This PDF file contains the front matter associated with SPIE Proceedings Volume 11014, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

Abstract: Aside from the daunting physics based challenges needed to maintain Moore's Law, it's becoming increasingly difficult to justify the huge foundry investments required. This probable constraint to future electronics provides sensor developers with a unique opportunity to create more "elegant" sensing design options. Instead of continually increasing the amount of data collected in an open-loop manner, we should consider providing "intelligent" feedback to and within sensors in a more "closed-loop" manner. In other words, allow the sensor to dynamically learn how best to allocate its sampling, processing, and communication resources based on both external objectives and internal conditions. Imbedding deep learning machines into sensors designs could help mitigate Moore's Law's possible demise.

Oceanic turbulence may be detrimental for optical signals in short-link communication and sensing systems, and, in particular, LIDARs. To characterize light propagation for LIDARs we first establish 4 x 4 matrix framework applicable for general ABCD optical systems, a wide variety of optical waves, including partially coherent and partially polarized, and a variety of targets. We then adopt this approach for situations in which a target embedded in the turbulent oceanic layer is to be sensed by a bi-static LIDAR. Under some conditions the monostatic transmission links may also exhibit the Enhanced BackScatter (EBS) effect, due to phase conjugation of incident and return waves. We will brie y discuss the appearance of the EBS in the oceanic turbulence. Our treatment of the oceanic fluctuations is based on the well-known temperature-salinity based power spectrum of the refractive-index and the Huygens-Fresnel integral propagation method.

Demand for underwater communication in the 21st century has been growing. Underwater communication technologies are needed in many different environments; ranging from the underwater environments of the seas to small ponds. Underwater communication technologies used in these environments are used in a multitude of ways, such as, monitoring evolutionary changes, surveillance, underwater sensor networks, and military communication between ships and underwater-unmanned vehicles. Traditionally underwater communication has been accomplished using acoustic techniques that have data rates that can range from the kb/s and into the Mb/s. Interest in the use of optical communications has increased with the growing need for higher data rates, as well as, a secure way of transmitting it. Optical communications are capable of providing data speeds in the Gb/s due to a larger available bandwidth. Optical communications are also inherently more secure due to the directional nature of the transmission. That said, there are a number of challenges to optical communications, which include scattering, absorption, extinction, and optical turbulence in underwater environments. This paper presents the performance of an inexpensive LED communication system operating in various emulated underwater environments. Specifically, we propagated ASCII data through various underwater environments, and analyzed system performance under the effects of salinity level, temperature turbulence effects, mechanically generated turbulence, as well as the misalignment between the transmitter and receiver. Relevant performance metrics studied include bit error rates, power received, and channel data rates.

Spatial optical coherence filtering is investigated as a means of reducing the amount of scattered light collected by an underwater laser system in turbid water. This approach exploits differences in coherence between unscattered and scattered laser light as a means of discrimination against scattered light prior to opto-electronic detection. An all optical filter is designed and tested that uses an axicon and a mask to pass the coherent, unscattered light while blocking the incoherent, scattered light. Experiments are performed in a laboratory water test tank to measure the effectiveness of the filter in reducing scattered light collection. The results obtained using the axicon filter are compared to those obtained using no filtering and using a conventional spatial filter. The axicon filter is shown to reduce the contribution of scattered light relative to either other test case.

Coral reef ecosystems is essential in healthy ocean and marine fishery. In the past decades, substantial of images and videos haven been collected from these cruises. These images are analyzed to quantify coral abundance in certain specific areas. However, the current manual analysis are time-consuming and labor intensive. In this paper, we proposes a fast automated tool for coral identification only based on sparse annotated labels by using deep learning method. There are two challenges to identify coral from such sparse labels and large images: one is to obtain denser labeled training data and the other is to improve the speed of testing on large images. In order to solves these problems, we propose a label augmentation algorithm to generate more labels and coarse-to-fine approach to find the location of corals quickly. Our methods were validated using the coral image dataset collected in Pulley Ridge region in the Gulf of Mexico, which substantial speed up the process of quantifying the corals while preserving accuracy.

