Dataset: Ocean aluminum cycle model output

Name: Ocean aluminum cycle model output containing model-predicted aluminum distribution and 60 different estimates of dust deposition
Published: 2024-03-15
Keywords: aluminum, GEOTRACES, Dust deposition, oceans

View Data: Data not available yet

Cite This Dataset

Project:

Collaborative research: Combining models and observations to constrain the marine iron cycle (Fe Cycle Models and Observations)

Principal Investigator:

Thomas Weber (University of Rochester)

Scientist:

Hairong Xu (University of Rochester)

BCO-DMO Data Manager:

Amber D. York (Woods Hole Oceanographic Institution, WHOI BCO-DMO)

Version:

Version Date:

2024-03-15

Restricted:

Validated:

Current State:

Data not available

Ocean aluminum cycle model output containing model-predicted aluminum distribution and 60 different estimates of dust deposition

Abstract:

Here present model output from a data-assimilating model of the ocean aluminum (Al) cycle. The model simulates inputs of Al from atmospheric dust, and internal cycling due to reversible scavenging and biological cycling, and assimilates data from the GEOTRACES archive. Full code and inputs for the model can be accessed at Zenodo (https://zenodo.org/records/10139317). This file contains the aluminum model-predicted aluminum distribution from the fully-optimized model, as well as 60 different estimates of dust deposition from a range of model configurations to quantify uncertainty in this process. A code is provided to synthesize and plot these results in a figure summarizing the oceanic aluminum budget as in Xu and Weber 2021.

Expand/Collapse All

Description

Data are presented in both the original Matlab data format (file "Xu_2021_model_output.mat") and a non-proprietary plain-text format "Xu_2021_model_output.zip" which contains a .txt file for each variable stored in Xu_2021_model_output.mat. See file descriptions for more information about the structure of these files.

Methods & Sampling

A model of the ocean aluminum (Al) cycle was developed in the MATLAB programming language, using the Ocean Circulation Inverse Model v2 to resolve physical transport. The model resolves deposition and dissolution of dust at the sea surface, and internal cycling of Al in the ocean by reversible scavenging, biological uptake, and sediment resuspension. The model is solved for the steady-state aluminum distribution, which is then compared to observations from the GEOTRACES Intermediate Data Product 2017 to optimize uncertain parameters. 60 different configurations of the model are optimized, representing all combinations of 12 dust deposition fields and 5 configurations of the OCIM transport model. The model equations are described in Xu and Weber 2021 (https://doi.org/10.1029/2021GB007049) and full model code and inputs are available for download on Zenodo (https://zenodo.org/records/10139317).

This dataset contains the main model results: the fully optimized model-predicted aluminum distribution, as well as estimates of Al supply to the surface ocean from dust deposition from the 60 model configurations described above.

Location:
Output from a global 3-dimensional model steady state model. Aluminum concentration is on a 3-dimensional grid, and dust deposition rates on a 2-dimensional grid. Grid information is found in the grid structure as follows:

grid.xt - Longitude
grid.yt - Latitude
grid.zt - Depth

Data Processing Description

The model was developed and run on the University of Rochester BlueHive computing cluster.

The model runs using custom code developed for this project and is available on the Zenodo repository (https://zenodo.org/records/10139317).

BCO-DMO Processing

Data within Xu_2021_model_output.mat was loaded using Matlab online version r2023b. Each variable was exported as individual plain-text files and added as a non-propriety form of this dataset (see Data File "Xu_2021_model_output.zip"). Struct contents of grid and basin_masks which were added into subfolders "grid" and "basin_masks."

>> load("/MATLAB Drive/Xu_2021_model_output.mat")
>> whos
Name Size Bytes Class Attributes

Alconc 91x180x24 3144960 double
Aldep 91x180x60 7862400 double
basin_masks 1x1 524832 struct
cost 1x60 480 double
grid 1x1 3279200 struct
mod_names 1x60 7798 cell

>> writematrix(Alconc,'Xu_2021_model_output/Alconc.txt');
>> writematrix(Aldep,'Xu_2021_model_output/Aldep.txt');
>> writematrix(basin_masks.SOC,'Xu_2021_model_output/basin_masks/SOC.txt');
>> writematrix(basin_masks.ATL,'Xu_2021_model_output/basin_masks/ATL.txt');
>> writematrix(basin_masks.PAC,'Xu_2021_model_output/basin_masks/PAC.txt');
>> writematrix(basin_masks.IND,'Xu_2021_model_output/basin_masks/IND.txt');
>> writematrix(cost,'Xu_2021_model_output/cost.txt');
>> writematrix(grid.x,'Xu_2021_model_output/grid/x.txt');
>> writematrix(grid.y,'Xu_2021_model_output/grid/y.txt');
>> writematrix(grid.z,'Xu_2021_model_output/grid/z.txt');
>> writematrix(grid.AREA,'Xu_2021_model_output/grid/AREA.txt');
>> writematrix(grid.M3d,'Xu_2021_model_output/grid/M3d.txt');
>> writecell(mod_names,'Xu_2021_model_output/mod_names.txt');