diff --git a/experimental/gen_data_citation.py b/experimental/gen_data_citation.py
new file mode 100644
index 0000000..67d2323
--- /dev/null
+++ b/experimental/gen_data_citation.py
@@ -0,0 +1,128 @@
+"""
+Charles Simpson 2020-10-13
+Script for generating data citations for CMIP6 data, for compliance with CMIP6 terms.
+"""
+import baspy as bp
+import pandas as pd
+import requests
+import xml.etree.ElementTree as ET
+
+
+def generate_citation(
+    MIP: str, Centre: str, Model: str, Experiment: str, Version: str,
+) -> dict:
+    """Generate a data citation from minimum necessary information.
+
+    The input is used to query CERA.
+    Metadata are downloaded as XML, and formatted as a citation string.
+    Input names follow baspy column names, as strings.
+    Output is a dictionary, including a citation string.
+
+    Example input:
+        MIP='CMIP',
+        Centre='MOHC',
+        Model='HadGEM3-GC31-LL',
+        Experiment='historical',
+        Version='v20190624
+    Example output:
+        {'MIP': 'CMIP', 'Centre': 'MOHC', 'Model': 'HadGEM3-GC31-LL', 'Experiment': 'historical', 'Version': 'v20190624', 'doi': '10.22033/ESGF/CMIP6.6109', 'publisher': 'Earth System Grid Federation', 'publicationYear': '2019', 'givenNames': ['Jeff', 'Matthew', 'Till', 'Martin', 'Tim'], 'familyNames': ['Ridley', 'Menary', 'Kuhlbrodt', 'Andrews', 'Andrews'], 'names': ['Ridley, J.', 'Menary, M.', 'Kuhlbrodt, T.', 'Andrews, M.', 'Andrews, T.'], 'citation': 'Ridley, J., Menary, M., Kuhlbrodt, T., Andrews, M., Andrews, T. (2019). MOHC HadGEM3-GC31-LL model output prepared for CMIP6 CMIP historical. v20190624. Earth System Grid Federation. https://doi.org/10.22033/ESGF/CMIP6.6109'}
+
+    """
+    url = (
+        "https://cera-www.dkrz.de/WDCC/ui/cerasearch/cerarest/"
+        + f"exportcmip6?input=CMIP6.{MIP}.{Centre}.{Model}.{Experiment}"
+        + "&wt=XML"
+    )
+    response = requests.get(url)
+    xml = response.text
+
+    root = ET.fromstring(xml)
+
+    root.tag
+    root.attrib
+
+    doi = ""
+    publisher = ""
+    publicationYear = ""
+    givenNames = []
+    familyNames = []
+
+    for child in root:
+        if "identifier" in child.tag and child.attrib["identifierType"] == "DOI":
+            doi = child.text
+        elif "creators" in child.tag:
+            for creator in list(child):
+                for x in list(creator):
+                    if "givenName" in x.tag:
+                        givenNames.append(x.text)
+                    elif "familyName" in x.tag:
+                        familyNames.append(x.text)
+        elif "publisher" in child.tag:
+            publisher = child.text
+        elif "publicationYear" in child.tag:
+            publicationYear = child.text
+
+    initials = [".".join([c for c in s if c.isupper()]) + "." for s in givenNames]
+    names = [", ".join([a, b]) for a, b in zip(familyNames, initials)]
+
+    citation = (
+        ", ".join(names)
+        + f" ({publicationYear}). {Centre} {Model} model output prepared for CMIP6 {MIP} {Experiment}. {Version}. {publisher}. https://doi.org/{doi}"
+    )
+
+    return {
+        "MIP": MIP,
+        "Centre": Centre,
+        "Model": Model,
+        "Experiment": Experiment,
+        "Version": Version,
+        "doi": doi,
+        "publisher": publisher,
+        "publicationYear": publicationYear,
+        "givenNames": givenNames,
+        "familyNames": familyNames,
+        "names": names,
+        "citation": citation,
+    }
+
+
+def df_to_citations(df: pd.DataFrame) -> pd.DataFrame:
+    """Generate data citations from a baspy dataframe.
+
+    Input: dataframe of model runs generated by baspy
+    Output: dataframe with citations generated.
+    """
+    results = []
+    for row in df.itertuples():
+        results.append(
+            generate_citation(
+                row.MIP, row.Centre, row.Model, row.Experiment, row.Version
+            )
+        )
+    return pd.DataFrame(results)
+
+
+if __name__ == "__main__":
+    experiments = ()
+
+    # Get some data with baspy
+    catlg = bp.catalogue(
+        dataset="cmip6",
+        Experiment=["historical", "ssp245",],
+        CMOR="day",
+        Var=["tas", "hurs", "tasmax", "tasmin"],
+        complete_var_set=True,
+    )
+    print("Demo generate_citation")
+    print(
+        generate_citation(
+            MIP="CMIP",
+            Centre="MOHC",
+            Model="HadGEM3-GC31-LL",
+            Experiment="historical",
+            Version="v20190624",
+        )
+    )
+
+    print("Demo generate_citations")
+    print(df_to_citations(catlg.sample(5).reset_index(drop=True))["citation"])