Omnibus

htdocs/plotting/auto/scripts/__init__.py

@@ -734,6 +734,10 @@
             "Top 10 Daily"
         ),
     },
+    {
+        "id": 247,
+        "label": "NWS Watch Warning Advisory (WaWa) Map + Population Stats",
+    },
     {
         "id": 70,
         "label": "Period between First and Last VTEC Product Each Year",

htdocs/plotting/auto/scripts200/p224.py

@@ -27,16 +27,20 @@
 that have been created, but not necessarily having an issuance time prior
 to the given timestamp.  This is why we can't have nice things!
 </p>
+
+<p><a href="/plotting/auto/?q=247">Autoplot 247</a> is closely related to
+this app and provides a map of WaWa + stats.</p>
 """
 from datetime import timezone
 from zoneinfo import ZoneInfo
 
 import pandas as pd
 from matplotlib.patches import Rectangle
+from pyiem.database import get_sqlalchemy_conn
 from pyiem.exceptions import NoDataFound
 from pyiem.nws.vtec import NWS_COLORS, get_ps_string
 from pyiem.plot import figure
-from pyiem.util import get_autoplot_context, get_sqlalchemy_conn, utc
+from pyiem.util import get_autoplot_context, utc
 from sqlalchemy import text
 
 PDICT = {

htdocs/plotting/auto/scripts200/p247.py

@@ -0,0 +1,304 @@
+"""
+This autoplot generates a map of active NWS Watch, Warnings, and Advisories
+active at the given timestamp.
+
+<p><strong>Footnote on "Active" WaWA</strong>: A more complex than it
+should be nuance to explain here is the concept of what is "active" at
+a given timestamp.  Let us consider a real world example.  On Monday
+afternoon, the NWS issues a Winter Storm Warning for an upcoming
+storm that goes "into effect" at noon on Tuesday.  You request this
+plot for a timestamp of 6 PM on that Monday.  Is the Winter Storm
+Warning included in this metric at that time?
+<ul>
+<li>Yes, if you select the option to include any WaWA that have been
+created, but may have an VTEC start time in the future yet.</li>
+<li>No, if you select the option to only include WaWA that have an
+issuance time before the given timestamp.</li>
+</ul>
+<br />The default setting here is the first option, to include any events
+that have been created, but not necessarily having an issuance time prior
+to the given timestamp.  This is why we can't have nice things!</p>
+
+<p>If you turn on statistics, the legend displays active events nationwide.</p>
+
+<p><a href="/plotting/auto/?q=224">Autoplot 224</a> is closely related to
+this app and provides a table of stats.</p>
+"""
+from datetime import timezone
+from io import StringIO
+
+import geopandas as gpd
+import httpx
+import pandas as pd
+from matplotlib.patches import Rectangle
+from pyiem.database import get_sqlalchemy_conn
+from pyiem.nws.vtec import NWS_COLORS, get_ps_string
+from pyiem.plot import MapPlot
+from pyiem.reference import SECTORS_NAME, Z_OVERLAY2, state_names
+from pyiem.util import get_autoplot_context, utc
+from sqlalchemy import text
+
+# shared with 224
+PDICT = {
+    "active": "Include WaWA that have been created or valid at the given time",
+    "within": "Include WaWA with VTEC issuance time before the given time",
+}
+PDICT2 = {
+    "none": "None",
+    "stats": "Include Population and Area Statistics",
+}
+
+
+def get_description():
+    """Return a dict describing how to call this plotter"""
+    desc = {"description": __doc__}
+    desc["defaults"] = {"_r": None}  # disables
+    sectors = SECTORS_NAME.copy()
+    sectors["nws"] = "NWS All"
+    sectors.update(state_names)
+    desc["arguments"] = [
+        {
+            "type": "datetime",
+            "name": "valid",
+            "default": utc().strftime("%Y/%m/%d %H00"),
+            "label": "Valid Timestamp (UTC):",
+            "min": "1986/01/01 0000",
+        },
+        {
+            "type": "select",
+            "options": PDICT,
+            "default": "active",
+            "name": "opt",
+            "label": "How to consider if an event is active [see footnote]",
+        },
+        {
+            "type": "select",
+            "options": sectors,
+            "name": "csector",
+            "default": "nws",
+            "label": "Select state/sector",
+        },
+        {
+            "type": "select",
+            "options": PDICT2,
+            "name": "add",
+            "default": "none",
+            "label": "Include Statistics in Legend (slows down plot)",
+        },
+    ]
+    return desc
+
+
+def plotsbw(mp, df):
+    """Do sbw plotting."""
+    df["color"] = df["ps"].apply(lambda x: NWS_COLORS.get(x, "k"))
+    for panel in mp.panels:
+        df2 = df.to_crs(panel.crs).cx[
+            slice(*panel.get_xlim()), slice(*panel.get_ylim())
+        ]
+
+        if df2.empty:
+            continue
+        df2.plot(
+            ax=panel.ax,
+            aspect=None,
+            edgecolor=df2["color"],
+            facecolor="None",
+            lw=2,
+            zorder=Z_OVERLAY2 + 3,
+        )
+
+
+def get_ps(pstr):
+    """."""
+    return (
+        get_ps_string(*pstr.split("."))
+        .replace("Advisory", "Adv")
+        .replace("Small Craft for", "SC")
+    )
+
+
+def make_legend(mp, df, statdf, year):
+    """Do legend work."""
+    if "zone_pop" not in statdf.columns:
+        # Build out an estimate of all combos
+        with get_sqlalchemy_conn("postgis") as conn:
+            domain = pd.read_sql(
+                text(
+                    f"""
+                select distinct phenomena || '.' || significance as key from
+                warnings_{year}
+                """
+                ),
+                conn,
+                index_col="key",
+            )
+    else:
+        domain = statdf
+
+    combos = None
+    for panel in mp.panels:
+        df2 = df.to_crs(panel.crs).cx[
+            slice(*panel.get_xlim()), slice(*panel.get_ylim())
+        ]
+
+        if df2.empty:
+            continue
+        combos = df2["ps"].unique().tolist()
+    if not combos or len(mp.panels) > 1:
+        combos = df["ps"].unique().tolist()
+    handles = []
+    labels = []
+    alphas = []
+    for key in domain.index.values:
+        present = False
+        if key in combos:
+            present = True
+            combos.remove(key)
+        handles.append(
+            Rectangle(
+                (0, 0),
+                1,
+                1,
+                fc=NWS_COLORS.get(key, "k"),
+                alpha=1 if present else 0.2,
+            )
+        )
+        label = get_ps(key)
+        if "zone_pop" in statdf.columns:
+            label += (
+                f"\n{statdf.at[key, 'final_pop']:,.0f} ppl,"
+                f" {statdf.at[key, 'final_area_sqkm']:,.0f} km$^2$"
+            )
+        labels.append(label)
+        alphas.append(1 if present else 0.2)
+    for key in combos:
+        handles.append(
+            Rectangle(
+                (0, 0),
+                1,
+                1,
+                fc=NWS_COLORS.get(key, "k"),
+            )
+        )
+        labels.append(get_ps(key))
+        alphas.append(1)
+    legend = mp.panels[0].ax.legend(
+        handles,
+        labels,
+        ncol=5,
+        bbox_to_anchor=(-0.03, -0.02),
+        loc="upper left",
+        fontsize=9,
+    )
+    for i, txt in enumerate(legend.get_texts()):
+        txt.set_alpha(alphas[i])
+
+
+def plotdf(mp, df):
+    """Do plotting."""
+    df["color"] = df["ps"].apply(lambda x: NWS_COLORS.get(x, "k"))
