Sensor Planning for Bayesian Nonparametric Target Modeling

Bayesian nonparametric models, such as the Gaussian process and the Dirichlet process, have been extensively applied for target kinematics modeling in various applications including environmental monitoring, traffic planning, endangered species tracking, dynamic scene analysis, autonomous robot navigation, and human motion modeling. As shown by these successful applications, Bayesian nonparametric models are able to adjust their complexities adaptively from data as necessary, and are resistant to overfitting or underfitting. However, most existing works assume that the sensor measurements used to learn the Bayesian nonparametric target kinematics models are obtained a priori or that the target kinematics can be measured by the sensor at any given time throughout the task. Little work has been done for controlling the sensor with bounded field of view to obtain measurements of mobile targets that are most informative for reducing the uncertainty of the Bayesian nonparametric models. To present the systematic sensor planning approach to leaning Bayesian nonparametric models, the Gaussian process target kinematics model is introduced at first, which is capable of describing time-invariant spatial phenomena, such as ocean currents, temperature distributions and wind velocity fields. The Dirichlet process-Gaussian process target kinematics model is subsequently discussed for modeling mixture of mobile targets, such as pedestrian motion patterns.

Novel information theoretic functions are developed for these introduced Bayesian nonparametric target kinematics models to represent the expected utility of measurements as a function of sensor control inputs and random environmental variables. A Gaussian process expected Kullback Leibler divergence is developed as the expectation of the KL divergence between the current (prior) and posterior Gaussian process target kinematics models with respect to the future measurements. Then, this approach is extended to develop a new information value function that can be used to estimate target kinematics described by a Dirichlet process-Gaussian process mixture model. A theorem is proposed that shows the novel information theoretic functions are bounded. Based on this theorem, efficient estimators of the new information theoretic functions are designed, which are proved to be unbiased with the variance of the resultant approximation error decreasing linearly as the number of samples increases. Computational complexities for optimizing the novel information theoretic functions under sensor dynamics constraints are studied, and are proved to be NP-hard. A cumulative lower bound is then proposed to reduce the computational complexity to polynomial time.

Three sensor planning algorithms are developed according to the assumptions on the target kinematics and the sensor dynamics. For problems where the control space of the sensor is discrete, a greedy algorithm is proposed. The efficiency of the greedy algorithm is demonstrated by a numerical experiment with data of ocean currents obtained by moored buoys. A sweep line algorithm is developed for applications where the sensor control space is continuous and unconstrained. Synthetic simulations as well as physical experiments with ground robots and a surveillance camera are conducted to evaluate the performance of the sweep line algorithm. Moreover, a lexicographic algorithm is designed based on the cumulative lower bound of the novel information theoretic functions, for the scenario where the sensor dynamics are constrained. Numerical experiments with real data collected from indoor pedestrians by a commercial pan-tilt camera are performed to examine the lexicographic algorithm. Results from both the numerical simulations and the physical experiments show that the three sensor planning algorithms proposed in this dissertation based on the novel information theoretic functions are superior at learning the target kinematics with

little or no prior knowledge

Melt climatology estimates for small to giant Antarctic icebergs, links to NetCDF files

We present four melt climatology estimates based on a simulation of Antarctic iceberg drift and melting that includes small, medium-sized, and giant tabular icebergs with a realistic size distribution. Drift and meltdown is simulated using vertical profiles of ocean currents, temperature, and salinity, which goes beyond the present standard in iceberg modeling. The climatology estimates based on simulations of small (SMA), 'small-to-medium'-sized (MED12 & MED123), and small-to-giant icebergs (ALL) exhibit differential characteristics: successive inclusion of larger icebergs leads to a reduced seasonality of iceberg melt and a shift of the mass input to the area north of 58°S, while less melt water is released into the coastal areas. This highlights the necessity to account for larger and giant icebergs in order to obtain accurate melt climatologies. The four monthly melt climatologies [mm/day] are available as netCDF files with 1°x1° spatial resolution and can be used, e.g., for sensitivity studies with uncoupled sea ice-ocean models, or as spatio-temporal templates for the redistribution of land ice from the Antarctic ice sheet over the Southern Ocean in climate models.