Boreal winter meteorological fronts manifest across the northern Gulf of Mexico as rapid 10-15° C drops in air temperature and accelerating northerly winds. The physical coastal ocean response across the Louisiana-Texas (LATEX) continental shelf system involves a complex interplay between coastal buoyancy, wind forcing, and intense thermal energy fluxes out of the ocean. Herein we combine numerical simulations, in situ optical surveys, and coincident satellite images derived from the Ocean and Land Colour Imager (OLCI) and other sensors to further unravel the mechanistic functioning and optical signatures of these complex events. The conspicuous optical gradients evident in color satellite images coincident with cold air outbreak (CAO) events appear to result from surface ventilation of sediment-laden bottom waters and wind/buoyancy-driven surface currents. The hyperspectral gradients associated with water mass types (sediment resuspension in marine waters versus freshwater effluent plumes) give rise to true color gradients that may be tracked with low spectral resolution color sensors at very high temporal resolution.

Herein we present an initial approach for assessing water color, specifically chromaticity, and determining if an accurate correlation can be made within chromaticity space between the water color and a hyperspectral synthetic data set. The water color assessed in this paper consist of remote sensing reflectance (R_rs) distributions from the Suomi-National Polarorbiting Partnership (SNPP) Visible Infrared Imaging Radiometer Suite (VIIRS), and the hyperspectral synthetic data set consists of R_rs distributions of natural marine waters. Where strong correlations exist, the hyperspectral R_rs reference data can be blended into the SNPP VIIRS R_rs data, thus creating a hyperspectral SNPP VIIRS spectra. Where applicable, the newly constructed VIIRS hyperspectral signature is compared to in situ data taken during a 2018 National Oceanic and Atmospheric Administration (NOAA) Calibration/Validation cruise. Given the proliferation of small, low-cost airborne platforms equipped with color imaging cameras, there exists tremendous potential to use and hyperspectrally enhance these data streams for ocean monitoring and scientific research. However, techniques for extracting traditional ocean radiant spectra from RGB data fields are new to oceanographic disciplines.

In September, 2018, Hurricane Florence made landfall in North Carolina as a Category 1 hurricane and inundated the eastern United States with significant rainfall. Precipitation from this slow moving storm event caused massive flooding. Outflow from this flooding carried suspended solids including sediments and other particulates as the rainwater worked its way through river and watershed systems toward the Atlantic Ocean. The Advanced Baseline Imager (ABI) on the NOAA Geostationary Operational Environmental Satellite - 16 (GOES-16) monitors the eastern United States. ABI data from GOES-16 is available every 5 minutes and provides a platform for studying the increased volume of river flow into the Atlantic Ocean. Data from the GOES-16 ABI covering the Atlantic waters off the eastern United States were downloaded after the Hurricane Florence event. Methodologies for atmospheric correction were used to generate water leaving radiance values from the GOES-16 ABI data sets. Using the multiple looks per day, the plumes of suspended solids were delineated and studied.