+    for panel in mp.panels:
+        for sig in ["S", "Y", "A", "W"]:
+            df2 = (
+                df[df["significance"] == sig]
+                .to_crs(panel.crs)
+                .cx[slice(*panel.get_xlim()), slice(*panel.get_ylim())]
+            )
+
+            if df2.empty:
+                continue
+            df2.plot(
+                ax=panel.ax,
+                aspect=None,
+                color=df2["color"],
+                zorder=Z_OVERLAY2 + 2,
+            )
+
+
+def plotter(fdict):
+    """Go"""
+    ctx = get_autoplot_context(fdict, get_description())
+    valid = ctx["valid"].replace(tzinfo=timezone.utc)
+    isscol = "issue" if ctx["opt"] == "within" else "product_issue"
+    with get_sqlalchemy_conn("postgis") as conn:
+        ugcdf = gpd.read_postgis(
+            text(
+                f"""
+            select w.ugc, simple_geom, phenomena, significance,
+            phenomena || '.' || significance as ps,
+            w.wfo || eventid || phenomena || significance as key
+            from warnings w
+            JOIN ugcs u on (w.gid = u.gid)
+            WHERE {isscol} <= :valid and expire >= :valid
+            ORDER by phenomena asc
+            """
+            ),
+            conn,
+            params={"valid": valid},
+            geom_col="simple_geom",
+        )
+        sbwdf = gpd.read_postgis(
+            text(
+                """
+            select geom, phenomena, significance,
+            phenomena || '.' || significance as ps,
+            wfo || eventid || phenomena || significance as key
+            from sbw
+            WHERE polygon_begin <= :valid and polygon_end > :valid
+            ORDER by phenomena asc
+            """
+            ),
+            conn,
+            params={"valid": valid},
+            geom_col="geom",
+        )
+    mp = MapPlot(
+        apctx=ctx,
+        nocaption=True,
+        figsize=(12.00, 12.00),
+        axes_position=[0.03, 0.38, 0.94, 0.54],
+    )
+    # Le Sigh
+    mp.fig.text(
+        0.1, 0.965, "NWS Watch, Warning, and Advisory Map", fontsize=20
+    )
+    mp.fig.text(
+        0.1,
+        0.94,
+        (
+            "Based on Unofficial IEM Archives, "
+            f"Map Valid: {valid:%d %b %Y %H:%M} UTC"
+        ),
+        fontsize=14,
+    )
+
+    if not ugcdf.empty:
+        if ctx["add"] == "stats" and valid > utc(2005, 10, 1):
+            uri = (
+                f"http://iem.local/plotting/auto/plot/224/opt:{ctx['opt']}::"
+                f"valid:{valid:%Y-%m-%d%%20%H%M}.csv"
+            )
+            with httpx.Client() as client:
+                res = client.get(uri, timeout=60)
+                if res.status_code != 200:
+                    raise ValueError("Failed to fetch data")
+                sio = StringIO(res.text)
+                statsdf = pd.read_csv(sio).set_index("key")
+        else:
+            statsdf = pd.DataFrame({"key": ugcdf["ps"].unique()}).set_index(
+                "key"
+            )
+        make_legend(mp, ugcdf, statsdf, valid.year - 1)
+        if not sbwdf.empty:
+            # remove rows from ugcdf where key is in sbwdf
+            ugcdf = ugcdf[~ugcdf["key"].isin(sbwdf["key"])]
+        plotdf(mp, ugcdf)
+        if not sbwdf.empty:
+            plotsbw(mp, sbwdf)
+    mp.drawcounties()
+    return mp.fig
+
+
+if __name__ == "__main__":
+    plotter({})

htdocs/sites/porclimo.php

@@ -9,6 +9,12 @@
 $network = $ctx->network;
 $metadata = $ctx->metadata;
 
+// silly bots
+if (strpos($network, "CLIMATE") === false){
+    header("HTTP/1.1 404 Not Found");
+    die();
+}
+
 $t = new MyView();
 $t->title = "Period of Record Daily Climatology";
 $t->sites_current = "porclimo";