Lost at sea: Genetic, oceanographic and meteorological evidence for storm-forced dispersal

[EN] For many species, there is broad-scale dispersal of juvenile stages and/or long-distance migration of individuals and hence the processes that drive these various wide-ranging move- ments have important life-history consequences. Sea turtles are one of these paradigmatic long-distance travellers, with hatchlings thought to be dispersed by ocean currents and adults often shuttling between distant breeding and foraging grounds. Here, we use multi- disciplinary oceanographic, atmospheric and genetic mixed stock analyses to show that juvenile turtles are encountered ‘downstream’ at sites predicted by currents. However, in some cases, unusual occurrences of juveniles are more readily explained by storm events and we show that juvenile turtles may be displaced thousands of kilometres from their expected dispersal based on prevailing ocean currents.

Lost at sea: Genetic, oceanographic and meteorological evidence for storm-forced dispersal

[EN] For many species, there is broad-scale dispersal of juvenile stages and/or long-distance migration of individuals and hence the processes that drive these various wide-ranging move- ments have important life-history consequences. Sea turtles are one of these paradigmatic long-distance travellers, with hatchlings thought to be dispersed by ocean currents and adults often shuttling between distant breeding and foraging grounds. Here, we use multi- disciplinary oceanographic, atmospheric and genetic mixed stock analyses to show that juvenile turtles are encountered ‘downstream’ at sites predicted by currents. However, in some cases, unusual occurrences of juveniles are more readily explained by storm events and we show that juvenile turtles may be displaced thousands of kilometres from their expected dispersal based on prevailing ocean currents.

Geographic distance, water circulation and environmental conditions shape the biodiversity of Mediterranean rocky coasts

Ecological connectivity is important for effective marine planning and biodiversity conservation. Our aim was to identify factors important in influencing variation in benthic community structure on shallow rocky reefs in 2 regions of the Mediterranean Sea with contrasting oceanographic regimes. We assessed beta (β) diversity at 146 sites in the littoral and shallow sublittoral from the Adriatic/Ionian Seas (eastern region) and Ligurian/Tyrrhenian Seas (western region) using a null modelling approach to account for variation in species richness. The distance decay relationship between species turnover within each region and geographic distance by sea was determined using generalised linear models. Mantel tests were used to examine correlations between β?diversity and connectivity by ocean currents, estimated from Lagrangian dispersal simulations. Variation in β diversity between sites was partitioned according to environmental and spatial components using a distance-based redundancy approach. Species turnover along a gradient of geographic distance was greater by a factor of 3 to 5 in the western region than the eastern region, suggesting lower connectivity between sites. β diversity was correlated with connectivity by ocean currents at both depths in the eastern region but not in the western region. The influ-OPEN ACCESS ence of spatial and environmental predictors of β diversity varied considerably between regions, but was similar between depths. Our results highlight the interaction of oceanographic, spatial and environmental processes influencing benthic marine β diversity. Persistent currents in the eastern region may be responsible for lower observed β diversity compared to the western region, where patterns of water circulation are more variable.

Surface currents of the Eastern Tropical Pacific Ocean

(PDF contains 39 pages.)

Submarine pyroclastic deposits formed at the Soufrière Hills volcano, Montserrat (1995–2003) : what happens when pyroclastic flows enter the ocean?

