A vertical graph

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import numpy as np
import random

# Generate random data
data = pd.DataFrame(index=pd.date_range(start="2020-01-01", periods=50, freq="30s"),
             data={"pm1":5+np.cumsum([random.uniform(-2, 3) for x in range(50)]),
                   "pm2_5":5+np.cumsum([random.uniform(-2, 3) for x in range(50)]),
                   "pm10":5+np.cumsum([random.uniform(-2, 3) for x in range(50)]),
                   "tvoc":5+np.cumsum([random.uniform(-2, 3) for x in range(50)])})

# Start a plot
fig = plt.figure(figsize=(4.5,8))
ax = fig.add_subplot()

# Prepare out data for stacking
# The .values property of pandas columns accesses the underlying numpy array of 
# numerical columns. Numpy arrays can just be added to one another elementwise.
stack_pm10 = data.pm10.values
stack_pm2_5 = data.pm10.values+data.pm2_5.values
stack_pm1 = data.pm10.values+data.pm2_5.values+data.pm1.values

# Call .fill_betweenx() or .fill_betweeny(), depending on your graph orientation, to
# build your stacked line chart
ax.fill_betweenx(data.index.values, 0, stack_pm10, facecolor="#a7a7a7")
ax.fill_betweenx(data.index.values, stack_pm10, stack_pm2_5, facecolor="#c0c0c0")
ax.fill_betweenx(data.index.values, stack_pm2_5, stack_pm1, facecolor="#dadada")

# Move the ticks and labels to the top off the graph
ax.xaxis.tick_top()

# Trim the margins
ax.margins(x=0, y=0)

# To make the y-axis top -> bottom, do this
ax.invert_yaxis()

# Matplotlib has a date formatter for axes, here we set hours:minutes
ax.yaxis.set_major_formatter(mdates.DateFormatter("%H:%M"))

# You can duplicate an axis with .twiny() (not sure why it's .twiny() instead of
# .twinx() here). Then set its limits the same as your original.
ax2 = ax.twiny()
ax2.set_xlim(ax.get_xlim())

# In our case, we want the axis title on the copy (up top), not the original
ax2.set_xlabel("Particulate concentration, ug/m3")

# Create a legend, which will be in the same order as we defined our areas in
# the .fill_betweenx()
ax.legend(["PM10 Coarse", "PM2.5 Fine", "PM1.0 Ultrafine"])

# This removes some of the margins
plt.tight_layout()

# Initial view, no extras
plt.savefig("0_demo.png")

# You can add a line, the v in axvline() means vertical.
# By saving to a different file, we get to keep each version of the graph.
# Update the legend too!
ax.axvline(x=50, color="black", linestyle="--")
ax.legend(["PM10 Coarse", "PM2.5 Fine", "PM1.0 Ultrafine", "Safe level"])
plt.savefig("1_line.png")

# Annotations have a few options for choosing xycoords= and textcoords=
# 'data' is useful for pointing to specific values, while 'axes fraction' is more
# convenient I find for finding an empty spot and putting text there.
ax.annotate('Annotation (00:05)', xy=(0, pd.to_datetime("2020-01-01 00:05")),  xycoords='data',
            xytext=(0.5, 0.5), textcoords='axes fraction',
            arrowprops=dict(facecolor='black', shrink=0.05),
            horizontalalignment='right', verticalalignment='top',
            )
plt.savefig("2_annotation.png")

# You can also afterwards plot another series
# Update the legend (or not) to have it show up there
ax.plot(data.tvoc, data.index)
ax.legend(["PM10 Coarse", "PM2.5 Fine", "PM1.0 Ultrafine", "Safe level", "TVOC"])
plt.savefig("3_extra_plot.png")

from PIL import Image

im1 = Image.open("./0_demo.png")
im2 = Image.open("./1_line.png")
im3 = Image.open("./2_annotation.png")
im4 = Image.open("./3_extra_plot.png")
im1.save("animation.gif", save_all=True, append_images=[im2, im3, im4], duration=500, loop=0)