🚩 Refresh crossover track plotting scripts

Main changes are an elev_filter setting in `deepicedrain.vizplots.plot_crossovers`, and the addition of annotated track numbers in the crossover_area map plot using GMT's quoted line functionality. Lots of other little bits too, including a refresh of the subglacial_lake ID (key) and name (value) pairs, a quick improvement to the inefficient brute-force crossover for-loop, and organizing figures inside 'figures/{lakename}' instead of just the 'figures' directory. Still a big chunk of local code yet to be committed, need to refactor and streamline the processing to be able to loop through many subglacial lakes in a more hands off manner!
weiji14 · Oct 26, 2020 · a63fad2 · a63fad2
1 parent 665995a
commit a63fad2
Show file tree

Hide file tree

Showing 3 changed files with 116 additions and 53 deletions.
diff --git a/atlxi_lake.ipynb b/atlxi_lake.ipynb
@@ -917,11 +917,14 @@
    "outputs": [],
    "source": [
     "# Save or load dhdt data from Parquet file\n",
-    "placename: str = \"Recovery\"  # \"Whillans\"\n",
-    "drainage_basins: gpd.GeoDataFrame = drainage_basins.set_index(keys=\"NAME\")\n",
-    "region: deepicedrain.Region = deepicedrain.Region.from_gdf(\n",
-    "    gdf=drainage_basins.loc[placename], name=\"Recovery Basin\"\n",
-    ")\n",
+    "placename: str = \"siple_coast\"  # \"Recovery\"  # \"Whillans\"\n",
+    "try:\n",
+    "    drainage_basins: gpd.GeoDataFrame = drainage_basins.set_index(keys=\"NAME\")\n",
+    "    region: deepicedrain.Region = deepicedrain.Region.from_gdf(\n",
+    "        gdf=drainage_basins.loc[placename], name=\"Recovery Basin\"\n",
+    "    )\n",
+    "except KeyError:\n",
+    "    pass\n",
     "df_dhdt: cudf.DataFrame = cudf.read_parquet(\n",
     "    f\"ATLXI/df_dhdt_{placename.lower()}.parquet\"\n",
     ")"
@@ -935,7 +938,7 @@
    "source": [
     "# Antarctic subglacial lake polygons with EPSG:3031 coordinates\n",
     "antarctic_lakes: gpd.GeoDataFrame = gpd.read_file(\n",
-    "    filename=\"antarctic_subglacial_lakes.geojson\"\n",
+    "    filename=\"antarctic_subglacial_lakes_3031.geojson\"\n",
     ")"
    ]
   },
@@ -948,17 +951,28 @@
    "outputs": [],
    "source": [
     "# Choose one draining/filling lake\n",
-    "draining: bool = False  # False\n",
-    "placename: str = \"Slessor\"  # \"Whillans\"\n",
+    "draining: bool = False\n",
+    "placename: str = \"Whillans\"  # \"Slessor\"  # \"Kamb\"  # \"Mercer\"  #\n",
     "lakes: gpd.GeoDataFrame = antarctic_lakes.query(expr=\"basin_name == @placename\")\n",
-    "lake = lakes.loc[lakes.maxabsdhdt.idxmin() if draining else lakes.maxabsdhdt.idxmax()]\n",
+    "lake = lakes.loc[lakes.inner_dhdt.idxmin() if draining else lakes.inner_dhdt.idxmax()]\n",
+    "# lake = lakes.query(expr=\"inner_dhdt < 0\" if draining else \"inner_dhdt > 0\").loc[63]\n",
     "lakedict = {\n",
-    "    76: \"Subglacial Lake Conway\",  # draining lake\n",
-    "    78: \"Whillans IX\",  # filling lake\n",
-    "    103: \"Slessor 45\",  # draining lake\n",
-    "    108: \"Slessor 23\",  # filling lake\n",
+    "    21: \"Mercer 2b\",  # filling lake\n",
+    "    40: \"Subglacial Lake Conway\",  # draining lake\n",
+    "    48: \"Subglacial Lake Whillans\",  # filling lake\n",
+    "    50: \"Whillans IX\",  # filling lake\n",
+    "    63: \"Kamb 1\",  # filling lake\n",
+    "    65: \"Kamb 12\",  # filling lake\n",
+    "    97: \"MacAyeal 1\",  # draining lake\n",
+    "    109: \"Slessor 45\",  # draining lake\n",
+    "    116: \"Slessor 23\",  # filling lake\n",
+    "    153: \"Recovery IX\",  # draining lake\n",
+    "    157: \"Recovery 3\",  # filling lake\n",
     "}\n",
-    "region = deepicedrain.Region.from_gdf(gdf=lake, name=lakedict[lake.name])"
+    "region = deepicedrain.Region.from_gdf(gdf=lake, name=lakedict[lake.name])\n",
+    "\n",
+    "print(lake)\n",
+    "lake.geometry"
    ]
   },
   {
@@ -1064,6 +1078,10 @@