The Soufrière Hills volcano, Montserrat, West Indies, has undergone a series of dome growth and collapse events since the eruption began in 1995. Over 90% of the pyroclastic material produced has been deposited into the ocean. Sampling of these submarine deposits reveals that the pyroclastic flows mix rapidly and violently with the water as they enter the sea. The coarse components (pebbles to boulders) are deposited proximally from dense basal slurries to form steep-sided, near-linear ridges that intercalate to form a submarine fan. The finer ash-grade components are mixed into the overlying water column to form turbidity currents that flow over distances >30 km from the source. The total volume of pyroclastic material off the east coast of Montserrat exceeds 280 × 106 m3, with 65% deposited in proximal lobes and 35% deposited as distal turbidites.

Mechanisms of formation of the Arabian Sea mini warm pool in a high-resolution Ocean General Circulation Model

An Ocean General Circulation Model of the Indian Ocean with high horizontal (0.25 degrees x 0.25 degrees) and vertical (40 levels) resolutions is used to study the dynamics and thermodynamics of the Arabian Sea mini warm pool (ASMWP), the warmest region in the northern Indian Ocean during January-April. The model simulates the seasonal cycle of temperature, salinity and currents as well as the winter time temperature inversions in the southeastern Arabian Sea (SEAS) quite realistically with climatological forcing. An experiment which maintained uniform salinity of 35 psu over the entire model domain reproduces the ASMWP similar to the control run with realistic salinity and this is contrary to the existing theories that stratification caused by the intrusion of low-salinity water from the Bay of Bengal into the SEAS is crucial for the formation of ASMWP. The contribution from temperature inversions to the warming of the SEAS is found to be negligible. Experiments with modified atmospheric forcing over the SEAS show that the low latent heat loss over the SEAS compared to the surroundings, resulting from the low winds due to the orographic effect of Western Ghats, plays an important role in setting up the sea surface temperature (SST) distribution over the SEAS during November March. During March-May, the SEAS responds quickly to the air-sea fluxes and the peak SST during April-May is independent of the SST evolution during previous months. The SEAS behaves as a low wind, heat-dominated regime during November-May and, therefore, the formation and maintenance of the ASMWP is not dependent on the near surface stratification.

Surface freshwater from Bay of Bengal runoff and Indonesian Throughflow in the Tropical Indian Ocean

According to recent estimates, the annual total continental runoff into the Bay of Bengal (BoB) is about 2950 km 3, which is more than half that into the entire tropical Indian Ocean (IO). Here we use climatological observations to trace the seasonal pathways of near surface freshwater from BoB runoff and Indonesian Throughflow (ITF) by removing the net contribution from precipitation minus evaporation. North of 20 degrees S, the amount of freshwater from BoB runoff and ITF changes with season in a manner consistent with surface currents from drifters. BoB runoff reaches remote regions of the Arabian Sea; it also crosses the equator in the east to join the ITF. This freshwater subsequently flows west across the southern tropical IO in the South Equatorial Current.

The Indian Ocean forecast system

In order to meet the ever growing demand for the prediction of oceanographic parametres in the Indian Ocean for a variety of applications, the Indian National Centre for Ocean Information Services (INCOIS) has recently set-up an operational ocean forecast system, viz. the Indian Ocean Forecast System (INDOFOS). This fully automated system, based on a state-of-the-art ocean general circulation model issues six-hourly forecasts of the sea-surface temperature, surface currents and depths of the mixed layer and the thermocline up to five-days of lead time. A brief account of INDOFOS and a statistical validation of the forecasts of these parametres using in situ and remote sensing data are presented in this article. The accuracy of the sea-surface temperature forecasts by the system is high in the Bay of Bengal and the Arabian Sea, whereas it is moderate in the equatorial Indian Ocean. On the other hand, the accuracy of the depth of the thermocline and the isothermal layers and surface current forecasts are higher near the equatorial region, while it is relatively lower in the Bay of Bengal.

