Add area_method:'latlon' to example configs, fix precompute_grid_area for radar coords

feng045 · feng045 · commit ad8d3210b403 · 2026-03-13T08:57:41.000-07:00
- Add area_method: 'latlon' to all 17 example config YAML files
- Fix precompute_grid_area() to use lat_coordname/lon_coordname from config
  when available (e.g., radar data where x/y coords are Cartesian meters)
- Use xarray dimension-aware indexing (isel/transpose) instead of hardcoded
  numpy positional slicing to handle 3D lat/lon arrays safely
- Update tests/README.md with test_area_method_utils and test_area_method_clouds
  descriptions
diff --git a/config/config_csapr500m_example.yml b/config/config_csapr500m_example.yml
@@ -133,6 +133,7 @@ z_coord_type: height
 # Specify data structure
 datatimeresolution: 0.25     # hours
 pixel_radius: 0.5      # km
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 ref_varname: dbz_comp
 
 # Tracking parameters
diff --git a/config/config_era5_z500_example.yml b/config/config_era5_z500_example.yml
@@ -44,6 +44,7 @@ feature_type: 'generic'
 # Specify data structure
 datatimeresolution: 24.0     # hours
 pixel_radius: 200.0      # km
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 time_dimname: 'time'
 x_dimname: 'lon'
 y_dimname: 'lat'
diff --git a/config/config_gridrad_mcs_example.yml b/config/config_gridrad_mcs_example.yml
@@ -52,6 +52,7 @@ landmask_y_coordname: 'lat'
 
 # Input dataset structure
 pixel_radius:  4.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'Tb'
diff --git a/config/config_gridrad_mcs_template.yml b/config/config_gridrad_mcs_template.yml
@@ -54,6 +54,7 @@ landfrac_thresh: [0.9, 1]  # Define the range of fraction for land (depends on w
 
 # Input dataset structure
 pixel_radius:  4.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'Tb'
diff --git a/config/config_himawari_mcs_example.yml b/config/config_himawari_mcs_example.yml
@@ -48,6 +48,7 @@ landmask_y_coordname: 'lat'
 
 # Specific to GPM Tb+IMERG combined dataset
 pixel_radius:  5.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'tbb_11'
diff --git a/config/config_imerg25km_mcs_tbpf.yml b/config/config_imerg25km_mcs_tbpf.yml
@@ -53,6 +53,7 @@ landfrac_thresh: [0.9, 1]  # Define the range of fraction for land (depends on w
 
 # Input dataset structure
 pixel_radius: 25.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 olr2tb: False
diff --git a/config/config_imerg_mcs_tbpf_example.yml b/config/config_imerg_mcs_tbpf_example.yml
@@ -52,6 +52,7 @@ landfrac_thresh: [0, 90]  # Define the range of fraction for land (depends on wh
 
 # Specific to GPM Tb+IMERG combined dataset
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'Tb'
diff --git a/config/config_mcs_pbc_idealized.yml b/config/config_mcs_pbc_idealized.yml
@@ -49,6 +49,7 @@ landmask_varname: 'landseamask'
 
 # Specific to GPM Tb+IMERG combined dataset
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname: 'Tb'
diff --git a/config/config_mcs_saag_example.yml b/config/config_mcs_saag_example.yml
@@ -53,6 +53,7 @@ landfrac_thresh: [0, 90]  # Define the range of fraction for land (depends on wh
 
 # Input dataset structure
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'tb'
diff --git a/config/config_mcs_tbpf_imergv7_mcsmip.yml b/config/config_mcs_tbpf_imergv7_mcsmip.yml
@@ -66,6 +66,7 @@ landfrac_thresh: [0, 90]  # Define the range of fraction for land (depends on wh
 
 # Input dataset structure
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname: Tb
diff --git a/config/config_mcs_tbpf_scream_healpix7.yml b/config/config_mcs_tbpf_scream_healpix7.yml
@@ -66,6 +66,7 @@ landfrac_thresh: [0, 90]  # Define the range of fraction for land (depends on wh
 
 # Input dataset structure
 pixel_radius:  50.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 olr2tb: True
diff --git a/config/config_mcs_tbpf_scream_healpix9.yml b/config/config_mcs_tbpf_scream_healpix9.yml
@@ -66,6 +66,7 @@ landfrac_thresh: [0, 90]  # Define the range of fraction for land (depends on wh
 
 # Input dataset structure
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 olr2tb: True
diff --git a/config/config_model25km_mcs_tbpf_example.yml b/config/config_model25km_mcs_tbpf_example.yml
@@ -46,6 +46,7 @@ landmask_varname: 'LANDMASK'
 
 # Input dataset structure
 pixel_radius:  25.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input dat
 olr2tb: True
diff --git a/config/config_mrms_mcs_composite_reflectivity.yml b/config/config_mrms_mcs_composite_reflectivity.yml
@@ -44,6 +44,7 @@ feature_type: 'generic'
 # Specify data structure
 datatimeresolution: 0.166667     # hours
 pixel_radius: 1.0      # km
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 time_dimname: 'time'
 x_dimname: 'longitude'
 y_dimname: 'latitude'
diff --git a/config/config_nexrad500m_example.yml b/config/config_nexrad500m_example.yml
@@ -133,6 +133,7 @@ z_coord_type: height
 # Specify data structure
 datatimeresolution: 0.0833     # hours
 pixel_radius: 0.5      # km
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 ref_varname: dbz_comp
 
