estimate Chl-a in the eastern and western parts of the Northwest Pacific region, respectively. The assessment of eutrophication was then conducted by the level and trend of satellite Chl-a.

Based on the processing of CZMIL data collected in Hawaii during a JALBTCX mission (2013) and in the Pacific for The Ocean Cleanup project (October, 2016), we demonstrate the possibility of reliably estimating the seawater column’s optical properties from lidar waveforms in deep clear (Jerlov class I and IB) waters. With minor improvements to the data processing method previously applied to Florida survey data (2003, 2006, 2012–2017), we estimate the diffuse attenuation coefficient at the wavelength of 532 nm, K_d (532), to be 0.045–0.060 m^-1 in both regions. The results are in good agreement with space satellite data for the days of the lidar surveys and with Jerlov’s Kd curves for water classes I and IB.

Seagrass beds provide habitat for invertebrate and fish species, many of which are economically important. In addition, they perform important physical functions such as trapping sediment particulates associated with dissipating wave energy, thus are helpful to maintain clear waters. We, here, generated the map of seagrass distribution using remotely sensed images to which atmospheric corrections and water column corrections had been applied. Then, the seagrass habitat distribution changes were calculated by seagrass habitat map. For this study, we selected Deukryang Bay located on the southern coast of the Korean peninsula. It is surrounded by small villages like Jinmok-ri and Ongam-ri. Zostera marina dominated at the bay, small amounts of Z. caulescens and Halophila nipponica are also distributed in this area. The results showed that image classifications to which the water column correction had been applied produced improved accuracies in all the classification algorithms we had employed. The object-based classification algorithm showed the highest accuracy, but it is effective method for the high spatial resolution remotely sensed images, consequently not suitable for monitoring changes of the long-term base. Thus, we applied the Mahalanobis distance method which had been known to suitable for medium spatial resolution images like Landsat. This study revealed that seagrass beds in the study area showed similar pattern of distribution during recent 20 years.

Shortly after strong rains or tropical storms, chlorophyll-a concentration (Chla), turbidity, and colored dissolved organic matter (CDOM) in the lagoon of New Caledonia may be tripled. This effect is visible as far as 50 km offshore. The Eastern Coast of New Caledonia is more impacted than the Western Coast because of its geomorphological structure. Higher frequency and intensity of the rains, and the dominance of highly erosive lateritic formations resulted from deep weathering of ultramafic rocks. At the deeper parts of the lagoon (> 20 m), increases of Chla, turbidity, and CDOM absorption can be detected by satellite imagery with their plumes extending off shore as a result of wind-driven circulation in the lagoon. Such increases agree with the oceanographic measurements of Chla, absorption, backscattering, turbidity and CDOM during extensive cruises. Satellite and in situ measurements allow tracing of particulates and dissolved matter in lagoon waters and up to the coral barrier reefs. The largest particles settle at a short distance from the coast influencing the sediments not far from the river outflows. Fine particles (< 10 μm), colloids and/or CDOM are exported to reefs where they can be beneficial through protection from high UV illumination (colloids) and feeding of the corals (CDOM). Phytoplankton and associated detritus, when not dominated by toxic filamentous cyanobacteria, can also feed coral reefs. Finally, the export of fine particles associated with colloids and/or CDOM can also contribute to the distribution of trace metals to the barrier reefs of the East Coast and offshore. The question whether this input of trace metals has an overall positive or negative effect on the coral reefs is still under debate.