     "# Parallelized paired crossover analysis\n",
     "futures: list = []\n",
     "for rgt1, rgt2 in itertools.combinations(rgts, r=2):\n",
+    "    # skip if same referencegroundtrack but different laser pair\n",
+    "    # as they are parallel and won't cross\n",
+    "    if rgt1[:4] == rgt2[:4]:\n",
+    "        continue\n",
     "    track1 = track_dict[rgt1][[\"x\", \"y\", \"h_corr\", \"utc_time\"]]\n",
     "    track2 = track_dict[rgt2][[\"x\", \"y\", \"h_corr\", \"utc_time\"]]\n",
     "    future = client.submit(\n",
@@ -1161,14 +1179,20 @@
     "pygmt.makecpt(cmap=\"batlow\", series=[sumstats[var][\"25%\"], sumstats[var][\"75%\"]])\n",
     "# Map frame in metre units\n",
     "fig.basemap(frame=\"f\", region=plotregion, projection=\"X8c\")\n",
-    "# Plot actual track points\n",
+    "# Plot actual track points in green\n",
     "for track in tracks:\n",
-    "    fig.plot(x=track.x, y=track.y, color=\"green\", style=\"c0.01c\")\n",
+    "    tracklabel = f\"{track.iloc[0].referencegroundtrack} {track.iloc[0].pairtrack}\"\n",
+    "    fig.plot(\n",
+    "        x=track.x,\n",
+    "        y=track.y,\n",
+    "        pen=\"thinnest,green,.\",\n",
+    "        style=f'qN+1:+l\"{tracklabel}\"+f3p,Helvetica,darkgreen',\n",
+    "    )\n",
     "# Plot crossover point locations\n",
     "fig.plot(x=df.x, y=df.y, color=df.h_X, cmap=True, style=\"c0.1c\", pen=\"thinnest\")\n",
-    "# PLot lake boundary\n",
+    "# Plot lake boundary in blue\n",
     "lakex, lakey = lake.geometry.exterior.coords.xy\n",
-    "fig.plot(x=lakex, y=lakey, pen=\"thin\")\n",
+    "fig.plot(x=lakex, y=lakey, pen=\"thin,blue,-.\")\n",
     "# Map frame in kilometre units\n",
     "fig.basemap(\n",
     "    frame=[\n",
@@ -1180,7 +1204,7 @@
     "    projection=\"X8c\",\n",
     ")\n",
     "fig.colorbar(position=\"JMR\", frame=['x+l\"Crossover Error\"', \"y+lm\"])\n",
-    "fig.savefig(f\"figures/crossover_area_{placename}.png\")\n",
+    "fig.savefig(f\"figures/{placename}/crossover_area_{placename}_{min_date}_{max_date}.png\")\n",
     "fig.show()"
    ]
   },
@@ -1252,7 +1276,7 @@
     "# Plot dashed line connecting points\n",
     "fig.plot(x=df_max.t, y=df_max.h, pen=f\"faint,blue,-\")\n",
     "fig.savefig(\n",
-    "    f\"figures/crossover_point_{placename}_{track1}_{track2}_{min_date}_{max_date}.png\"\n",
+    "    f\"figures/{placename}/crossover_point_{placename}_{track1}_{track2}_{min_date}_{max_date}.png\"\n",
     ")\n",
     "fig.show()"
    ]
@@ -1264,8 +1288,8 @@
    "outputs": [],
    "source": [
     "# Plot all crossover height points over time over the lake area\n",
-    "fig = deepicedrain.vizplots.plot_crossovers(df=df_th, regionname=region.name)\n",
-    "fig.savefig(f\"figures/crossover_many_{placename}_{min_date}_{max_date}.png\")\n",
+    "fig = deepicedrain.plot_crossovers(df=df_th, regionname=region.name)\n",
+    "fig.savefig(f\"figures/{placename}/crossover_many_{placename}_{min_date}_{max_date}.png\")\n",
     "fig.show()"
    ]
   },
@@ -1280,10 +1304,15 @@
     "normfunc = lambda h: h - h.iloc[0]  # lambda h: h - h.mean()\n",
     "df_th[\"h_norm\"] = df_th.groupby(by=\"track1_track2\").h.transform(func=normfunc)\n",
     "\n",
-    "fig = deepicedrain.vizplots.plot_crossovers(\n",
-    "    df=df_th, regionname=region.name, elev_var=\"h_norm\"\n",
+    "fig = deepicedrain.plot_crossovers(\n",
+    "    df=df_th,\n",
+    "    regionname=region.name,\n",
+    "    elev_var=\"h_norm\",\n",
+    "    elev_filter=3 * abs(df.h_X).median(),\n",
+    ")\n",
+    "fig.savefig(\n",
+    "    f\"figures/{placename}/crossover_many_normalized_{placename}_{min_date}_{max_date}.png\"\n",
     ")\n",
-    "fig.savefig(f\"figures/crossover_many_normalized_{placename}_{min_date}_{max_date}.png\")\n",
     "fig.show()"
    ]
   },
@@ -1315,7 +1344,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.3"
+   "version": "3.8.6"
   }
  },
  "nbformat": 4,

