Abstract:

The dataset includes output from a computer model designed to simulate the formation of an intense diatom bloom that occurred soon after the Deepwater Horizon incident, the subsequent formation of diatom aggregates and the scavenging of dispersed oil as the aggregates sank through the water column and eventually settled on the seafloor. This intense diatom/oil sedimentation event was captured by a sediment trap moored near to the spill site in August/September of 2010. Model sensitivity data are also included.

Suggested Citation:

Simone Francis and Uta Passow. 2018. Modeling study of oil transport to seafloor by sinking diatom aggregates following an intense bloom. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/N7MW2FM5

Purpose:

The purpose of the model is: (a) to demonstrate the potential for sinking diatom aggregates to scavenge dispersed oil from the water column and deposit it on the seafloor on time scales and in quantities consistent with those measured by sediment traps, and (b) to uncover the most important parameters and mechanisms that impact oil sedimentation amounts and rates.

Data Parameters and Units:

Bloom layer: Time (days); Nitrate concentration (uM); Chlorophyll a concentration (mg/m^3); Total particulate organic carbon concentration (mg/m^3); Total particulate organic carbon flux (mg m^2/d); Particulate organic carbon concentration (mg/m^3); Spherical diameter (mm) Fluxes at trap depth: Time (days); Total particulate organic carbon flux (mg m^2/d); Depth (m); Total oil flux (mg m^2/d) Average fluxes at trap depth: Time (days); Averaged total particulate organic carbon flux (mg m^2/d); Averaged total oil flux (mg m^2/d) Baseline parameter values: Parameter name; Symbol; Baseline value; Units; Sticking probability (dimensionless); Carbon to chlorophyll ratio (mg/mg); Carbon to nitrogen ratio (mol/mol); Cell diameter (um); Maximum aggregate diameter (mm); Fractal dimension (dimensionless); Average shear (per second); Half-saturation concentration for nitrate uptake (uM); Mixed layer depth (m); Maximum growth rate (per day); Initial nitrate concentration (uM); Initial oil concentration (ppb); Particulate organic carbon per cell (pg/cell); Oil degradation rate (per day); Particulate organic carbon degradation rate (per day); Sinking velocity (m/d) Sensitivity results: Model run; Parameter varied; Value; Units; Total particulate organic carbon (POC) deposition (mg C/m^2); Total oil deposition (mg oil/m^2); Oil carbon as % of POC (%); % of total oil scavenged (%)

Methods:

The model used in this study begins with a simple mixed-layer particle aggregation model that is extended to include scavenging of dispersed oil by aggregates, degradation of oil and aggregate carbon during transit through the water column, and sedimentation of oil-containing aggregates to the seafloor. The particle aggregation model is used to simulate the formation of a diatom bloom, the subsequent coagulation of diatom cells into aggregates and the sinking of these aggregates out of the bloom layer. It is based on coagulation theory and includes three mechanisms for particle collision: Brownian motion, wherein the random motion of particles results in collision; turbulent shear, wherein particles moving at different speeds due to velocity gradients come into contact; and differential sedimentation, wherein particles sinking at different speeds collide. The coagulation equations are solved numerically using the sectional method, in which the particle size spectrum is broken into discrete size classes (sections) in such a way as to make the calculations very efficient. The bloom is initiated by a pulse of nitrate. Cells divide according to the Monod model, where growth is a function of a maximum growth rate, ambient nitrate concentration and the half-saturation concentration for nitrate uptake. Growth slows as the nitrate concentration decreases and stops altogether when nitrate is depleted. Cells continue to aggregate and sink from the bloom layer even after growth stops. Cell loss due to grazing is not included in the model. The aggregation model is coupled to a scavenging and degradation model which simulates incorporation of dispersed oil into the aggregates via scavenging as well as degradation of diatom carbon and oil as the aggregates sink through the water column toward the seafloor. Dispersed oil is assumed to be evenly distributed throughout the water column, and all oil contained in the volume of water traversed by a sinking aggregate is assumed to be swept up, or scavenged, by that aggregate. This is equivalent to assuming collision between aggregates and oil droplets by differential sedimentation with a sticking probability of 1. The amount of material (both diatom carbon and oil) that reaches the seafloor is a function of the sinking velocities of the aggregates and the degradation rates of the respective components. Sinking velocities of aggregates that form in the bloom layer are assumed to be a function of size, and are specified according to the results of in-house experiments on diatom aggregates, data from which are available on GRIIDC (doi: 10.7266/N7H70D81). As the diatom-oil aggregates sink through the water column, the organic carbon and oil they contain degrade at temperature-dependent (or, correspondingly, depth-dependent) rates taken from the literature. These additions to the basic aggregation model allow this one-dimensional Aggregation-Scavenging-Degradation-Sedimentation (ASDS) model to provide time series estimates of the amount of cell carbon and oil that reaches the deep ocean. The model is used to calculate fluxes at 1540 m, which was the depth at which a sediment trap was moored near the Deepwater Horizon spill site in August and September of 2010. This trap captured a large diatom and oil sedimentation event. A comparison of the modeled carbon and oil fluxes at the trap depth with those measured by the trap required integrating the modeled fluxes over the 21-day deployment period for each cup of the trap. The total amounts of diatom carbon and oil deposited on the seafloor as a result of the diatom bloom and its scavenging of dispersed oil were estimated by integrating deep fluxes over the duration of the bloom, i.e. without the restriction of a 21-day trap collection period. This allows for comparison between our baseline model run and sensitivity runs done with different parameter values. The sensitivity of the model to changes in parameter values was tested by systematically varying the value of one parameter at a time and comparing the results with those of the baseline model run. A total of sixteen parameters were varied, generally from plus or minus 25 to 50% of their nominal values. Metrics used for comparison with the baseline model were total diatom carbon and total oil deposited at depth over the course of the bloom.

Individual Files: