GISS Zone Anomalies
Table of Contents
Introduction
This is a look at the Godard Institute for Space Studies' surface temperature data. I'll look at the mean difference between the yearly mean and the historical "normal" mean.
Set Up
Imports
Python
from functools import partial
from pathlib import Path
import os
PyPi
from dotenv import load_dotenv
from holoviews import opts
import holoviews
from holoviews.plotting.links import RangeToolLink
from bokeh.models import CrosshairTool, HoverTool
from bokeh.palettes import Category20
import hvplot.pandas
import pandas
import seaborn
This Project
from bartleby_the_penguin.tangles.embed_bokeh import EmbedBokeh
from graeae.tables import CountPercentage
Load Dotenv
load_dotenv(".env")
HoloViews Backend
holoviews.extension("bokeh", width=800)
The Embedder
OUTPUT_PATH = Path("../../files/posts/giss/giss-zone-anomalies/")
Embed = partial(EmbedBokeh,
folder_path=OUTPUT_PATH)
Some Settings
class Plots:
width = 1200
height = 600
font = "Open Sans"
font_size = "24pt"
line_width = 3
tools = ["hover"]
blue = seaborn.color_palette()[0]
light_blue = Category20[3][1]
red = seaborn.color_palette()[3]
yellow = seaborn.color_palette()[1]
green = seaborn.color_palette()[2]
gray = seaborn.color_palette()[7]
The Data
The data is the Combined Land-Surface Air and Sea-Surface Water Temperature Anomolies' Zonal Annual Means which shows the different annual mean for each zone in a given year.
zone_path = Path(os.environ.get("ZONES")).expanduser()
assert zone_path.is_file()
with zone_path.open() as reader:
giss = pandas.read_csv(reader)
print(giss.describe())
Year Glob NHem SHem 24N-90N \
count 139.000000 139.000000 139.000000 139.000000 139.000000
mean 1949.000000 0.032302 0.056043 0.008561 0.077698
std 40.269923 0.336896 0.393435 0.301848 0.464606
min 1880.000000 -0.490000 -0.540000 -0.490000 -0.580000
25% 1914.500000 -0.200000 -0.220000 -0.235000 -0.280000
50% 1949.000000 -0.070000 -0.010000 -0.080000 0.020000
75% 1983.500000 0.215000 0.210000 0.265000 0.255000
max 2018.000000 0.980000 1.260000 0.710000 1.500000
24S-24N 90S-24S 64N-90N 44N-64N 24N-44N EQU-24N \
count 139.000000 139.000000 139.000000 139.000000 139.000000 139.000000
mean 0.036115 -0.018561 0.111079 0.117770 0.027698 0.027626
std 0.331384 0.295695 0.917715 0.516729 0.356416 0.326111
min -0.650000 -0.470000 -1.640000 -0.710000 -0.590000 -0.720000
25% -0.215000 -0.250000 -0.545000 -0.270000 -0.200000 -0.230000
50% -0.030000 -0.100000 0.020000 0.000000 -0.070000 0.000000
75% 0.255000 0.230000 0.660000 0.360000 0.135000 0.240000
max 0.970000 0.700000 3.050000 1.440000 1.060000 0.930000
24S-EQU 44S-24S 64S-44S 90S-64S
count 139.000000 139.000000 139.000000 139.000000
mean 0.045683 0.020432 -0.069353 -0.078129
std 0.343385 0.312688 0.269380 0.732359
min -0.580000 -0.430000 -0.540000 -2.570000
25% -0.210000 -0.220000 -0.265000 -0.490000
50% -0.030000 -0.080000 -0.090000 0.050000
75% 0.290000 0.260000 0.180000 0.410000
max 1.020000 0.780000 0.450000 1.270000
So we have years from 1880 through 2018.
print(giss.iloc[0])
Year 1880.00 Glob -0.18 NHem -0.31 SHem -0.06 24N-90N -0.38 24S-24N -0.17 90S-24S -0.01 64N-90N -0.97 44N-64N -0.47 24N-44N -0.25 EQU-24N -0.21 24S-EQU -0.13 44S-24S -0.04 64S-44S 0.05 90S-64S 0.67 Name: 0, dtype: float64
Tidying the Data
I'm going to use the locations for names so I'll make them something more meaningful for me. I don't know if these are, strictly speaking, the right names, but it's close enough.