diff --git a/atlxi_lake.py b/atlxi_lake.py
@@ -343,11 +343,14 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 
 # %%
 # Save or load dhdt data from Parquet file
-placename: str = "Recovery"  # "Whillans"
-drainage_basins: gpd.GeoDataFrame = drainage_basins.set_index(keys="NAME")
-region: deepicedrain.Region = deepicedrain.Region.from_gdf(
-    gdf=drainage_basins.loc[placename], name="Recovery Basin"
-)
+placename: str = "siple_coast"  # "Recovery"  # "Whillans"
+try:
+    drainage_basins: gpd.GeoDataFrame = drainage_basins.set_index(keys="NAME")
+    region: deepicedrain.Region = deepicedrain.Region.from_gdf(
+        gdf=drainage_basins.loc[placename], name="Recovery Basin"
+    )
+except KeyError:
+    pass
 df_dhdt: cudf.DataFrame = cudf.read_parquet(
     f"ATLXI/df_dhdt_{placename.lower()}.parquet"
 )
@@ -356,23 +359,34 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 # %%
 # Antarctic subglacial lake polygons with EPSG:3031 coordinates
 antarctic_lakes: gpd.GeoDataFrame = gpd.read_file(
-    filename="antarctic_subglacial_lakes.geojson"
+    filename="antarctic_subglacial_lakes_3031.geojson"
 )
 
