Precip distribution

README.md

@@ -40,6 +40,7 @@ Some installation support for CMIP participating modeling groups is available: p
 
 [PMP versions](https://github.com/PCMDI/pcmdi_metrics/releases)
 ------------
+- [v2.5.0](https://github.com/PCMDI/pcmdi_metrics/releases/tag/v2.5.0) - New metric added: Precipitation Benchmarking -- distribution. Graphics updated
 - [v2.4.0](https://github.com/PCMDI/pcmdi_metrics/releases/tag/v2.4.0) - New metric added: AMO in variability modes
 - [v2.3.2](https://github.com/PCMDI/pcmdi_metrics/releases/tag/v2.3.2) - CMEC interface updates
 - [v2.3.1](https://github.com/PCMDI/pcmdi_metrics/releases/tag/v2.3.1) - Technical update

conda-env/dev.yml

@@ -19,9 +19,13 @@ dependencies:
   - eofs=1.4.0
   - seaborn=0.11.1
   - enso_metrics=1.1.1
+  - netcdf4=1.6.0
+  - regionmask=0.9.0
+  - rasterio=1.2.10
+  - shapely=1.8.0
   # Testing
   # ==================
-  - pre_commit=2.15.0
+  - pre_commit=2.20.0
   - pytest=6.2.5
   - pytest-cov=3.0.0
   # Developer Tools

pcmdi_metrics/precip_distribution/README.md

@@ -0,0 +1,28 @@
+# Precip distribution metrics
+
+Reference: Ahn, M.-S., P. A. Ullrich, P. J. Gleckler, J. Lee,  A. C. Ordonez, and A. G. Pendergrass, 2022: Evaluating Precipitation Distributions at Regional Scales: A Benchmarking Framework and Application to CMIP5 and CMIP6. Geoscientific Model Development (Submitted)
+
+## Driver code:
+- `precip_distribution_driver.py`
+
+## Parameter codes:
+- `param/`
+  - `precip_distribution_params_IMERG.py`
+  - `precip_distribution_params_TRMM.py`
+  - `precip_distribution_params_CMORPH.py`
+  - `precip_distribution_params_GPCP.py`
+  - `precip_distribution_params_PERSIANN.py`
+  - `precip_distribution_params_ERA5.py`
+  - `precip_distribution_params_cmip5.py`
+  - `precip_distribution_params_cmip6.py`
+
+## Run scripts:
+- `scripts_pcmdi/`
+  - `run_obs.bash`
+  - `run_parallel.wait.bash`
+
+## Note
+- Input data: daily averaged precipitation
+- This code should be run for a reference observation initially as some metrics (e.g., Perkins score) need a reference.
+- After completing calculation for a reference observation, this code can work for multiple datasets at once.
+- This benchmarking framework provides three tiers of area averaged outputs for i) large scale domain (Tropics and Extratropics with separated land and ocean) commonly used in the PMP , ii) large scale domain with clustered precipitation characteristics (Tropics and Extratropics with separated land and ocean, and separated heavy, moderate, and light precipitation regions), and iii) modified IPCC AR6 regions shown in the reference paper.

pcmdi_metrics/precip_distribution/lib/__init__.py

@@ -0,0 +1,19 @@
+from .argparse_functions import AddParserArgument  # noqa
+from .lib_precip_distribution import (  # noqa
+    CalcBinStructure,
+    CalcMetricsDomain,
+    CalcMetricsDomain3Clust,
+    CalcMetricsDomainAR6,
+    CalcP10P90,
+    CalcPscore,
+    CalcRainMetrics,
+    MakeDists,
+    MedDomain,
+    MedDomain3Clust,
+    MedDomainAR6,
+    Regrid,
+    getDailyCalendarMonth,
+    oneyear,
+    precip_distribution_cum,
+    precip_distribution_frq_amt,
+)

pcmdi_metrics/precip_distribution/lib/argparse_functions.py

@@ -0,0 +1,79 @@
+def AddParserArgument(P):
+    P.add_argument(
+        "--mip", type=str, dest="mip", default=None, help="cmip5, cmip6 or other mip"
+    )
+    P.add_argument(
+        "--exp", type=str, dest="exp", default=None, help="amip, cmip or others"
+    )
+    P.add_argument("--mod", type=str, dest="mod", default=None, help="model")
+    P.add_argument(
+        "--var", type=str, dest="var", default=None, help="pr or other variable"
+    )
+    P.add_argument(
+        "--frq", type=str, dest="frq", default=None, help="day, 3hr or other frequency"
+    )
+    P.add_argument(
+        "--modpath", type=str, dest="modpath", default=None, help="data directory path"
+    )
+    P.add_argument(
+        "--results_dir",
+        type=str,
+        dest="results_dir",
+        default=None,
+        help="results directory path",
+    )
+    P.add_argument(
+        "--case_id", type=str, dest="case_id", default=None, help="case_id with date"
+    )
+    P.add_argument(
+        "--prd",
+        type=int,
+        dest="prd",
+        nargs="+",
+        default=None,
+        help="start- and end-year for analysis (e.g., 1985 2004)",
+    )
+    P.add_argument(
+        "--fac",
+        type=str,
+        dest="fac",
+        default=None,
+        help="factor to make unit of [mm/day]",
+    )
+    P.add_argument(
+        "--res",
+        type=int,
+        dest="res",
+        nargs="+",
+        default=None,
+        help="list of target horizontal resolution [degree] for interporation (lon, lat)",
+    )
+    P.add_argument("--ref", type=str, dest="ref", default=None, help="reference data")
+    P.add_argument(
+        "--ref_dir",
+        type=str,
+        dest="ref_dir",
+        default=None,
+        help="reference directory path",
+    )
+    P.add_argument(
+        "--exp", type=str, dest="exp", default=None, help="e.g., historical or amip"
+    )
+    P.add_argument("--ver", type=str, dest="ver", default=None, help="version")
+    P.add_argument(
+        "--cmec",
+        dest="cmec",
+        default=False,
+        action="store_true",
+        help="Use to save CMEC format metrics JSON",
+    )
+    P.add_argument(
+        "--no_cmec",
+        dest="cmec",
+        default=False,
+        action="store_false",
+        help="Do not save CMEC format metrics JSON",
+    )
+    P.set_defaults(cmec=False)
+
+    return P