name_remap = {
"Glob": "Global",
"NHem": "Northern Hemisphere",
"SHem": "Southern Hemisphere",
"24N-90N": "Northern Extratropics",
"24S-24N": "Tropics",
"90S-24S": "Southern Extratropics",
"64N-90N": "North Frigid",
"44N-64N": "North Temperate",
"24N-44N": "North Sub-Tropic",
"EQU-24N": "Tropic of Cancer",
"24S-EQU": "Tropic of Capricorn",
"44S-24S": "South Sub-Tropic",
"64S-44S": "South Temperate",
"90S-64S": "South Frigid",
}
giss = giss.rename(columns=name_remap)
print(giss.columns)
Index(['Year', 'Global', 'Northern Hemisphere', 'Southern Hemisphere',
'Northern Extratropics', 'Tropics', 'Southern Extratropics',
'North Frigid', 'North Temperate', 'North Sub-Tropic',
'Tropic of Cancer', 'Tropic of Capricorn', 'South Sub-Tropic',
'South Temperate', 'South Frigid'],
dtype='object')
giss_tidy = giss.melt(id_vars=["Year"], var_name="location",
value_name="Difference From Normal")
print(giss_tidy.iloc[0])
Year 1880 location Global Difference From Normal -0.18 Name: 0, dtype: object
location = CountPercentage(giss_tidy.location)
# plot = location.table.hvplot.table()
# Embed(plot, "location_table")()
location()
| Value | Count | Percentage |
|---|---|---|
| Southern Extratropics | 139 | 7.14 |
| Tropics | 139 | 7.14 |
| North Temperate | 139 | 7.14 |
| North Frigid | 139 | 7.14 |
| Northern Extratropics | 139 | 7.14 |
| North Sub-Tropic | 139 | 7.14 |
| South Sub-Tropic | 139 | 7.14 |
| Southern Hemisphere | 139 | 7.14 |
| Global | 139 | 7.14 |
| South Frigid | 139 | 7.14 |
| South Temperate | 139 | 7.14 |
| Tropic of Cancer | 139 | 7.14 |
| Northern Hemisphere | 139 | 7.14 |
| Tropic of Capricorn | 139 | 7.14 |
The Frigid North
This is a plot of the temperature anomalies for the Frigid North using a Range Finder. You can select a range of years using the mini-plot below the main one.
# The Tools
hover = HoverTool(
tooltips = [
("Year", "@Year"),
("Temperature Anomaly (C)", "@{North Frigid}{0.00}"),
],
formatters={"North Frigid": "numeral"},
mode="vline",
)
# The Parts
curve = holoviews.Curve(
giss, "Year",
("North Frigid", "Temperature Anomaly (C)"),
)
main = curve.relabel("North Frigid Zone").opts(width=Plots.width,
height=Plots.height,
labelled=["y"],
fontsize=Plots.font_size,
color=Plots.blue,
tools=["pan", "zoom_in",
hover, "reset"])
range_finder = curve.opts(width=Plots.width, height=100, yaxis=None, default_tools=[])
line = holoviews.HLine(0)
RangeToolLink(range_finder, main)
layout = (main * line + range_finder).cols(1)
layout = layout.opts(
opts.Layout(shared_axes=False, merge_tools=False),
opts.HLine(color="black", alpha=0.2, line_width=2))
Embed(layout, "north_frigid_zone")()
This plot is kind of unusual in that the column name has a space in it ("North Frigid") so in the hover tool you have to put braces around it.
If we were only looking at one zone I would think this is the most useful type, but when I tried to add multiple zones I couldn't get it to work. HoloViews… I also tried to change the font and couldn't figure out how to do it. I really don't get what bokeh and HoloViews are doing with their documentation.
All Zones Dropdown
This creates a plot with all the zones but only showing one at a time (you select which one using a dropdown). To get it you specify the dropdown column using groupby.