The SAtellite VAlidation Navy Tool (SAVANT) was developed by the Navy to help facilitate the assessment of the stability and accuracy of ocean color satellites using ground truth (insitu) platform and buoy stations positioned around the globe and support methods for match-up protocols. This automated, continuous monitoring system for satellite ocean color sensors employs a website interface to extract and graph coincident satellite and insitu data in near-real-time. Available satellite sensors include MODerate resolution Imaging Spectrometer (MODIS) onboard the Aqua satellite, Visible Infrared Imaging Radiometer Suite (VIIRS) onboard Suomi National Polar-orbitting Partnership (SNPP) & Joint Polar Satellite Sensor (JPSS), Ocean and Land Colour Instrument (OLCI) onboard the Sentinel 3A and Geostationary Ocean Color Imager (GOCI) onboard the Communication, Ocean and Meteorological Satellite (COMS). SAVANT houses an extensive match-up data set covering nineteen plus years (2000- 2019) of coincident global satellite and ground truth spectral Normalized Water Leaving Radiance (nLw) data allowing users to evaluate the accuracy of ocean color sensors spectral water leaving radiance at specific ground truth sites that provide continuous data. The tool permits changing different match-up constraints and evaluating the effects on the match-up uncertainty. Results include: a) the effects of spatial selection (using single satellite pixel versus 3x3 and 5x5 boxes, all centered around the insitu location), b) time difference between satellite overpass and ground truth observations, c) and satellite and solar zenith angles. Match-up uncertainty analyses was performed on VIIRS SNPP at the AErosol RObotic NETwork Ocean Color (AeroNET-OC) Wave Current surge Information System (WavCIS) site, maintained by NRL and the Louisiana State University (LSU) in the North Central Gulf of Mexico onboard the Chevron platform CSI-06. The VIIRS SNPP and AeroNET-OC assessment determined optimal satellite ocean color cal/val match-up protocols that reduced uncertainty in the derived satellite products.

We examine the potential for ocean color (OC) retrievals using a neural network (NN) technique recently developed by us to make up for the lack of a 678 nm fluorescence band on VIIRS, previously available on MODIS and important for Karenia brevis harmful algal bloom (KB HABs) retrievals. NN uses VIIRS Remote Sensing Reflectance (Rrs) at 486, 551 and 671 nm to retrieve phytoplankton absorption at 443nm, from which both chlorophyll [Chla] concentrations and KB HABs can be inferred. NN retrievals are compared with retrievals obtained using different ocean color algorithms for both complex and open ocean waters. Retrievals were compared using both satellite match-ups and in-situ radiometric Rrs measurements against sample measurements of different field campaigns in the WFS and Atlantic coasts, 2014-18. VIIRS KB HABs retrievals in the WFS, using NN and other algorithms, are compared against retrievals obtained for different algorithms using satellite observations and in-situ radiometric Rrs measurements against sample measurements, 2017-18, for both the WFS and Atlantic coasts. Retrieval statistics showed (i) the important impact of temporal inter-intra pixel variations and sample depth considerations in complex bloom waters. These limit satellite retrieval accuracies and utility; and (ii) particularly for high chlorophyll bloom waters, better retrieval accuracies were obtained with NN followed by OCI/OCx algorithms. Likely rationales: the longer Rrs wavelengths used with NN are less vulnerable (i) to atmospheric correction inadequacies than the deeper blue wavelengths used with other algorithms, and (ii) to spectral interference by CDOM in more complex waters.

Detecting drifting icebergs is an important task to avoid threats to navigation and offshore activities. Government and companies use aerial reconnaissance and shore-based observation platforms to detect these icebergs. However, in some areas with harsh weather conditions only satellite imagery can be used to monitor this risk. In this work, we propose the use of deep Convolutional Neural Networks to detect and classify these small remotely sensed targets as ships or icebergs. In this work, we use satellite radar imagery composed of two bands. The image patches have a resolution below 6K pixels and are noisy. To solve this challenge, we developed a deep convolutional network architecture and optimized its hyperparameters for this classification. The obtained results show that the proposed deep convolutional network achieves a very interesting accuracy for the classification of icebergs vs. ships with radar satellite images.

Results are presented for measurements of the total and polarized radiances from the ocean surface by a state-of-the-art snapshot hyperspectral imager, which simultaneously acquires spectra with 4nm spectral resolution in the wavelength range of 450-950nm within a 40° field-of-view. The imager does not require any along track movement and allows the continuous collection of hyperspectral imagery from stationary structures or slow moving platforms such as ships or helicopters. In addition, a computer controlled filter wheel is installed in front of the imager allowing for division-oftime Stokes vector images from the ocean surface. Experiments are aimed at the application of the multi-angular polarimetric measurements for the retrieval of water parameters in addition to the ones retrieved from traditional unpolarized methods. Several sets of measurements used in the analysis were acquired from ocean platform in the NYC area, Duck, NC and from shipborne observations in the Gulf of Mexico and along the Florida coast. Measurements made by the imager are compared with simulations using a vector radiative transfer code showing good agreement. Analysis of pixel-to-pixel variability of the total and polarized above water radiance for the viewing angles of 20°-60° in different wind conditions enable the estimation of uncertainties in measurements of these radiances in un-polarized and polarized modes for the whole spectral range, thus setting requirements for the quality of polarized measurements. Impacts of aerosols on spectral variations of both the radiance and the polarized terms of the Stokes vector are studied.

