This folder "model runs" include model code, input data and output data in netcdf format.

Following the instructions readers are able to generate the data that is used in our paper, in his/her local machine:
1) create a folder to run the model, for example, named with "COA"
2) copy the content from folder "shared_files" into "COA"
3) depending on which model run you want to repeat, you have to choose an experimental run folder to copy things from. For example, "oliv10_con" includes the essential files to generate model run "oliv10_con". If you want to generate the output for this run, you will have to copy "results" to "COA" and "02" to "COA/updates"
4) In "results" folder, our model output is located in netcdf format. You have to go to "make.in" file in this folder, and change the forth line started with:
"version_directory …."   
into the directory where your "COA" folder is located: ".../COA"
5) compile the executive file. If you are using mac or linux, just command "../mk e" in terminal within "results" folder
6) Execute "./Yuming > runlog &" to run the model within "results" folder

Our UVic model version is based on the Kiel version as published by Keller, D., Oschlies, A. und Eby, M. (2012) A new marine ecosystem model for the University of Victoria Earth System Climate Model. Geoscientific Model Development, 5 . pp. 1195-1220. DOI 10.5194/gmd-5-1195-2012.

For our Coastal Ocean Alkalinization (COA) experiments we modified four  .F and .h files to UVic model Kiel version and add a new one (coa.F), for details see the list below.

COA-experiments require the option "o_coa" to be turned on in mk.in. Please check the mentioned files for further details. Example is given in "Oliv100_Omega3.4" run folder.   
Please search key word “olivine” for locating the relevant module.


- "coa.F" located at "updates/02/source/common/"
where olivine particle properties and dissolution process are characterized

- "gosbc.F" located at "updates/02/source/common/"
where COA induced alkalinity is added whenever air-sea boundary conditions are updated.

- "csbc.h" located at "updates/02/source/common/"
where COA related public variables are defined

- "diag.h" located at "updates/02/source/mom/"
where COA related diagnostic (global average) public variables are defined

-"diag.F" located at "updates/02/source/mom/"
where COA related diagnostic variables are updated

-"mom_tsi.F" located at "updates/02/source/mom/"
where COA related global average output is written to the tsi.nc output file.

Algorithm to simulate COA was initially introduced by Ellias Yuming Feng (see Thesis: Modeling Assessment of Climate Engineering, University Kiel, 2017). Code has been improved thereafter by Wolfgang Koeve (wkoeve@geomar.de). 

In tsi file, global averaged data has been stored. The related variables are namely:
F_OLIVINE:  global averaged added olivine in alkalinity unit (Alk umol m-2 s-1)
F_OLIVINE_DIS: global averaged dissolved olivine in alkalinity unit (Alk umol m-2 s-1)
F_OLIVINT: global total olivine addition (Pmol yr-1)
F_OLIVINT_DIS：global total olivine dissolution (Pmol yr-1)

In tavg.nc file, global distributed data in each grid box has been stored. The related variable is:
F_alk: global distributed olivine dissolution in each grid box in alkalinity unit (Alk umol m-2 s-1)

For model runs with weekly output, only global sea surface temperature, aragonite omega and pCO2 are saved in respective folders. Readers have to conduct model runs themselves if they want to check other variables.