tropics

Code to investigate TROPICS' ability to accurately measure the diurnal cycle in MSU/AMSU tropospheric temperature observations.

Compile RTTOV

Follow rttov_install_notes to install RTTOV.

Setup Environment

mamba create -n rttov2 -c conda-forge hdf5 gfortran_linux-64 netcdf4 netcdf-fortran ipython matplotlib xarray h5py xesmf xcdat==0.4

Download Data

• Needed to setup copernicus for ERA5 downloads in "cds" conda environment (https://cds.climate.copernicus.eu/api-how-to)
• Downloaded the constants and orbit files
• In download/ folder (via compute that can download and via screen if needed): * conda activate cds * bash download_era5_wrapper.sh YYYY

Run RTTOV

• In rttov folder:
  • ./submit_monthly_jobs.sh YYYY

Process Data

• In process folder:
  • simulate_tmt.py [runs with 12 cores; use a full node]
  • sample_like_tropics.py [runs with 20 cores; use a full node; ~3 hour run time for 5 years / 6 satellites; informed by plot_satellites.py]
  • create_reference_diurnal_cycle.py
  • infer_diurnal_cycle.py [can use submit_python_job.py] (e.g., with the following configurations):
    • nadir_4sats_0.25K
    • 5fps_4sats_0.25K