Input data; new to GETM

[Dan Bartley]

Hello GETM community!

I am a newcomer to GETM and have been spending some time trying to feel my way through the development of a new model. I have been focused to far on the input data, specifically the meteo.nc, bdy.2d.nc, and bdy.3d.nc files. I’ve ended up with a few questions below that I’m struggling to clear up by myself. May I please ask if someone would be able to help me with the below? In each case I’ve tried to explain my current approach, followed by the questions I’m currently having. I’d be very grateful for any answers anyone can provide, or any corrections to my approach that may be wrong!

meteo.nc

• I have so far been working with meteorological data downloaded from the ERA5 reanalysis hourly dataset on single levels. The variables I have been using are: mean sea level pressure, total precipitation, 10m u component of wind, 10m v component of wind, 2m temperature, 2m dewpoint temperature, and total cloud cover. In getm.inp I am setting calc_met=true, and fwf_method=3.
• My main two questions in regards to this were about the calculated evaporation and the precipitation:
  • For the precipitation, I think I am confusing myself about the units. The ERA5 page provides total precipitation as rainfall accumulated in the 1 hour ending at the validity date and time, and the units are m of water in the ERA5 grid box if it were evenly spread over the box. My initial understanding of this was that I provide this data to GETM in m, and set the precip_factor to 1/3600=0.0002777... I am now doubting myself and wondering if some additional account needs to be made for the accumulation area and time, with regards to the interpolation onto my GETM grid and the timing. Can you help clear this up for me?
  • If my 2m dewpoint temperature is provided with the name “dev2”, is that enough for GETM to know that I have provided the dewpoint temperature as opposed to the specific/relative humidity, or is there another setting I have missed in order to specify this? Currently I am receiving outputs of a calculated evaporation, but I’m unfamiliar with how this is calculated so I was unsure how to verify that my input is being interpreted correctly.

bdy.2d.nc

• Up to now I have been putting 2D tidal data into bdy.2d.nc. I have been obtaining this from a regional model downloaded from https://www.tpxo.net/regional that covers my domain, and extracting tidal elevations and velocities (not transports) using the TMD matlab package (https://www.esr.org/research/polar-tide-models/tmd-software/). I am having some difficulty finding the numerical implementation of the Flather boundary condition, and I think this is causing me some uncertainty. May I ask if someone could give me a quick rundown of this or direct me to where I may find it? Specifically:
  • For which grid points should I be obtaining the tidal data? Currently I am expecting it to be the relevant U- and V- points at the very edges of the grid (i.e. those masked as “4” according to the Lateral Boundary Conditions section of the documentation), but I am not sure about this.
  • Am I understanding correctly that for each boundary (east, north, west, south), I only need to input the velocity components that are normal to each boundary? i.e. only u-components are needed for cells on the eastern and western boundaries, and only v-components are needed for cells on the northern and southern boundaries, with corner cells needing both components?
  • In runs of the model so far, outputs of velocity seem always to be zero at the boundaries - is this expected behaviour or would it indicate a problem?

bdy.3d.nc

• Up to now I have been inputting 3D temperature and salinity data downloaded from HYCOM into the bdy.3d.nc file. For each boundary I have located the nearest line of latitude/longitude available in HYCOM. Then for each depth in the HYCOM data, I have done a simple interpolation of values for each of the boundary cell T-point longitudes/latitudes. I have then put this in the bdy.3d.nc file, where my understanding is that zax is a depth below the surface in metres. I have set bdy3d_vel=false in getm.inp, and have only provided temperature and salinity in the bdy.3d.nc file, meaning that the only boundary velocities I am specifying are in the bdy.2d.nc file.
  • This is the input file for which I am more uncertain about the approach; could you please let me know if I ought to be doing any of the above differently?
  • Am I correct in thinking that GETM interpolates the boundary data for the sigma levels using linear interpolation within each column?

Apologies that there’s quite a few things here but as a newcomer it would certainly help me out if someone can help with my questions, or advise on any of my approaches you happen to notice might not be correct. Thank you very much for any help you can give!

Best wishes,

Dan Bartley