This project was initiated as a part of the cloud direct insertion task in the PANDA-C project. It leverages the work of Jamie Bresch https://github.com/jamiebresch/obs2ioda to retrieve obs data. After reading the ABI L1b and L2 data, they are interpolated or super-obbed to the MPAS unstructured mesh. The processed data can be written either in the MPAS file format or in the IODA format (v1 for now). This repository can be also applicable for AHI (by Ivette Hernández Baños). Initial prototype code was designed and written by Yonggang G. Yu.
- Interpolate fields from obs (geostationary satellite radiance obs) to model grid
source mpas-jedi environment
git clone https://github.com/NCAR/obs2mpas.git
cd obs2mpas
mkdir build; cd build
cmake ../ ; make -j4
cd ../test/abi
ln ../../build/bin/obs2mpas.x .
./obs2mpas.x
NOTE: For NCAR HPC Derecho, the actual execution of this program should be done in the computational node.
main.F90
- 0. get argument from command line
- 1. read namelist
- 2. read observations (ABI or AHI) lat/lon & data
- 3. read MPAS lat/lon
- 4. build and search kd-tree
- 5. re-organize the matching pairs
- 6. interpolate the obs fields into model mesh either superob or nearest neighbor.
- 7a. Write the interpolated fields to MPAS file
- 7b. Write the interpolated fields to IODA file
&main_nml
f_mpas_latlon = '' , ! MPAS file path/name to read lat & lon information
f_mpas_out = '' , ! MPAS file for writing the interpolated ABI fields
l_read_indx = .true. or .false., ! read index and counnt for matching ABI-MPAS pairs
l_write_indx = .true. or .false., ! write index and counnt for matching ABI-MPAS pairs
l_superob = .true., ! .true.= mesh-based superob, .false.= nearest-neighbor
l_write_o2m_iodav1 = .true., ! .true. = write superob/nearest-neighbor into ioda v1 file
! .false.= write superob/nearest-neighbor into MPAS file
&data_nml
This section is the same as https://github.com/jamiebresch/obs2ioda