This is the repository for the modified Larval TRNASport Lagrangian (LTRANS) model (Liang et al., 2021). The original LTRANS model described in North et al. 2008. The modified code contains the implementation of the parameterization including the effects of subsurface currents on buoyant particles (Liang et al., 2018). The example is an application of the model on riverborne plastics over the Louisiana/Texas continental shelf.
- Installation
Create a local copy of LTRANS_microplastics
by
git clone https://github.com/lsuocean/LTRANS_microplastics
- Change to the input directory
cd LTRANS_microplastics
- Create input and output directory (as an example, choose 0.0005m/s for the rising speed of the buoyant particles, i.e.,
sink= 0.0005
inLTRANS.data
)
mkdir input
mkdir input/ROMS_solution
mkdir output
mkdir output/0.0005
- Download the ROMS solution in netcdf (nc) format from http://barataria.tamu.edu:8080/thredds/catalog.html, and rename the downloaded nc file (say the original file name for year 2008 is ROMS_2008.nc) as 20080001.nc, and save the file in directory input/ROMS_solution
cd input/ROMS_solution
mv ROMS_2008.nc 20080001.nc
- Create the particle location Initial_part_location.csv.
You can find the example of Initial_part_location.csv
and user guide on https://github.com/LTRANS/LTRANSv.2b.
- Here we use 1200 particles in total, therefore numpar=1200 in LTRANS.data
- Revise the setup in LTRANS.data
- The location of nc file
NCgridfile='./input/ROMS_solution/20080001.nc'
- Setting for the ROMS input files
$romsoutput
prefix='./input/ROMS_solution/2008' ! NetCDF Input Filename prefix
suffix='.nc' ! NetCDF Input Filename suffix
filenum = 0001 ! Number in first NetCDF input filename
numdigits = 4 ! Number of digits in number portion of file name (with leading zeros)
startfile = .FALSE.
$end
- LTRANS output path
outpath = './output/0.0005/' ! Location to write output .csv and/or .nc files
- rising speed for the buoyant particles
sink= 0.0005 ! in m/s
- compile the model (use the correct PATH and LIBDIR for NETCDF and HDF5, see below as an example for LSU supercomputer)
- Here use ifort as the default Fortran compiler
export NETCDF=/usr/local/packages/netcdf/4.1.3/Intel-13.0.0
export PATH=/usr/local/packages/netcdf/4.1.3/Intel-13.0.0/bin:$PATH
export NETCDF_LIBDIR=/usr/local/packages/netcdf/4.1.3/Intel-13.0.0/lib
export LD_LIBRARY_PATH=/usr/local/packages/netcdf/4.1.3/Intel-13.0.0/lib:/usr/local/packages/hdf5/1.8.9/Intel-13.0.0/lib/:${LD_LIBRARY_PATH}
export NETCDF_INCDIR=/usr/local/packages/netcdf/4.1.3/Intel-13.0.0/include
export HDF5_INCDIR=/usr/local/packages/hdf5/1.8.9/Intel-13.0.0/include
export HDF5_LIBDIR=/usr/local/packages/hdf5/1.8.9/Intel-13.0.0/lib
make clean
make
- Run the model (the screen output will be saved in jobout.dat)
./LTRANS.exe > jobout.dat
The output files will be saved in './output/0.0005/'
.
- numpar: total number of partilces
- sink: rising speed (positive value) in m/s --- it is used to replace variable
wb
in the subroutinefind_currents
inLTRANS.f90
- The effective diffusivity k_{xx} and k_{yy} in equations (6a) and (6b) in Liang et al. (2021) are computed in the new subroutine
find_current
in the newLTRANS.f90
ask11
andk22
, which are used to calculate the subgrid-scale displacementXgrd
andYgrd
[i.e., x_{sgs} and y_{sgs} in equations (5a) and (5b) in Liang et al. (2021)]. - The weighted-averaged velocities, i.e., ubar and vbar in equations (3a) and (3b) in Liang et al. (2021) are calculated and saved in variable
ubar
andvbar
in the new subroutinefind_current
in the newLTRANS.f90
, respectively.
For reference, please cite:
J.-H. Liang, J. Liu, M. C. Benfield, D. Justic, D Holstein, B. Liu, R. Hetland, D. Kobashi, C. Dong, and W. Dong, 2021. Including the Effects of Subsurface Currents on Buoyant Particles in Lagrangian Particle Tracking Models: Model Development and its Application to the Study of Riverborne Plastics over the Louisiana/Texas Shelf. Ocean Modelling. doi:10.1016/j.ocemod.2021.101879.
If you have any questions regarding the model, please feel free to open an issue or send an email to Dr. Jun-Hong Liang (jliang@lsu.edu) at Louisiana State University. You are welcome to visit our website for more information about our research: https://faculty.lsu.edu/liang/