Estimating wind stress at the ocean surface from scatterometer observations

Wind stress is the most important ocean forcing for driving tropical surface currents. Stress can be estimated from scatterometer-reported wind measurements at 10 m that have been extrapolated to the surface, assuming a neutrally stable atmosphere and no surface current. Scatterometer calibration is designed to account for the assumption of neutral stability; however, the assumption of a particular sea state and negligible current often introduces an error in wind stress estimations. Since the fundamental scatterometer measurement is of the surface radar backscatter (sigma-0) which is related to surface roughness and, thus, stress, we develop a method to estimate wind stress directly from the scatterometer measurements of sigma-0 and their associated azimuth angle and incidence angle using a neural network approach. We compare the results with in situ estimations and observe that the wind stress estimations from this approach are more accurate compared with those obtained from the conventional estimations using 10-m-height wind measurements.

Data assimilation using ensemble transform Kalman filter (ETKF) in ROMS model for Indian Ocean

Study of Oceans dynamics and forecast is crucial as it influences the regional climate and other marine activities. Forecasting oceanographic states like sea surface currents, Sea surface temperature (SST) and mixed layer depth at different time scales is extremely important for these activities. These forecasts are generated by various ocean general circulation models (OGCM). One such model is the Regional Ocean Modelling System (ROMS). Though ROMS can simulate several features of ocean, it cannot reproduce the thermocline of the ocean properly. Solution to this problem is to incorporates data assimilation (DA) in the model. DA system using Ensemble Transform Kalman Filter (ETKF) has been developed for ROMS model to improve the accuracy of the model forecast. To assimilate data temperature and salinity from ARGO data has been used as observation. Assimilated temperature and salinity without localization shows oscillations compared to the model run without assimilation for India Ocean. Same was also found for u and v-velocity fields. With localization we found that the state variables are diverging within the localization scale.

Technologies for Measuring Currents in Coastal Environments, Portland, Maine, October 26-28, 2005:workshop proceedings

The Alliance for Coastal Technologies (ACT) Workshop entitled "Technologies for Measuring Currents in Coastal Environments" was held in Portland, Maine, October 26-28, 2005, with sponsorship by the Gulf of Maine Ocean Observing System (GoMOOS), an ACT partner organization. The primary goals of the event were to summarize recent trends in nearshore research and management applications for current meter technologies, identify how current meters can assist coastal managers to fulfill their regulatory and management objectives, and to recommend actions to overcome barriers to use of the technologies. The workshop was attended by 25 participants representing state and federal environmental management agencies, manufacturers of current meter technologies, and researchers from academic institutions and private industry. Common themes that were discussed during the workshop included 1) advantages and limitations of existing current measuring equipment, 2) reliability and ease of use with each instrument type, 3) data decoding and interpretation procedures, and 4) mechanisms to facilitate better training and guidance to a broad user group. Seven key recommendations, which were ranked in order of importance during the last day of the workshop are listed below. 1. Forums should be developed to facilitate the exchange of information among users and industry: a) On-line forums that not only provide information on specific instruments and technologies, but also provide an avenue for the exchange of user experiences with various instruments (i.e. problems encountered, cautions, tips, advantages, etc). (see References for manufacturer websites with links to application and technical forums at end of report) b) Regional training/meetings for operational managers to exchange ideas on methods for measuring currents and evaluating data. c) Organize mini-meetings or tutorial sessions within larger conference venues. 2. A committee of major stakeholders should be convened to develop common standards (similar to the Institute of Electrical and Electronics Engineers (IEEE) committee) that enable users to switch sensors without losing software or display capabilities. (pdf contains 28 pages)

Seasonality in the coastal ocean of North America, 35°N - 48°N [abstract]

EXTRACT (SEE PDF FOR FULL ABSTRACT): The seasonal cycles of coastal wind stress, adjusted sea level height (ASL), shelf currents and water temperatures off the west coast of North America (35°N to 48°N) were estimated by fitting annual and semiannual harmonics to data from 1981-1983. Longer records of monthly ASL indicate that these two harmonics adequately represent the long-term monthly average seasonal cycle, and that the current measurement period is long enough to define the seasonal cycles, with relatively small errors in magnitude and phase.