diff --git a/content/gsi/05-tutorial.qmd b/content/gsi/05-tutorial.qmd index 3e565f1..d28eb90 100644 --- a/content/gsi/05-tutorial.qmd +++ b/content/gsi/05-tutorial.qmd @@ -2,8 +2,239 @@ title: "Tutoriales" --- -Ejemplo asimilando todo +This tutorial is intended to showcase some of the capabilities of the GSI system. For a more complete set of examples, please visit the official website for GSI: [variational methods](https://dtcenter.org/community-code/gridpoint-statistical-interpolation-gsi) and \[Kalman Filter\]() methods. -gsi +The tutorial uses the serial Ensemble Square Root Filter (EnSRF) algorithm but could also be used to run the Local Ensemble Kalman Filter (LETKF) algorithm if the GSI system code is compiled using a intel compiler[^1]. It also uses a version of the code that have the capability of assimilating GOES-16 observations. -letkf +[^1]: The current version of the GSI system compiled using gnu returns an error when the LETKF algorithm is used. + +## Moving parts + +The first component to run the tutorial is the GSI system code that can be found in this repository: [github.com/paocorrales/comGSIv3.7_EnKFv1.3](https://github.com/paocorrales/comGSIv3.7_EnKFv1.3). Alternatively, the original code without modifications is available [here](https://dtcenter.org/community-code/gridpoint-statistical-interpolation-gsi). The repository includes an example script `compile_gsi_yakaira` to compile the code. In any case it is important to include the option `-DBUILD_WRF=ON` to use the ENKF algorithm with WRF. + +Then, clone or download the repository with the scripts, namelists and specific config files from: . This repository also includes a bash scripts to download the necessary data to run the tutorial. This data is also available as a [Zenodo record](https://zenodo.org/records/10439645). + +### Observations + +The observations included in this tutorial are: + +**Conventional observations:** surface and upper air observations (meteorological stations, radiosondes, airplanes, etc.). It also includes wind derived from satellite. The bufr file used here is created from the prepbufr file at 12 UTC of 2018/11/22 and includes also observations from automatic station networks in Argentina and other countries. + +- File: `cimap.20181122.t12z.01h.prepbufr.nqc` + +**Radiance observations:** radiances from polar and geostationary satellites. Observations from polar satellites comes from the Global Data Assimilation System (GDAS) Model: . The GOES-16 observations (ABI sensor) are derived from the public netcdf files published by NOAA. + +Files: + +- `abig16.20181122.t12z.bufr_d` +- `1bamua.20181122.t12z.bufr_d` +- `ssmisu.20181122.t12z.bufr_d` +- `1bhrs4.20181122.t12z.bufr_d` +- `airsev.20181122.t12z.bufr_d` +- `mtiasi.20181122.t12z.bufr_d` +- `1bmhs.20181122.t12z.bufr_d` +- `atms.20181122.t12z.bufr_d` +- `satwnd.20181122.t12z.bufr_d` + +### Background + +The background files includes the 10-member ensemble generated using the WRF-ARW numerical model for a regional domain centered in the center and northern Argentina. For more information about the model configuration see + +- The 00 subfolder includes a 10-member ensemble and the ensemble mean to run the GSI system using the ENKF version. + +- The 01 to 10 subfolders include the background at the analysis time and files every 10 minutes inside the assimilation window to run the GSI system using the [FGAT method](#fgat) + +## Structure + +By cloning the tutorial repo and downloading the associated data with the provied script you will end up with the following folder structure. + +```{bash} +tutorial_gsi/ +├── download_data.sh +├── fix +│   ├── global_satinfo.txt +│   ├── satbias_ang +│   ├── satbias_in +│   └── satbias_pc_in +├── GUESS +│   └── 20181122120000 +│   ├── 00 +│   │   ├── wrfarw.ensmean +│   │   ├── wrfarw.mem001 +│   │   ├── wrfarw.mem002 +│   │   ├── wrfarw.mem003 +│   │   ├── wrfarw.mem004 +│   │   ├── wrfarw.mem005 +│   │   ├── wrfarw.mem006 +│   │   ├── wrfarw.mem007 +│   │   ├── wrfarw.mem008 +│   │   ├── wrfarw.mem009 +│   │   └── wrfarw.mem010 +│   ├── 01 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 02 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 03 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 04 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 05 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 06 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 07 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 08 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   ├── 09 +│   │   ├── wrf_inou1 +│   │   ├── wrf_inou2 +│   │   ├── wrf_inou3 +│   │   ├── wrf_inou4 +│   │   ├── wrf_inou5 +│   │   ├── wrf_inou6 +│   │   └── wrf_inou7 +│   └── 10 +│   ├── wrf_inou1 +│   ├── wrf_inou2 +│   ├── wrf_inou3 +│   ├── wrf_inou4 +│   ├── wrf_inou5 +│   ├── wrf_inou6 +│   └── wrf_inou7 +├── namelists +│   ├── comenkf_namelist.sh +│   └── comgsi_namelist.sh +├── OBS +│   ├── 1bamua.20181122.t12z.bufr_d +│   ├── 1bhrs4.20181122.t12z.bufr_d +│   ├── 1bmhs.20181122.t12z.bufr_d +│   ├── abig16.20181122.t12z.bufr_d +│   ├── airsev.20181122.t12z.bufr_d +│   ├── atms.20181122.t12z.bufr_d +│   ├── cimap.20181122.t12z.01h.prepbufr.nqc +│   ├── mtiasi.20181122.t12z.bufr_d +│   ├── satwnd.20181122.t12z.bufr_d +│   └── ssmisu.20181122.t12z.bufr_d +├── README.md +├── run_enkf.sh +└── run_gsi.sh +``` + +A `GSI` folder will be created when running the `run_gsi.sh` script and a `ENKF` folder will be created when running the `run_enkf.sh` that performs the analysis. + +## Running GSI {#fgat} + +As we focus on running GSI with the Kalman Filter method, the first stem is to run GSI as Observation operator. So, the system will compare the observations with the background state and and save that information in diagnostic files. + +The example used in this tutorial is relatively small, so while you may need a HPC system for real cases, this one can be run in a small server or even a computer with at least 10 processors. + +Here are the \~20 first lines of the script: + +```{bash} +#PBS -N TEST-1-GSI +#PBS -m abe +#PBS -l walltime=03:00:00 +#PBS -l nodes=1:ppn=24 +#PBS -j oe + +BASEDIR=/home/paola.corrales/datosmunin3/tutorial_gsi # Path to the tutorial folder +GSIDIR=/home/paola.corrales/datosmunin3/comGSIv3.7_EnKFv1.3 # Path to where the GSI/EnKF code is compiled +FECHA_INI='11:00:00 2018-11-22' # Init time (analysis time - $ANALISIS) +ANALISIS=3600 # Assimilation cycle in seconds +OBSWIN=1 # Assimilation window in hours +N_MEMBERS=10 # Ensemble size +E_WE=200 # West-East grid points +E_SN=240 # South-North grid points +E_BT=37 # Vertical levels + + +export OMP_NUM_THREADS=1 +GSIPROC=10 + +set-x +``` + +In principle, you only need to change `BASEDIR` and `GSIDIR` variables that are the path to the tutorial folder and the path to where GSI is compiled (the code expects to find a `build` folder with the executable files. + +So, with that, you can run the script or send it to a queue. + +### Possible issues {#issues-gsi} + +The script assume many things, in particular, where the configuration files, observations and background files are located. If you chance the structure of the folders and files, make sure to do the same in the script. + +The other very possible issue is machine dependent. GSI creates files with the information of the observations and backgroud called `pe\*something`. Those files are later concatenated in `diag_*` files. This process depends on listing all the types of observations with some regex magic: + +```{bash} +listall=`ls pe* | cut -f2 -d"." | awk '{print substr($0, 0, length($0)-2)}' | sort | uniq ` + + for type in $listall; do + count=`ls pe*${type}_${loop}* | wc -l` + if [[ $count -gt 0 ]]; then + cat pe*${type}_${loop}* > diag_${type}_${string}.${ANAL_TIME} # For binary diag files + fi + done +``` + +I had to slightly change that first line every time I changed machines. So, if you dont see a bunch of `diag*` files in the `GSI` folder after running the script this is probably the reason. + +### Did it work? + +If you get a `exit 0` at the end, it probably means that everything went well. However, I recommend you check a few things to make sure everything went *really* well. + +- Check that all the `diag*` files are there. You will get 1 file per member and ensemble mean for each type of observation. If you don't see any of these files, check the [Issues](@issues-gsi) section. If you are missing the files for 1 type of observation, that probably means that the bufr file with the observations was not read properly or that is missing in the folder. Check if the script is linking the correct file to the `GSI` folder. +- Check the statistics for each type of observations. You will find this information in the `fit_.` files. Each one may have different information or structure depending on the type of observation, but make sure to check the number of observations read and keep by the system. This information is also included in the `stdout` file, you can sear `READ_*` to find the section in the file. + +If you got an error number instead and, if you are lucky, the error code may be described in `gsimain.f90`. + +## Running ENKF + +## Running GSI using the FGAT method