Characterizing turbulent instabilities in transitional and developed turbulent boundary layers is of critical importance to the development of methods to suppress such instabilities, which has important implications for research into drag reduction and energy-efficient propulsion methods. This includes the cancellation of Tollmien-Schlichting (TS) waves in the transitional and the counteracting of large-scale coherent structures in the developed turbulent boundary layer. To efficiently respond to instabilities near a boundary, the instabilities have to be sensed and described with a high level of accuracy. We visualize and measure velocity fluctuations in a laminar to turbulent flow tank to study the development of TS waves for active cancellation and identify boundary layers streaks for the development of stabilization methods. The flow structure in the tank is described through dye experiments, Particle Image Velocimetry (PIV), novel high-frequency fiber-optics flow sensors, and Acoustic Doppler Velocimeters (ADV), with the goal of identifying TS waves in transitional turbulence, as well as boundary layer streaks in developed turbulence. The impact of passive elastic boundary materials on the flow is investigated to aid the development of active actuated membranes aimed at reducing boundary layer turbulence and drag. In addition to flow measurements from various sensors, we employ computational fluid dynamics (CFD) to emulate the laboratory setting and complement the measurements. The CFD representation of the laboratory tank is implemented as high-resolution large-eddy simulation with elastic boundary conditions simulating the compliant boundary. The combined setup is a critical tool in the ongoing development of methods for active boundary later control.

Hurricanes are tropical weather systems that play a significant role in the exchange of energy (heat) between the atmosphere and ocean. Sea surface temperature (SST) variability and mixed layer depth (MLD) modulates the intensity of tropical cyclones. TRMM Microwave Imager (TMI) data corresponding to five hurricanes that traversed across the Gulf of Mexico (GOM) and Caribbean Sea were analyzed for evaluating the response of upper-ocean parameters to passing hurricanes. Data on SST, rainfall, wind speed, and MLD (from HYCOM) were compared from pre- and post-storm conditions. Daily data from the satellites were extracted from TMI data archives (http://www.remss.com/missions/tmi/) and further averaged during the time of hurricanes traversing across the GOM. Pre-storm conditions were generated by averaging data from the immediate days before hurricanes entering the GOM. Fast-moving Hurricane Charley cooled the West Florida Shelf by merely ~ 0.5°C, while Hurricane Ivan came through the following month as a much larger and slow-moving storm and a cooling impact up to 2°C. SST cooling from Hurricane Ivan extended over a wider region covering most of eastern GOM. The following year, Hurricanes Dennis, Katrina, Rita, and Wilma caused substantial ocean surface cooling with an observed maximum of 3.8°C. Precipitation from the hurricanes also exceeded 2.5-3.0 mm/hr, particularly along the right side of the hurricane path. Wind speed distribution showed significant spatial variability with maximum winds observed along the eastern side of the tracks. MLD data extracted from HYCOM model archives also responded dynamically to the passing hurricanes with the cooling extended to more than 100 m water depth after a major hurricane.

We investigate how oceanic turbulence impacts symbol-error-probability of free-space optical communication systems operating underwater and based on phase-shift keying heterodyne coherent receivers. We make use of the oceanic turbulence power spectrum and we estimate symbol error probabilities for M-ary phase shift keying as a function of the normalized aperture of the receiver: D/r₀.