Local changes in specular reflections of visible sunlight on the ocean surfaces can be captured effectively by satellite sensors operating in the visible range of the electromagnetic spectrum. This causes the sun-glint imagery to closely resemble the oceanic images obtained using Synthetic Aperture Radar (SAR) further allowing the identification of the various fine scale structures and patterns of the ocean. Moreover, at relevant spatial resolutions, cloud-free conditions as well as optimum relative positions of the sensor, sun and the wave front it is possible to image ocean waves, wave transformations and refraction patterns using Satellite Sun-glint imagery (SSGI). In the present study, Landsat OLI imagery captured along the coast of Brest, France is used to derive ocean wave characteristics such as wavelength, direction, amplitude and then mapped to better understand the process of wave transformation. The 2D fast Fourier transform technique has been used on Band 5 (NIR, 0.851 - 0.879μm) to derive the wavelength of swell waves in nearshore regions as well as to analyze the wavelength change. Furthermore, owing to the detector configuration of Landsat 8 OLI there is a small time lag between the channel acquisitions. This effectively helps to infer the space-time characteristics of the surface waves using the cross channel correlation between Band 5 and Band 6 subsequently enabling removal of the directional ambiguity associated with the wave spectra obtained from the analysis. The main purpose of this study is to demonstrate the importance of SSGI in deriving relevant coastal information which can be further utilized for bathymetry, surface current and wave motion determinations.

Remote sensing analysis is routinely used to map flood inundation during flooding events or retrospectively for planning and research activities. Quantification of the depth of floodwater is important for emergency response, relief operations, damage assessment etc. The Floodwater Depth Estimation Tool (FwDET) calculates water depth based on topographic analysis using standard GIS tools within a Python script. FwDET’s low input requirements (DEM and inundation polygon) and high computational efficiency lend it as a useful tool for emergency response and large-scale applications. Operational use of FwDET is described herein as part of emergency response activation of the Global Flood Partnership (GFP) during the 2017 USA Hurricane Season and May 2018 flooding in Sri Lanka. Use of FwDET during Hurricanes Harvey (Texas and Louisiana), Irma (Florida) and Maria (Puerto Rico) demonstrated its utility by producing large-scale water depth products at near-real-time at relatively high spatial resolution. Despite FwDET’s success, limitations of the tool stemmed from bureaucratic disallowance of non-governmental remote sensing products by U.S. federal emergency response agencies, misclassified remotely sensed floodwaters and challenges obtaining global high resolution DEMs specifically for the aforementioned Sri Lankian flooding. While global-scale DEM products at 30m resolution are freely available, these datasets are of integer precision and thus have limited vertical resolution. This limitation is significant primarily in flat (e.g. coastal) locations and flooded domains comprised of relatively small patches of water.

Remote-sensing of ocean colour has an advantage over any other biological data source for monitoring long-term global changes in phytoplankton biomass, due to its spatial and temporal sampling capabilities. Chlorophyll-a concentration (Chl-a) provide a proxy for phytoplankton biomass. The Operational Land Imager (OLI) is a multispectral radiometer hosted on the recently launched Landsat8 satellite, which is a potential tool for ocean color radiometry because it includes narrow band, high signal-to-noise ratios (SNRs) and the addition of a band centered at 443 nm, has competitive advantage in phytoplankton pigment Chl-a estimation compare with previous Landsat instruments. The aim of this work was to evaluate the performance of the standard NASA algorithm OC3 type for Landsat-8 OLI in determining Chl-a concentrations in both turbid estuary and clear open sea waters of the north South China Sea, in which the empirical coefficients were tuned by using field data and used 443-, 561-, and 655nm bands instead of 443-, 482 and 561nm bands. The standard OC3-based algorithm for OLI performed well in the Southeast continental shelf of Hainan Island (HNI). While empirical algorithm should be developed in the Pearl River estuary (PRE), and the comparisons between estimated and in situ measured Chl-a produced R² reaching 0.88 and APD <30%. Furthermore, we assessed Chl-a products by conducting cross-validation with concurrent MODIS-Aqua and NPP VIIRS data, which demonstrate good consistency and minor deviation in HNI waters, while demonstrate good consistency but large deviation in the PRE waters. Our findings demonstrate the potential of high resolution OLI Chl-a products to study short-lasting events and capture fine-scale features in the marine environment in different cases waters. The OLI Chl-a products using standard OC3-based algorithm performed well in the case I waters, while regional algorithm should be developed basing on large field data in the estuary waters.