# the Tools
hover = HoverTool(
tooltips = [
("Year", "@Year"),
("Difference from Normal", "@Difference From Normal"),
],
formatters={"Difference From Normal": "numeral"},
mode="vline",
)
plot = giss_tidy.hvplot("Year", groupby="location", width=Plots.width, dynamic=False, tools=[hover])
line = holoviews.HLine(0)
layout = plot * line
file_name = "zones.html"
plot_file = OUTPUT_PATH.joinpath(file_name)
holoviews.save(layout, plot_file)
print("[[file:{}][Link To Plot]]".format(file_name))
All Zones Overlay
The dropdown is a convenient way to look at different zones individually but make it hard to compare different zones. I'm going to put all of them on one plot first to get an overall picture.
plot = giss_tidy.hvplot("Year", "Difference From Normal",
by="location",
title="Difference by Zone",
ylabel="Difference From Normal",
width=Plots.width,
height=Plots.height).opts(
fontsize=Plots.font_size,
)
line = holoviews.HLine(0)
layout = plot * line
Embed(layout, "all_zones")()
This is kind of a messy plot, but it does show that the North Frigid and South Frigid zones seem to have the largest anomalies.
Hemisphere Spikes
hemisphere_map = {
'Global': "Global",
'Northern Hemisphere': "Northern",
'Southern Hemisphere': "Southern",
'Northern Extratropics': "Northern",
'Tropics': "Tropics",
'Southern Extratropics': "Southern",
'North Frigid': "Northern",
'North Temperate': "Northern",
'North Sub-Tropic': "Northern",
'Tropic of Cancer': "Northern",
'Tropic of Capricorn': "Southern",
'South Sub-Tropic': "Southern",
'South Temperate': "Southern",
'South Frigid': "Southern",
}
giss_tidy["hemisphere"] = giss_tidy.location.map(hemisphere_map)
Separate out the hemispheres for plotting.
northern = giss_tidy[giss_tidy.hemisphere=="Northern"]
southern = giss_tidy[giss_tidy.hemisphere=="Southern"]
global_anomalies = giss_tidy[giss_tidy.location=="Global"]
Plot.
# The Tools
hover = HoverTool(
tooltips = [
("Year", "@Year"),
("Difference from Normal", "@{Difference From Normal}"),
("Hemisphere", "@Hemisphere"),
],
formatters={"Difference From Normal": "numeral"},
)
northern_spikes = holoviews.Spikes(northern, "Year", ("Difference From Normal",
"Difference From Normal"),
group="Hemisphere",
label="Northern")
southern_spikes = holoviews.Spikes(southern, "Year", ("Difference From Normal",
"Difference From Normal"),
group="Hemisphere",
label="Southern")
overlay = (northern_spikes
* holoviews.Scatter(northern_spikes)
* southern_spikes
* holoviews.Scatter(southern_spikes)
)
overlay.opts(
opts.Spikes("Hemisphere.Northern", color=Plots.blue),
opts.Spikes("Hemisphere.Southern",
fontsize=Plots.font_size,
color=Plots.red,
width=Plots.width,
height=Plots.height),
opts.Scatter(size=10, tools=[hover]),
)
line = holoviews.HLine(0)
layout = overlay * line
bokeh_layout = holoviews.render(layout)
bokeh_layout.title.text_font = "open sans"
bokeh_layout.legend.location = "top_left"
embed = Embed(bokeh_layout, "all_hemispheres")
embed._figure = bokeh_layout
embed()
This one is a little hard to understand at first. All of the zones are still being plotted, but they are grouped by their hemisphere, so they look different from just the Hemisphere temperatures that come later. Mostly what you're seeing is a highlightt of the most anomalous zone for each year.
I also couldn't figure out how to get the zone information into the hovertool. With hvPlot you can pass in extra columns, but the Spikes didn't like it when I tried… such a frustrating library.
Anyway, it looks like the Northern Hemisphere started out below average while the Southern started out above average, then during the baseline period it sort of flipped, and now they're both above average, but the Northern is even more so.
Hemispheres
This plot shows what happens when you average the zones by hemisphere.
hemispheres = ["Global", "Northern Hemisphere", "Southern Hemisphere"]
hemispheres = giss_tidy[giss_tidy.location.isin(hemispheres)]
plot = hemispheres.hvplot("Year", "Difference From Normal",
ylabel="Difference From Normal (C)",
by="location",
title="Difference by Hemisphere",
fontsize=Plots.font_size,
width=Plots.width, height=Plots.height)
line = holoviews.HLine(0)
layout = plot * line
Embed(layout, "hemispheres")()
This shows a different view than the Zones grouped by Hemisphere, as all the values start out below normal and move in sort of a linear fashion upwards. This probably is the result of the fact that the Frigid Zones are more extreme in their differences than the other zones, so averaging all of the zones supresses them. At least that's my initial thought. Looking at the previous plot, though, it looks quite different - the lowest value in the Hemispheres plot comes in 1917, while the zones grouped by hemisphere puts it at 1930 and 1917 actually doesn't look so special. Maybe they're calculated differently.