 # %%
 # Choose one draining/filling lake
-draining: bool = False  # False
-placename: str = "Slessor"  # "Whillans"
+draining: bool = False
+placename: str = "Whillans"  # "Slessor"  # "Kamb"  # "Mercer"  #
 lakes: gpd.GeoDataFrame = antarctic_lakes.query(expr="basin_name == @placename")
-lake = lakes.loc[lakes.maxabsdhdt.idxmin() if draining else lakes.maxabsdhdt.idxmax()]
+lake = lakes.loc[lakes.inner_dhdt.idxmin() if draining else lakes.inner_dhdt.idxmax()]
+# lake = lakes.query(expr="inner_dhdt < 0" if draining else "inner_dhdt > 0").loc[63]
 lakedict = {
-    76: "Subglacial Lake Conway",  # draining lake
-    78: "Whillans IX",  # filling lake
-    103: "Slessor 45",  # draining lake
-    108: "Slessor 23",  # filling lake
+    21: "Mercer 2b",  # filling lake
+    40: "Subglacial Lake Conway",  # draining lake
+    48: "Subglacial Lake Whillans",  # filling lake
+    50: "Whillans IX",  # filling lake
+    63: "Kamb 1",  # filling lake
+    65: "Kamb 12",  # filling lake
+    97: "MacAyeal 1",  # draining lake
+    109: "Slessor 45",  # draining lake
+    116: "Slessor 23",  # filling lake
+    153: "Recovery IX",  # draining lake
+    157: "Recovery 3",  # filling lake
 }
 region = deepicedrain.Region.from_gdf(gdf=lake, name=lakedict[lake.name])
 
+print(lake)
+lake.geometry
+
 # %%
 # Subset data to lake of interest
 placename: str = region.name.lower().replace(" ", "_")
@@ -445,6 +459,10 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 # Parallelized paired crossover analysis
 futures: list = []
 for rgt1, rgt2 in itertools.combinations(rgts, r=2):
+    # skip if same referencegroundtrack but different laser pair
+    # as they are parallel and won't cross
+    if rgt1[:4] == rgt2[:4]:
+        continue
     track1 = track_dict[rgt1][["x", "y", "h_corr", "utc_time"]]
     track2 = track_dict[rgt2][["x", "y", "h_corr", "utc_time"]]
     future = client.submit(
@@ -511,14 +529,20 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 pygmt.makecpt(cmap="batlow", series=[sumstats[var]["25%"], sumstats[var]["75%"]])
 # Map frame in metre units
 fig.basemap(frame="f", region=plotregion, projection="X8c")
-# Plot actual track points
+# Plot actual track points in green
 for track in tracks:
-    fig.plot(x=track.x, y=track.y, color="green", style="c0.01c")
+    tracklabel = f"{track.iloc[0].referencegroundtrack} {track.iloc[0].pairtrack}"
+    fig.plot(
+        x=track.x,
+        y=track.y,
+        pen="thinnest,green,.",
+        style=f'qN+1:+l"{tracklabel}"+f3p,Helvetica,darkgreen',
+    )
 # Plot crossover point locations
 fig.plot(x=df.x, y=df.y, color=df.h_X, cmap=True, style="c0.1c", pen="thinnest")
-# PLot lake boundary
+# Plot lake boundary in blue
 lakex, lakey = lake.geometry.exterior.coords.xy
-fig.plot(x=lakex, y=lakey, pen="thin")
+fig.plot(x=lakex, y=lakey, pen="thin,blue,-.")
 # Map frame in kilometre units
 fig.basemap(
     frame=[
@@ -530,7 +554,7 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
     projection="X8c",
 )
 fig.colorbar(position="JMR", frame=['x+l"Crossover Error"', "y+lm"])
-fig.savefig(f"figures/crossover_area_{placename}.png")
+fig.savefig(f"figures/{placename}/crossover_area_{placename}_{min_date}_{max_date}.png")
 fig.show()
 
 
@@ -587,14 +611,14 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 # Plot dashed line connecting points
 fig.plot(x=df_max.t, y=df_max.h, pen=f"faint,blue,-")
 fig.savefig(
-    f"figures/crossover_point_{placename}_{track1}_{track2}_{min_date}_{max_date}.png"
+    f"figures/{placename}/crossover_point_{placename}_{track1}_{track2}_{min_date}_{max_date}.png"
 )
 fig.show()
 