In multinational defense operations, either EU or NATO driven, the exchange of surveillance and reconnaissance data and information is an essential aspect to be able to act promptly. Coordinated processes and agreements are the basis, distribution architectures, services, interfaces and formats the prerequisite. In the NATO context, the Joint ISR (Intelligence, Surveillance and Reconnaissance) process supports the execution of surveillance and reconnaissance tasks. The Coalition Shared Data (CSD) concept and the associated specifications, interfaces and information models defined in STANAGs (Standardization Agreements), as well as the NATO ISR Interoperability Architecture (NIIA), facilitate the exchange of information based on the described processes. The EU uses CISE (Common Information Sharing Environment) and MARSUR (Maritime Surveillance), which are based on NATO-like principles. Within this frame, the OCEAN2020 (Open Cooperation for European mAritime awareNess) project, funded by the European Union's Preparatory Action on Defense Research and implemented by the European Defense Agency, sees 42 partners from 15 EU countries working to network future maritime surveillance and interdiction missions at sea integrating drones and unmanned submarines into fleet operations. Here data and information will be integrated in a comprehensive (maritime) picture of developing situations for military commanders on different unit levels. Maritime Operation Centers (MOC) on a national and EU level can be connected with operational commands/units to exchange information. With its remote-acting units equipped with only temporary and often narrow-band network connections, the Navy places particular demands on architecture (s). This paper focuses on the challenge to define flexible architectures for maritime operations.

Sea Surface Temperature (SST) products at NOAA are produced from multiple polar-orbiting and geostationary sensors using the Advanced Clear-Sky Processor for Ocean (ACSPO) enterprise system. Data of several high-resolution (~1 km) sensors onboard US and EUMETSAT polar-orbiting platforms are processed, including two VIIRSs (onboard NPP and N20), three AVHRR FRAC (onboard Metop-A, -B, and -C), and two MODISs (onboard Terra and Aqua). L1b data of each platform/sensor are processed independently, and two SST products generated: swath (L2P) and 0.02° mapped equal-grid L3U. All L2P and L3U are consistently reported in 10-min granules in a Group for High-Resolution SST (GHRSST) Data Specification Version 2 (GDS2) format, and assimilated in several gridded gap-free L4 analyses, which reconcile data from individual platforms, sensors and overpasses, and fill in cloud obscured regions by optimal interpolation. The L4 feature resolution is degraded, compared to input L2P/3U, with no measure provided of this degradation or identification which grids contain clear-sky observations versus those created by estimation (modeling). There is currently no global SST product based on real observations from all available platforms/sensors, without modeled data. As a result, users either have to rely on L4 products (without knowing what data come from real observations versus those modeled, and how much the satellite data have been smoothed), or deal with huge and ever growing data volumes from multiple L2P/3U data files, and learn how to fuse/aggregate those, themselves. In response to multiple users’ requests, NOAA started developing a new multi-sensor, high-resolution, sensor-agnostic gridded L3 SST product, with no modeled data added, which maximally preserves the original sensors’ resolution. In creating such collated and super-collated (L3C/S) products, several issues must be addressed, including minimizing the effect of residual cloud leakages, which are always present in the L2P/3U data, on the L3C/S product, while maximally preserving the feature-resolution present in the original satellite imagery. This aspect of data fusion is the focus of this study.

New-generation geostationary sensors, including the Advanced Baseline Imager (ABI) onboard GOES-16/17 and the Advanced Himawari Imager (AHI, a twin to ABI) onboard Himawari-8, capture infrared images with 2km nadir resolution every 10 or 15 minutes. This high temporal resolution is a unique feature of geostationary sensors, facilitating studies of SST diurnal variability. A collated-in-time geostationary SST product developed for the NOAA Advanced Clear-Sky Processor for Oceans (ACSPO) SST system is able to exploit the temporal information in geostationary SST images by using the temporal context to separate the effects of faster moving clouds and other atmospheric formations from the slower evolution of the SST field. This significantly improves spatial coverage by using measurements from the nearest clear-sky looks in time, reduces the overall noise in the SST time series, and allows for a more accurate characterization of SST diurnal variability. Moreover, the data volume is significantly reduced by reporting at a reduced hourly temporal rate, which is sufficient to resolve the main features in the diurnal cycle. We present an overview of the collation algorithm, sample validation data, and examples of ocean phenomena, such as thermal fronts, diurnal warming, and tidal motion, which have been observed by new-generation sensors. The improved spatial coverage and temporal resolution give us an unprecedented opportunity to investigate the sub-daily time evolution of these phenomena.