 # Tracking parameters
diff --git a/config/config_sipam4km_example.yml b/config/config_sipam4km_example.yml
@@ -166,6 +166,7 @@ z_coord_type: height
 # Specify data structure
 datatimeresolution: 0.2     # hours
 pixel_radius: 4.0      # km
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 ref_varname: dbz_comp
 
 # Tracking parameters
diff --git a/config/config_wrfregrid_mcs_tbpf_example.yml b/config/config_wrfregrid_mcs_tbpf_example.yml
@@ -77,6 +77,7 @@ write_native: False
 
 # Input dataset structure
 pixel_radius:  10.0  # [km] Spatial resolution of the input data
+area_method: 'latlon'  # 'fixed' (default) or 'latlon' for latitude-dependent grid cell area
 datatimeresolution: 1.0  # [hour] Temporal resolution of the input data
 # Variable names in the input data
 tb_varname:  'tb'
diff --git a/pyflextrkr/ft_utilities.py b/pyflextrkr/ft_utilities.py
@@ -1119,6 +1119,10 @@ def precompute_grid_area(config, first_file=None):
     input_format = config.get("input_format", "netcdf")
     x_coordname = config.get("x_coordname", "lon")
     y_coordname = config.get("y_coordname", "lat")
+    # For radar data, lat/lon may be in separate variables (e.g., point_latitude)
+    # If lat_coordname/lon_coordname are defined, use those instead
+    lat_coordname = config.get("lat_coordname", None)
+    lon_coordname = config.get("lon_coordname", None)
     geolimits = config.get("geolimits", None)
 
     # Read lat/lon coordinates
@@ -1145,8 +1149,27 @@ def precompute_grid_area(config, first_file=None):
                 "to read lat/lon for grid area computation."
             )
         ds = xr.open_dataset(first_file, decode_timedelta=False)
-        lat = ds[y_coordname].values
-        lon = ds[x_coordname].values
+        # Use lat_coordname/lon_coordname if available (e.g., radar data
+        # where x_coordname/y_coordname are Cartesian meters)
+        _lat_name = lat_coordname if lat_coordname else y_coordname
+        _lon_name = lon_coordname if lon_coordname else x_coordname
+        da_lat = ds[_lat_name]
+        da_lon = ds[_lon_name]
+
+        # If lat/lon are 3D (e.g., radar point_latitude(z, y, x)),
+        # select the first level using dimension name from config
+        z_dimname = config.get("z_dimname", None)
+        y_dimname = config.get("y_dimname", "lat")
+        x_dimname = config.get("x_dimname", "lon")
+        if da_lat.ndim == 3 and z_dimname and z_dimname in da_lat.dims:
+            da_lat = da_lat.isel({z_dimname: 0})
+            da_lon = da_lon.isel({z_dimname: 0})
+        # Ensure (y, x) dimension order and convert to numpy
+        if da_lat.ndim == 2:
+            da_lat = da_lat.transpose(y_dimname, x_dimname)
+            da_lon = da_lon.transpose(y_dimname, x_dimname)
+        lat = da_lat.values
+        lon = da_lon.values
         ds.close()
 
     # Mesh 1D to 2D if needed
diff --git a/tests/README.md b/tests/README.md
@@ -5,7 +5,7 @@ Tests are written with [pytest](https://docs.pytest.org/) and are split into thr
 
 | Tier | Files | Runs on | Needs data? |
 |------|-------|---------|-------------|
-| Unit / synthetic | `test_steiner_func.py`, `test_echotop_func.py`, `test_vertical_coordinate.py`, `test_idcells_synthetic.py`, `test_example.py` | GitHub CI + local | No |
+| Unit / synthetic | `test_steiner_func.py`, `test_echotop_func.py`, `test_vertical_coordinate.py`, `test_idcells_synthetic.py`, `test_area_method_utils.py`, `test_area_method_clouds.py`, `test_example.py` | GitHub CI + local | No |
 | Local regression | `test_regression_local.py` | Local / HPC only | Yes (demo outputs) |
 | End-to-end demos | `run_demo_tests.py` | Local / HPC only | Auto-downloads |
 
@@ -110,6 +110,20 @@ temporary directory, then runs the full cell-identification pipeline:
 Verifies that the known convective cell in the synthetic data is detected correctly.
 No real data needed.
 
+### `test_area_method_utils.py`
+Unit tests for the `area_method: 'latlon'` utility functions: `compute_grid_area`,
+`save_grid_area`, `load_grid_area`, `get_pixel_area`, and `precompute_grid_area`.
+Also tests `sort_renumber2vars` with a 2D `grid_area` array to verify that
+feature areas are computed correctly with variable-size pixels.
+Uses synthetic lat/lon grids and temporary files — no real data needed.
+
+### `test_area_method_clouds.py`
+Unit tests for `label_and_grow_cold_clouds` with 2D `pixel_area` (grid area)
+support. Verifies that cloud labelling, area thresholding, and feature
+renumbering work correctly when pixel areas vary across the domain (as they
+do for latitude-dependent grids). Tests both the fixed (scalar) and
+latlon (2D array) modes to ensure backward compatibility.
+
 ### `test_regression_local.py`
 Validates the output of full demo runs (e.g., `demo_cell_nexrad.sh`).
 These tests check that track statistics files contain reasonable values