 # %%
 # Plot all crossover height points over time over the lake area
-fig = deepicedrain.vizplots.plot_crossovers(df=df_th, regionname=region.name)
-fig.savefig(f"figures/crossover_many_{placename}_{min_date}_{max_date}.png")
+fig = deepicedrain.plot_crossovers(df=df_th, regionname=region.name)
+fig.savefig(f"figures/{placename}/crossover_many_{placename}_{min_date}_{max_date}.png")
 fig.show()
 
 # %%
@@ -603,10 +627,15 @@ def find_clusters(X: cudf.core.dataframe.DataFrame) -> cudf.core.series.Series:
 normfunc = lambda h: h - h.iloc[0]  # lambda h: h - h.mean()
 df_th["h_norm"] = df_th.groupby(by="track1_track2").h.transform(func=normfunc)
 
-fig = deepicedrain.vizplots.plot_crossovers(
-    df=df_th, regionname=region.name, elev_var="h_norm"
+fig = deepicedrain.plot_crossovers(
+    df=df_th,
+    regionname=region.name,
+    elev_var="h_norm",
+    elev_filter=3 * abs(df.h_X).median(),
+)
+fig.savefig(
+    f"figures/{placename}/crossover_many_normalized_{placename}_{min_date}_{max_date}.png"
 )
-fig.savefig(f"figures/crossover_many_normalized_{placename}_{min_date}_{max_date}.png")
 fig.show()
 
 # %%
diff --git a/deepicedrain/vizplots.py b/deepicedrain/vizplots.py
@@ -35,7 +35,7 @@ class IceSat2Explorer(param.Parameterized):
     plot_variable = param.Selector(
         default="dhdt_slope", objects=["referencegroundtrack", "dhdt_slope", "h_corr"]
     )
-    cycle_number = param.Integer(default=7, bounds=(2, 7))
+    cycle_number = param.Integer(default=7, bounds=(2, 8))
     dhdt_range = param.Range(default=(1.0, 10.0), bounds=(0.0, 20.0))
     rasterize = param.Boolean(default=True)
     datashade = param.Boolean(default=False)
@@ -271,6 +271,7 @@ def plot_crossovers(
     time_var: str = "t",
     track_var: str = "track1_track2",
     spacing: float = 2.5,
+    elev_filter: float = 1.0,
 ) -> pygmt.Figure:
     """
     Plot to show how elevation is changing at many crossover points over time.
@@ -316,6 +317,9 @@ def plot_crossovers(
         Provide as a 'dy' increment, this is passed on to `pygmt.info` and used
         to round up and down the y axis (elev_var) limits for a nicer plot
         frame. Default is 2.5.
+    elev_filter : float
+        Minimum elevation change required for the crossover point to show up
+        on the plot. Default is 1.0 (metres).
 
     Returns
     -------
@@ -359,7 +363,7 @@ def plot_crossovers(
     ):
         df_ = df.loc[indexes].sort_values(by=time_var)
         # plot only > 1 metre height change
-        if df_[elev_var].max() - df_[elev_var].min() > 1.0:
+        if df_[elev_var].max() - df_[elev_var].min() > elev_filter:
             track1, track2 = track1_track2.split("x")
             fig.plot(
                 x=df_[time_var],
@@ -374,5 +378,6 @@ def plot_crossovers(
             fig.plot(
                 x=df_[time_var], y=df_[elev_var], Z=i, pen=f"faint,+z,-", cmap=True
             )
-    fig.legend(S=0.5, position="JMR+JMR+o0.2c", box="+gwhite+p1p")
+    with pygmt.config(FONT_ANNOT_PRIMARY="9p"):
+        fig.legend(S=0.8, position="JTR+jTL+o0.2c", box="+gwhite+p1p")
     return fig