Dynamic modeling of long-term sedimentation in the Yadkin River basin

Modeling of sediment transport in relation to changing land-surface conditions against a background of considerable natural variability is a challenging area in hydrology. Bayesian dynamic linear models (DLMs) however, offer opportunities to account for non-stationarity in relationships between hydrologic input and basin response variables. Hydrologic data are from a 40 years long record (1951–1990) from the 5905 km² Yadkin River basin in North Carolina, USA. DLM regressions were estimated between log-transformed volume-weighted sediment concentration as a response and log-transformed rainfall erosivity and river flow, respectively, as input variables. A similar regression between log-transformed river flow and log-transformed basin averaged rainfall was also analyzed. The dynamic regression coefficient which reflects the erodibility of the basin decreased significantly between 1951 and 1970, followed by a slowly rising trend. These trends are consistent with observed land-use shifts in the basin. Bayesian DLMs represent a substantial improvement over traditional monotonic trend analysis. Extensions to incorporate multiple regression and seasonality are recommended for future applications in hydrology.     

Markov chain Monte Carlo (MCMC) sampling methods to determine optimal models,model resolution and model choice for Earth Science problems

We present an overview of Markov chain Monte Carlo, a sampling method for model inference and uncertainty quantification. We focus on the Bayesian approach to MCMC, which allows us to estimate the posterior distribution of model parameters, without needing to know the normalising constant in Bayes' theorem. Given an estimate of the posterior, we can then determine representative models (such as the expected model, and the maximum posterior probability model), the probability distributions for individual parameters, and the uncertainty about the predictions from these models. We also consider variable dimensional problems in which the number of model parameters is unknown and needs to be inferred. Such problems can be addressed with reversible jump (RJ) MCMC. This leads us to model choice, where we may want to discriminate between models or theories of differing complexity. For problems where the models are hierarchical (e.g. similar structure but with a different number of parameters), the Bayesian approach naturally selects the simpler models. More complex problems require an estimate of the normalising constant in Bayes' theorem (also known as the evidence) and this is difficult to do reliably for high dimensional problems. We illustrate the applications of RJMCMC with 3 examples from our earlier working involving modelling distributions of geochronological age data, inference of sea-level and sediment supply histories from 2D stratigraphic cross-sections, and identification of spatially discontinuous thermal histories from a suite of apatite fission track samples distributed in 3D.     

Depth of activation on a mixed sediment beach

The relationship between wave height and depth of sediment activation is evaluated on an estuarine beach to determine whether activation depth is less in pebbles than sand. Rods with washers were used to monitor three excavated beach plots filled with (1) pebbles with mean grain size of 11.5 mm; (2) sand and granules; and (3) sand, granules and pebbles. Plots were monitored for 26 events over 27 days. Significant wave heights ranged from 0.18 to 0.40 m and activation depths from 0.02 to 0.12 m. Activation depths in the pebble plot were less than the other two plots when waves reworked sediment not activated during previous tidal cycles. Proportionality coefficients for activation depth to wave height, when net change was < 0.02 m, were 0.24 in the pebble plot and 0.30–0.31 in the other plots when experimental fill sediments remained and 0.22 to 0.23 in all three plots over the entire monitoring period, which included activation of newly deposited native sediment. Results suggest that for similar wave heights, activation depths in pebbles is lower than in sand, granules and pebbles or sand and granules, but once waves have reworked the sediment there is little difference in activation depths.     