NOAA produces operational sea surface temperature (SST) data products from two VIIRS sensors, flown onboard NPP (launched in Oct 2011) and NOAA-20 (N20, aka. JPSS-1 prior to launch; launched in Nov 2017). These first two satellites in the US new generation Joint Polar Satellite System (JPSS) series will be followed by JPSS-2 to -4, planned for launch in 2022, 2026, and 2031, respectively. The goal of VIIRS SST Reanalyses (RANs) is to generate stable, accurate and consistent data products, which currently include Level 2P (L2P; swath projection; 26 Gb/day) and gridded 0.02° L3U (0.45 Gb/day). The RAN comprises multiples steps, including 1) RDR-to-SDR conversion, i.e., reprocessing of all available VIIRS L0 (“Raw Data Records”, RDRs) into improved L1b (“Sensor Data Records”, SDRs), using latest and most accurate sensor calibration, 2) pre-processing, which includes destriping the radiances, resampling VIIRS imagery (to minimize the effect of bow-tie deletions and distortions), and aggregating original 86-sec granules into 10-min SDR granules; 3) feeding those into the latest NOAA Advanced Clear-Sky Processor for Oceans (ACSPO) enterprise SST code (currently, version 2.61), and producing L2P; 4) gridding L2P data and producing L3U; 5) matching L2P/3U data with several accurate Level 4 (L4) analyses, and with quality controlled in situ SSTs from the NOAA in situ SST Quality Monitor, iQuam, system; 6) calculation of the corresponding performance statistics – global daily mean biases and standard deviations, SDs, of various paired SST differences, ΔTs, stratified by day and night, and displaying them in the NOAA SST Quality Monitor (SQUAM) web-based system; 6) generation of SST imagery over ~30 regional targets and routine monitoring in another NOAA system, ACSPO Regional Monitor for SST (ARMS); 7) calculation of brightness temperature (BT) differences between measured BTs and those simulated using the fast Community Radiative Transfer Model (CRTM), as a part of ACSPO L2P processing, and displaying in another NOAA web-based system, MICROS; 8) product archival at NOAA (CoastWatch and NCEI) and at NASA/JPL PO.DAAC. NPP RAN1, performed in late 2015 jointly with UW/CIMSS using ACSPO v2.40, covered a period from Mar’2012-Dec’2015. The data from Jan’2016 – on have been supplemented from operational ACSPO products with various versions (2.41, 2.60, and 2.61). Some issues, unresolved in RAN1, are now being addressed in RAN2. The most important features of RAN2 include the addition of N20 data; fixing quarterly spikes in SST time series (resulting from the VIIRS black-body warm-ups/cool-downs, WUCDs); and using a consistent ACSPO version 2.61 for the full NPP and N20 records. As of this writing, 64 months of NPP (Jan 2014 – Apr 2019) and 16 months of N20 (Jan 2018 – Apr 2019) RAN2 data have been generated. The remaining two years of NPP data (Feb 2012 – Dec 2013), are being processed. The SST records appear stable, and consistent with in situ data and across NPP/N20. The RAN2 data are currently being archived at the NASA/JPL PO.DAAC, and NOAA CoastWatch and NCEI archives.

