Merge branch 'master' of github.com:djgagne/hagelslag into djgagne

.github/workflows/python-package-conda.yml

@@ -10,14 +10,7 @@ jobs:
 
     steps:
     - uses: actions/checkout@v2
-    - name: Install grib dependencies
-      shell: bash -l {0}
-      run: |
-        sudo add-apt-repository "deb http://security.ubuntu.com/ubuntu xenial-security main"
-        sudo apt update
-        sudo apt install libjasper1 libjasper-dev libopenjp2-7 libopenjp2-tools libopenjp2-7-dev
-        ldconfig -p | grep libjasper
-    - name: "Set up Python 3.8"
+    - name: "Set up Python 3"
       uses: conda-incubator/setup-miniconda@v2
       with:
         mamba-version: "*"

config/MPAS_interpolated_HRRR.config

@@ -0,0 +1,70 @@
+import numpy as np
+import pandas as pd
+from hagelslag.processing.ObjectMatcher import shifted_centroid_distance
+from hagelslag.processing.ObjectMatcher import centroid_distance, time_distance
+
+# NOTE: HRRR variables must be listed in the following format:
+# {HRRR_VARIABLE_NAME}-{HRRR_level}
+# For example, Composite Reflectivity (REFC) which is at the (entire_atmosphere) level
+# would be listed as 'REFC-entire_atmosphere'
+
+# 'ensemble_name' must be lsited as 'HRRR-ZARR'
+# 'model_path' must be "hrrrzarr/sfc/"
+# 'end_hour' must be no more than n-1 number of forecast hours in model
+
+# Support for model runs with different forecast lengths, use 'HRRR_alt_end_hour' for an alternative
+# forecast length that is used for each model run hour listed in 'HRRR_alt_hours'. If neither is provided,
+# 'end_hour' will be used for all model runs.
+
+## output path
+scratch_path = "/glade/scratch/cbecker/MPAS_interp_HRRR/"
+
+# Historical runs
+date_index = pd.date_range(start='2023-04-24', end='2023-05-31', freq='1D', tz='UTC').to_pydatetime()
+
+# Real Time runs
+# Use pd.Timedelta to correspond with delay in data availability from hour that script is submitted
+# date_index = pd.DatetimeIndex([pd.Timestamp.utcnow().strftime("%Y-%m-%d-%H")]) - pd.Timedelta(hours=3)
+
+ensemble_members = ["1", "2", "3", "4", "5"]
+
+config = dict(dates=date_index,
+              start_hour=1,
+              end_hour=132,
+              HRRR_alt_end_hour=47,
+              HRRR_alt_run_hours=[0, 6, 12, 18],
+              watershed_variable="refl10cm_max",
+              ensemble_name="MPAS-interp",
+              ensemble_members=ensemble_members,
+              model_path="/glade/campaign/mmm/parc/schwartz/HWT2023/mpas/",
+              segmentation_approach="hyst",
+              model_watershed_params=(35, 50),
+              size_filter=12,
+              gaussian_window=1,
+              mrms_path=None,
+              mrms_variable="MESH_Max_60min_00.50",
+              mrms_watershed_params=(13, 1, 125, 100, 100),
+              object_matcher_params=([shifted_centroid_distance], np.array([1.0]),
+                                     np.array([24000])),
+              track_matcher_params=([centroid_distance, time_distance],
+                                     np.array([80000, 2])),
+              storm_variables=["refl10cm_max", "updraft_helicity_max", "updraft_helicity_max01",
+                               "updraft_helicity_max03"],
+              potential_variables=[],
+              tendency_variables=[],
+              shape_variables=["area", "eccentricity", "major_axis_length", "minor_axis_length", "orientation"],
+              variable_statistics=["mean", "max", "min"],
+              csv_path=scratch_path + "track_data_MPAS_interp_hrrr_3km_csv_refl/",
+              geojson_path=scratch_path + "track_data_MPAS_interp_hrrr_3km_json_refl/",
+              nc_path=scratch_path + "track_data_MPAS_interp_hrrr_3km_nc_refl/",
+              patch_radius=48,
+              unique_matches=True,
+              closest_matches=True,
+              match_steps=True,
+              train=False,
+              single_step=True,
+              label_type="gamma",
+              segment_full_object=True,
+              model_map_file="/glade/work/cbecker/hagelslag/mapfiles/mpas_interp_hrrr_map_2023.txt",
+              mask_file=None,
+              output_geojson_grid=True)

hagelslag/data/MPASHrrrGrid.py

@@ -0,0 +1,33 @@
+from .ModelGrid import ModelGrid
+import numpy as np
+from os.path import join
+class MPASHrrrGrid(ModelGrid):
+    """
+    Loads model output from the NCAR MMM 1 and 3 km WRF runs on Cheyenne.
+
+    """
+
+    def __init__(self, run_date, variable, start_date, end_date, member, path):
+        self.path = path
+        self.forecast_hours = np.arange((start_date - run_date).total_seconds() / 3600,
+                                        (end_date - run_date).total_seconds() / 3600 + 1, dtype=int)
+        filenames = []
+        for hour in self.forecast_hours:
+            filename = join(path, run_date.strftime("%Y%m%d%H"), "post", f"mem_{member}",
+                            f"interp_mpas_3km_{run_date.strftime('%Y%m%d%H')}_mem{member}_f{hour:03}.nc")
+            filenames.append(filename)
+        super(MPASHrrrGrid, self).__init__(filenames, run_date, start_date, end_date, variable)
+
+    def format_var_name(self, variable, var_list):
+
+        z_index = None
+        if variable in var_list:
+            var_name = variable
+        elif "_PL" in variable:
+            var_parts = variable.split("_")
+            var_name = "_".join(var_parts[:-1])
+            p_level = int(var_parts[-1])
+            z_index = np.where(self.pressure_levels == p_level)[0][0]
+        else:
+            raise KeyError("{0} not found in {1}".format(variable, ", ".join(var_list)))
+        return var_name, z_index

hagelslag/data/ModelOutput.py

@@ -15,6 +15,7 @@
 from .NCARStormEventModelGrid import NCARStormEventModelGrid
 from .SSEFModelGrid import SSEFModelGrid
 from .VSEModelGrid import VSEModelGrid
+from .MPASHrrrGrid import MPASHrrrGrid
 
 
 class ModelOutput(object):
@@ -194,6 +195,15 @@ def load_data(self):
                                    self.end_date,
                                    self.path)
             self.data, self.units = mg.load_data()
+
+        elif self.ensemble_name.upper() == "MPAS-INTERP":
+            mg = MPASHrrrGrid(self.run_date,
+                              self.variable,
+                              self.start_date,
+                              self.end_date,
+                              self.member_name,
+                              self.path)
+            self.data, self.units = mg.load_data()
         else:
             print(self.ensemble_name + " not supported.")
 

mapfiles/mpas_interp_hrrr_map_2023.txt

@@ -0,0 +1,14 @@
+proj=lcc
+a=6371229.0
+b=6371229.0
+lat_0=38.5
+lon_0=262.5
+lat_1=38.5
+lat_2=38.5
+units=m
+dx=3000
+dy=3000
+sw_lon=-122.7195
+sw_lat=21.13812
+ne_lon=-60.91719
+ne_lat=47.84219