Updated convert script to get netcdf files from daily (_DD_) Ameriflu…

Ameriflux/Diurnal/convert_AMF_to_netcdf_daily.py

@@ -0,0 +1,122 @@
+#!/usr/bin/env python3
+"""
+Convert AmeriFlux daily CSV files (_DD_) to CF-compliant NetCDF files.
+
+- Only keeps variables that work with daily data:
+    GPP (GPP_DT_VUT_REF), PR, TAS
+- Radiation, RECO, and NEE variables are skipped
+- Output is stored in _output folder
+- Site names are extracted from the original CSV filenames
+"""
+
+import numpy as np
+import pandas as pd
+import xarray as xr
+from pathlib import Path
+
+# ----------------------------
+# Configuration
+# ----------------------------
+RAW_DIR = Path("_raw")
+OUTPUT_DIR = Path("_output")
+OUTPUT_DIR.mkdir(exist_ok=True)
+
+# ----------------------------
+# Variable mapping
+# (long_name, short_name, CSV column, units)
+# ----------------------------
+VARMAP = [
+    ("gross_primary_productivity", "gpp", "GPP_DT_VUT_REF", "gC m-2 d-1"),
+    # ("net_ecosystem_exchange", "nee", "NEE_VUT_REF", "gC m-2 d-1"),  # Not used
+    # ("ecosystem_respiration", "reco", "RECO_VUT_REF", "gC m-2 d-1"),  # Not used
+    ("precipitation", "pr", "P_F", "mm d-1"),
+    ("surface_air_temperature", "tas", "TA_F", "K"),
+    # Radiation variables skipped for daily processing
+    # ("surface_downward_shortwave_radiation", "rsds", "SW_IN_F", "W m-2"),
+    # ("surface_upward_shortwave_radiation", "rsus", "SW_OUT", "W m-2"),
+    # ("surface_downward_longwave_radiation", "rlds", "LW_IN_F", "W m-2"),
+    # ("surface_upward_longwave_radiation", "rlus", "LW_OUT", "W m-2"),
+]
+
+# ----------------------------
+# Read all daily CSVs
+# ----------------------------
+data_by_var = {v[1]: [] for v in VARMAP}
+site_names = []
+time_index = None
+
+print("Reading daily CSV files...")
+
+for csvfile in sorted(RAW_DIR.glob("*_DD*.csv")):
+    # Extract site name from filename
+    # If filename is ICs_2025_DD.csv → site = ICs
+    # If filename is AMF_ICs_DD_2025.csv → site = ICs
+    stem_parts = csvfile.stem.split("_")
+    if stem_parts[0] == "AMF":   # if there's an "AMF_" prefix
+        site = stem_parts[1]
+    else:
+        site = stem_parts[0]
+
+    print(f"  → {site}")
+
+    df = pd.read_csv(csvfile)
+
+    # ----------------------------
+    # Time parsing
+    # ----------------------------
+    if "TIMESTAMP" in df.columns:
+        df["time"] = pd.to_datetime(df["TIMESTAMP"], format="%Y%m%d")
+    elif "TIMESTAMP_START" in df.columns:
+        df["time"] = pd.to_datetime(df["TIMESTAMP_START"], format="%Y%m%d")
+    else:
+        raise ValueError(f"No recognizable timestamp column in {csvfile.name}")
+
+    df = df.set_index("time").sort_index()
+
+    if time_index is None:
+        time_index = df.index
+
+    # ----------------------------
+    # Extract variables
+    # ----------------------------
+    for long_name, short, amf_col, units in VARMAP:
+        if amf_col not in df.columns:
+            print(f"    ⚠️  Missing {amf_col} at {site}, filling with NaNs")
+            values = np.full(len(time_index), np.nan)
+        else:
+            values = df.reindex(time_index)[amf_col].to_numpy()
+        data_by_var[short].append(values)
+
+    site_names.append(site)
+
+# ----------------------------
+# Create NetCDF files by variable
+# ----------------------------
+print("\nCreating NetCDF files in _output/ ...")
+for long_name, short, amf_col, units in VARMAP:
+    data_array = np.vstack(data_by_var[short]).T  # shape: (time, site)
+
+    da = xr.DataArray(
+        data_array,
+        dims=("time", "site"),
+        coords={
+            "time": time_index,
+            "site": site_names,
+        },
+        name=short,
+        attrs={
+            "long_name": long_name,
+            "units": units,
+            "source": "AmeriFlux daily _DD_ CSV",
+            "product": "VUT_REF" if "VUT" in amf_col else "standard",
+        },
+    )
+
+    ds = da.to_dataset()
+
+    outfile = OUTPUT_DIR / f"{short}_daily.nc"
+    ds.to_netcdf(outfile)
+    print(f"  ✔ Saved {outfile}")
+
+print("\nAll done!")
+