Simulation of the sea clutter signal by using digital signal processing (DSP) platform for radar performance testing is an prospective and widely used technique in radar engineering application. Based on the digital signal processor TMS320C6748 as the hardware platform and the autoregressive model, a method of generating non-stationary K distributed sea clutter is proposed. In DSP realization of the non-stationary sea clutter sequence simulation, its amplitude usually satisfies the K distribution, and the amplitude envelope and Doppler spectral center frequency will change with time. The simulation results show that the probability distribution and time-frequency characteristics of the sea clutter are in good agreement with the theoretical values.

It is important to understand the flow of marine debris for environmental research purposes, since marine debris causes extensive damage to coastal environments. Due to its small size, most marine debris in the ocean cannot be confirmed directly, even when a high-spatial-resolution satellite image is used. Thus, to extract candidate pixels containing possible marine debris, pixels with spectra that differ from those of the surrounding ocean are identified. As a first step towards identifying and monitoring marine debris, a method using spectral angle mapper (SAM) algorithm in n-dimensional space corresponding to the satellite spectral bands was previously proposed. In this paper, a method to discriminate marine debris from white-crested waves is proposed using the distance from the origin in an n-dimensional scatter diagram. Moreover, it is also discussed that the relationship between the distance from the coast and the amount of marine debris depends on the locations of the sea currents and neighbouring rivers.

MODIS chlorophyll-a concentration (chla), sea surface temperature (SST), and photosynthetically active radiation (PAR) were used to perform a geographically weighted regression (GWR) analysis within a 150-km buffer of the Brazilian coast. The correlation was between chla as the regressed variable and SST or PAR as the predictors. Both a GWR and a Bayesian GWR (BGWR) were used for evaluating the variables. Colored matrices were plotted for displaying beta values, significance, residuals, and t-statistics. Coefficients of determination (R²) were computed for all months. Also, the ratio of the GWR beta estimates and the 95% confidence interval BGWR estimates was computed. Results showed overall better R² for SST than for PAR regression but also better beta estimates for PAR than for SST in relation to BGWR beta significance range. Northern regions of the Brazilian coast exhibited lower statistical significance. July had lowest GWR beta values and best significance, January highest beta values and worst significance, and April and October highly variable results.

Satellite ocean-color project offices routinely generate Level 2 and Level 3 daily Photo-synthetically Available Radiation (PAR) products. Accuracy is currently evaluated against in-situ measurements from buoys and fixed platforms at a few locations, but specifying algorithm (and other) uncertainties on a pixel-by-pixel basis is needed to assess product quality. Expressing uncertainties requires modeling the measurement, identifying all possible error sources (e.g., noise in the input variables, imperfect/incomplete mathematical model), and determining the combined uncertainty. In the present study, algorithm uncertainties associated with PAR products are considered, i.e., those due to model approximations and parameter errors (e.g., decoupling effects of clouds and clear atmosphere, neglecting diurnal variability of clouds, using aerosol climatology) assuming that the input variables (TOA reflectance at wavelengths in the PAR spectral range) are known perfectly. A procedure is provided to estimate and provide, for each pixel of a product, this uncertainty component of the total uncertainty budget, which is expected to dominate. The bias and standard deviation of the daily PAR estimates are calculated as a function of clear sky PAR and cloud factor (i.e., the effect of clouds on daily PAR). The uncertainty characterization is accomplished using an extended simulation dataset covering the 2003–2012 time period using hourly MERRA-2 input data. The large number of data points allows one to sample well atmospheric variability and in particular many variations of daytime cloudiness, for all latitudes. Selected maps of global daily and monthly PAR and associated uncertainties (bias, standard deviation), obtained from MERIS data, are analyzed. Comparisons with match-up data at the COVE calibration/evaluation site reveal that experimental uncertainties are similar to the theoretical uncertainties obtained from simulated data.