13.4 Initial Conditions

13.4.1 CONUS (NEON/FIA/BADM) Initial Conditions.

To convert initial condition data into the PEcAn Standard and then into the model formats we follow three main steps:

  1. Downloading vegetation info -converting the vegetation info into .rd file format
  2. Creating ensemble members based on veg file -resamples veg file to create ensmeble members
  3. Converting ensemble members to model specific format

Common Questions regarding IC Data:

How do I know if my data product is supported by ic_process workflow? -The current data products supported by ic_process are: NEON, FIA, and BADM. That said you if your vegetation files are registered with BETY and abide by the BETY formats and column names you will be able to use ic_process to create ensemble members and convert the ensemble members to model specific format. See section 6.3.3 for more info. Can I use ic_process outside of the PEcAn workflow? -Short answer is yes, ic_process can be called as a function all by itself outside of do_conversions. What about soil initial conditions? -Please reference section 6.3.6 for more info.

ic_process workflow:

Required inputs for ic_process are: settings object with and sections (see pecan.xml section for more info), dir or location where you would to store the output files, input object that contains all info from settings\(run\)inputs$poolinitcond, and the overwrite setting the default is overwrite = FALSE.

get_veg_module() This function will either download your vegetation info from: NEON_veg, FIA, BADM sources or query BETY for existing veg files previously downloaded. The input and dbfile ids of your veg file are then added to a list getveg.id.

ens_veg_module() This function will create ensemble member ncdf files by resampling the veg file supplied in getveg.id.

put_veg_module() This function will convert the ensemble member ncdf files into model specific format. Currently the supported models are ED2 and SIPNET.

13.4.2 North America (NA) Initial Conditions.

To create initial condition files across North America, you will need to strictly follow the script located at ~/pecan/modules/assim.sequential/inst/anchor/IC_prep_anchorSites.Rmd. Within the script we will be following those main steps: 1. Loading settings.xml file and specify paths. 2. Downloading/extracting estimations of four major carbon/water pools (leaf, wood, soil C, soil water) into by-site and by-ensemble tables. 3. Doing unit conversion. For each ensemble of each site, we will be preparing the poolinfo object consisting of converted pool estimations. 4. We will finally be writing the NC files through the PEcAn.SIPNET::veg2model.SIPNET function. 5. Within the loop, we will store the NC file paths into the settings object and rewrite the settings into the XML file to the destination.

Within the script we proposed the following new datasets for handling the NA initial condition preparations:

  1. The leaf carbon is initialized with MODIS LAI observations and the SLA for the corresponding PFT.
  2. The above ground biomass (AGB) is initialized with the 2010 global AGB map (DOI: https://doi.org/10.3334/ORNLDAAC/1763).
  3. The soil moisture (SM) is initialized with the SM estimations starting from 1978 (DOI: 10.24381/cds.d7782f18).
  4. The soil organic carbon (SOC) is initialized with ISCN SOC estimations (data already prepared on PEcAn, use PEcAn.data.land::iscn_soc to load) based on the level 2 ecoregion map (pre-downloaded using the following link: https://www.epa.gov/eco-research/ecoregions).