MODIS has 16 Thermal Emissive Bands (TEBs) whose wavelengths range from 3.7μm to 14.4μm and are calibrated using scan-by-scan observations of its on-board blackbody (BB). Specific Earth surface targets are used to track the long-term consistency, stability, and relative bias between the two MODIS instruments onboard the Terra and Aqua satellites. Previous studies displayed that cold targets have exhibited a relative bias (between Terra and Aqua MODIS Band 31 (11μm)) of less than 0.10K for more than a decade. However, these brightness temperatures are typically outside the MODIS BB calibration range. Hence, in-situ and remotely-sensed (RS) sea surface temperature (SST) measurements provide useful references for the more characteristic higher scene temperatures. Prior literature indicates that the MODIS TEBs have not been extensively evaluated for calibration stability using warmer targets for Collection 6.1 (C6.1), as it was recently implemented. This manuscript expands on the techniques previously applied to all TEBs using SST ocean buoys as reference for C6 by extending the long-term calibration stability and relative bias (between Terra and Aqua MODIS) discussion to present day and incorporating RS data provided by The Group for High Resolution Sea Surface Temperature (GHRSST) as proxy.

GOES-17 (G17), the second satellite in the NOAA GOES-R series with a new Advanced Baseline Imager (ABI) radiometer onboard, was launched in Mar’2018 and declared the NOAA operational GOES-West satellite in Feb’2019. Sea Surface Temperature (SST) is one of the key geophysical products derived from ABI brightness temperatures (BTs) using the NOAA Advanced Clear-Sky Processor for Ocean (ACSPO) enterprise SST system.

Following the launch of G17, an issue was discovered with its ABI loop heat pipe (LHP) that should maintain the ABI electronics (and in particular, Focal Plane Module, FPM) at their intended temperatures. There are two main implications. During normal operations, the G17 FPM temperature is elevated compared to the specifications (and compared to the ABI twin sensor onboard G16, which has been the NOAA operational GOES-East satellite since Dec’2017), leading to overall noisier BTs. During nighttime, especially in some seasons, when more sunlight impinges directly on the G17 ABI, its FPM temperature is elevated even further and becomes very unstable, resulting in increased noise and degraded ABI calibration (due to increased and band-specific emission from the focal plane itself), rendering measured BTs completely unusable for SST retrievals.

The increased noise of the G17 ABI instrument necessitates changes in the ACSPO clear-sky mask (in particular, its spatial uniformity test), and in the collation-in-time algorithm introduced in ACSPO v2.60 for G16. When the ABI temperature is further elevated, its BTs in the longwave IR bands remain biased (although the calibration algorithm is expected to account for changes in the instrument temperature). When the temperature is near its maximum values, they saturate. The collation-in-time algorithm can partially fill in the periods of saturation, while the remaining biases may potentially be addressed empirically. We discuss the challenges imposed by the G17 ABI LHP and resulting BT anomalies, and our progress with mitigating those in the ACSPO SST products at NOAA. Future plans include tweaking ACSPO algorithms, to generate the best possible SST out of G17 BT data, reprocessing of G17 SST data, archival of data with the NASA/JPL Physical Oceanography Distributed Active Archive Center (PO.DAAC) and NOAA NCEI, and work with users to iteratively improve the processing algorithms and derived SST products.

Environmental engineering remote sensing platforms using hyperspectral imagery are often responsible for monitoring coastal regions in order to safeguard national waters. This objective requires determining subsurface turbulent structure from surface water spatial measurements for flow state assessment and decision-making. The inability of remote sensing platforms to penetrate the water column at depth because of turbulence-induced sediment-concentration modulation necessitates using models that dynamically link surface and subsurface structures. A hidden Markov model is applied to large-eddy simulated three-dimensional turbulent flow for the purpose of exploring the feasibility of constructing a system model possessing diagnostic/prognostic statistical power for turbulent state evolution. The data-driven model is based on machine-learning techniques that rely on data statistical covariance structure. Initial results suggest strong nonlinear coupling between the mean flow directed vorticity, cross mean flow velocity, and sediment concentration. In addition, a Bayesian-based state-action estimation algorithm is employed that demonstrates which turbulent feature variables should be focused on at specific times, given the desire to reach a known goal state, and given only a limited number of observations. Such a model gives experimentalists time- and resource-saving guidance for determining what turbulent variables to measure at different times in order to reach a known turbulent goal state.