Adding recipe for GODAS

feedstock/godas.py

@@ -0,0 +1,50 @@
+"""
+NCEP Global Ocean Data Assimilation System (GODAS)
+"""
+import apache_beam as beam
+from pangeo_forge_recipes.patterns import MergeDim, ConcatDim, FilePattern
+from pangeo_forge_recipes.transforms import OpenURLWithFSSpec, OpenWithXarray, StoreToZarr, Indexed, T
+variables = ['sshg', 'thflx'] 
+years = [1980, 1981, 1982]
+
+def make_full_path(variable, time):
+    return f"https://downloads.psl.noaa.gov/Datasets/godas/{variable}.{time}.nc"
+
+variable_merge_dim = MergeDim("variable", variables)
+time_concat_dim = ConcatDim("time", years)
+
+## preprocessing transform
+
+class Preprocess(beam.PTransform):
+    """
+    Set variables to be coordinates
+    """
+
+    @staticmethod
+    def _set_bnds_as_coords(item: Indexed[T]) -> Indexed[T]:
+        """
+        The netcdf lists some of the coordinate variables as data variables. 
+        This is a fix which we want to apply to each dataset.
+        """
+        index, ds = item
+        new_coords_vars = ['date', 'timePlot']
+        ds = ds.set_coords(new_coords_vars)
+        return index, ds
+
+    def expand(self, pcoll: beam.PCollection) -> beam.PCollection:
+        return pcoll | beam.Map(self._set_bnds_as_coords)
+
+
+pattern = FilePattern(make_full_path, variable_merge_dim, time_concat_dim, file_type="netcdf4")
+
+GODAS = (
+    beam.Create(pattern.items())
+    | OpenURLWithFSSpec()
+    | OpenWithXarray(file_type=pattern.file_type)
+    | Preprocess() # New preprocessor
+    | StoreToZarr(
+        target_chunks={'time':120},
+        store_name="GODAS.zarr",
+        combine_dims=pattern.combine_dim_keys,
+    )
+)

feedstock/meta.yaml

@@ -34,6 +34,8 @@ recipes:
             - producer
             - licensor
       license: "CC-BY-4.0"
+  - id: "GODAS"
+    object: "godas:GODAS"
   - id: CASM
     object: "casm:CASM"
     description: A long-term Consistent Artificial-intelligence based Soil Moisture dataset based on machine learning and remote sensing

feedstock/requirements.txt

@@ -1 +1,3 @@
 git+https://github.com/pangeo-forge/pangeo-forge-recipes.git@beam-refactor
+fsspec==2023.5.0
+gcsfs==2023.5.0