Extratropics
zones = ["Global", "Northern Extratropics", "Southern Extratropics", "Tropics"]
zones = giss_tidy[giss_tidy.location.isin(zones)]
plot = zones.hvplot("Year", "Difference From Normal", by="location",
ylabel="Difference From Normal (C)",
fontsize=Plots.font_size,
title="Tropics and Extratropics",
width=Plots.width, height=Plots.height)
line = holoviews.HLine(0)
layout = plot * line
Embed(layout, "tropics")()
The Zones
I did a previous plot "by zone" but in that case I was including all the zones given in the data. In this case we're only looking at the non-overlapping zones - so this chops up the planet by the smallest latitudinal zones that don't overlapy.
zones = ["North Frigid", "North Temperate", "North Sub-Tropic",
"Tropic of Cancer", "Tropic of Capricorn",
"South Frigid", "South Temperate", "South Sub-Tropic"]
zones = giss_tidy[giss_tidy.location.isin(zones)]
plot = zones.hvplot("Year", "Difference From Normal", by="location",
ylabel="Difference From Normal (C)",
title="Difference By (Non-Overlapping) Zones",
fontsize=Plots.font_size,
width=Plots.width, height=Plots.height)
line = holoviews.HLine(0)
layout = plot * line
Embed(layout, "zones")()
Tabbed Plots
giss_tidy = giss_tidy.rename(columns={"location": "Location"})
class PlotGroups:
hemisphere = "Hemishpere"
tropics = "Tropics"
climate_zones = "Zones"
group_map = {
"Global": PlotGroups.hemisphere,
"Northern Hemisphere": PlotGroups.hemisphere,
"Southern Hemisphere": PlotGroups.hemisphere,
"Northern Extratropics": PlotGroups.tropics,
"Southern Extratropics": PlotGroups.tropics,
"Tropics": PlotGroups.tropics,
"North Frigid": PlotGroups.climate_zones,
"North Sub-Tropic": PlotGroups.climate_zones,
"North Temperate": PlotGroups.climate_zones,
"Tropic of Cancer": PlotGroups.climate_zones,
"Tropic of Capricorn": PlotGroups.climate_zones,
"South Temperate": PlotGroups.climate_zones,
"South Sub-Tropic": PlotGroups.climate_zones,
"South Frigid": PlotGroups.climate_zones,
}
giss_tidy["plot_group"] = giss_tidy.Location.map(group_map)
hemispheric = giss_tidy[giss_tidy.plot_group==PlotGroups.hemisphere]
tropical = giss_tidy[giss_tidy.plot_group==PlotGroups.tropics]
zones = giss_tidy[giss_tidy.plot_group==PlotGroups.climate_zones]
assert (len(hemispheric.Location.unique())
+ len(tropical.Location.unique())
+ len(zones.Location.unique())) == len(giss_tidy.Location.unique())
assert len(hemispheric.Location.unique()) == 3
assert len(tropical.Location.unique()) == 3
assert len(zones.Location.unique()) == 8, zones.Location.unique()
cross_hair = CrosshairTool()
plot = (hemispheric.hvplot(x="Year", y="Difference From Normal",
ylabel="Difference From Historical Mean",
by="Location",
width=Plots.width,
height=Plots.height,
tools=[cross_hair]
).relabel("Hemispheres")
* tropical.hvplot(x="Year", y="Difference From Normal",
ylabel="Difference From Historical Mean",
by="Location",
width=Plots.width,
height=Plots.height,
tools=[cross_hair]).relabel(
"Tropics vs Extra-Tropics")
* zones.hvplot(x="Year", y="Difference From Normal",
ylabel="Difference From Historical Mean",
by="Location",
width=Plots.width,
height=Plots.height,
tools=[cross_hair],
).relabel("Climate Zones")
).opts(tabs=True, fontsize=Plots.font_size)
Embed(plot, "tabbed_plots")()