Nitrogen dynamics within a wind-driven eddy

Wind-driven cyclonic eddies are hypothesized to relieve nutrient stress and enhance primary production by the upward displacement of nutrient-rich deep waters into the euphotic zone. In this study, we measured nitrate (NO₃⁻), particulate carbon (PC), particulate nitrogen (PN), their stable isotope compositions (δ¹⁵N-NO₃⁻, δ¹³C-PC and δ¹⁵N-PN, respectively), and dissolved organic nitrogen (DON) within Cyclone Opal, a mature wind-driven eddy generated in the lee of the Hawaiian Islands. Sampling occurred in March 2005 as part of the multi-disciplinary E-Flux study, approximately 4–6 weeks after eddy formation. Integrated NO₃⁻ concentrations above 110 m were 4.8 times greater inside the eddy (85.8±6.4 mmol N m⁻²) compared to the surrounding water column (17.8±7.8 mmol N m⁻²). Using N-isotope derived estimates of NO₃⁻ assimilation, we estimated that 213±59 mmol m⁻² of NO₃⁻ was initially injected into the upper 110 m Cyclone Opal formation, implying that NO₃⁻ was assimilated at a rate of 3.75±0.5 mmol N m⁻² d⁻¹. This injected NO₃⁻ supported 68±19% and 66±9% of the phytoplankton N demand and export production, respectively. N isotope data suggest that 32±6% of the initial NO₃⁻ remained unassimilated. Self-shading, inefficiency in the transfer of N from dissolved to particulate export, or depletion of a specific nutrient other than N may have led to a lack of complete NO₃⁻ assimilation. Using a salt budget approach, we estimate that dissolved organic nitrogen (DON) concentrations increased from eddy formation (3.8±0.4 mmol N m⁻²) to the time of sampling (4.0±0.09 mmol N m⁻²), implying that DON accumulated at rate of 0.83±1.3 mmol N m⁻² d⁻¹, and accounted for 22±15% of the injected NO₃⁻. Interestingly, no significant increase in suspended PN and PC, or export production was observed inside Cyclone Opal relative to the surrounding water column. A simple N budget shows that if 22±15% of the injected NO₃⁻ was shunted into the DON pool, and 32±6% is unassimilated, then 46±16% of the injected NO₃⁻ remains undocumented. Alternative loss processes within the eddy include lateral exchange of injected NO₃⁻ along isopycnal surfaces, remineralization of PN at depth, as well as microzooplankton grazing. A 9-day time series within Cyclone Opal revealed a temporal depletion in δ¹⁵N-PN, implying a rapid change in the N source. A change in NO₃⁻ assimilation, or a shift from NO₃⁻ fueled growth to assimilation of a ¹⁵N-deplete N source, may be responsible for such observations.     

Preliminary landfill siting and related analysis using simple modeling techniques

The evaluation of existing and new landfills requires considerable consideration of environmental factors for the protection of groundwater resources. The initial evaluation of sites can be time-consuming and expensive depending on the assessment approach selected. General planning methods, such as the DRASTIC procedure, coupled with the use of analytical or numerical groundwater models can provide useful site selection tools by improving the quality and quantity of analysis. This paper presents an overview of some promising methods for landfill site evaluation and suggests alternative data sources where few data are present. Examples of some of the methods are given.     

Large-sample estimates of wind fluctuations over the ocean

The best quality wind data from the Norwegian sector of the North Sea, consisting of 3662 20-min time series measured at the top of the Statfjord A drilling derrick, are analyzed. Identification of Autoregressive wind models with Akaike's AIC and Achwarz's BIC measures appears to give rather arbitrary results. Spectral estimation with FFT- and AIC-identified AR-methods give almost identical results in the mean. At the higher frequencies (% MathType!MTEF!2!1!+-% feaafeart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqtubsr% 4rNCHbGeaGqiVu0Je9sqqrpepC0xbbL8F4rqqrFfpeea0xe9Lq-Jc9% vqaqpepm0xbba9pwe9Q8fs0-yqaqpepae9pg0FirpepeKkFr0xfr-x% fr-xb9adbaqaaeGaciGaaiaabeqaamaabaabaaGcbaGaamOzaaaa!36D7!\[f\] > 10^–2 s^–1) the spectrum is estimated to follow the usual inertial subrange law with little variability. The small-scale turbulent intensity is estimated to be very low, even in hurricane conditions. Comparatively, the low-frequency (% MathType!MTEF!2!1!+-% feaafeart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqtubsr% 4rNCHbGeaGqiVu0Je9sqqrpepC0xbbL8F4rqqrFfpeea0xe9Lq-Jc9% vqaqpepm0xbba9pwe9Q8fs0-yqaqpepae9pg0FirpepeKkFr0xfr-x% fr-xb9adbaqaaeGaciGaaiaabeqaamaabaabaaGcbaGaamOzaaaa!36D7!\[f\] ~ 10^–3 s^–1) fluctuations are more energetic than expected. None of the chosen low-frequency characteristica appear to be significantly linearly correlated to the available mean weather variables. However, some nonlinear relations appear to exist.     

Larval Pacific Herring (Clupea pallasi) Survival in Suspended Sediment

Pacific herring early life stages provide good model systems for studying effects of suspended sediments on estuarine organisms. To investigate effects on the herring larval stage, we used environmentally relevant particle concentrations for San Francisco Bay (200?C400?mg/L of particles <50???m in size) and exposure times of 16?h in a novel two-pump sediment suspension mesocosm. There were no mortalities during the 16-h suspended sediment incubation. Following sediment exposure, larvae were cultured in sediment-free water for up to 10?days during which survival and condition were measured. None were affected by previous sediment treatment. Four criteria for larval condition included growth, heart rate, prey capture, and critical swimming velocity. These results provide a framework for implementing regulatory decisions on anthropogenic activities such as dredging.