Showcase — NO₂ over China and the 2020 COVID lockdown (Sentinel-5P)¶

When China locked down in early 2020, traffic and industry collapsed and the fall in air pollution was visible from orbit. Sentinel-5P/TROPOMI maps tropospheric NO₂. This notebook maps Feb–Mar 2019 vs 2020 with pyramids' Dataset.plot() and builds a three-year monthly time series.

What this notebook does¶

Map mean NO₂ for Feb–Mar 2019, 2020 and the difference with Dataset.plot().
Build a monthly time series 2019–2021 (the seasonal cycle, lockdown dip, rebound).

Needs GEE_SERVICE_ACCOUNT / GEE_SERVICE_KEY and pyramids-gis[viz].

Setup¶

First the imports and a bit of noise-suppression. pyramids provides Dataset (reading + plotting); earthlens provides the unified EarthLens entry point.

In [1]:

Copied!





import os
import warnings
from datetime import date

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from dotenv import find_dotenv, load_dotenv
from loguru import logger
from pyramids.dataset import Dataset  # pyramids' own reading + plotting

from earthlens import EarthLens

warnings.filterwarnings("ignore")
logger.remove()
import os
import warnings
from datetime import date

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from dotenv import find_dotenv, load_dotenv
from loguru import logger
from pyramids.dataset import Dataset  # pyramids' own reading + plotting

from earthlens import EarthLens

warnings.filterwarnings("ignore")
logger.remove()

2026-06-13 18:11:16 | INFO | pyramids.base.config | Logging is configured.

Credentials¶

The notebook reads its credentials from a .env file at the repository root (loaded with python-dotenv; your PATH is left untouched). Create that .env with at least:

GEE_SERVICE_ACCOUNT="your-sa@your-project.iam.gserviceaccount.com"
GEE_SERVICE_KEY="/path/to/your/service-account-key.json"

In [2]:

Copied!





load_dotenv(find_dotenv(usecwd=True))

SERVICE_ACCOUNT = os.environ["GEE_SERVICE_ACCOUNT"]
SERVICE_KEY = os.environ["GEE_SERVICE_KEY"]
if any(x is None for x in (SERVICE_ACCOUNT, SERVICE_KEY)):
    raise ValueError("Missing credentials in .env")
load_dotenv(find_dotenv(usecwd=True))

SERVICE_ACCOUNT = os.environ["GEE_SERVICE_ACCOUNT"]
SERVICE_KEY = os.environ["GEE_SERVICE_KEY"]
if any(x is None for x in (SERVICE_ACCOUNT, SERVICE_KEY)):
    raise ValueError("Missing credentials in .env")

A helper for plotting derived grids¶

plot_arr() wraps a single-band derived array (a NO₂ column, or the year-on-year difference) back into a Dataset — reusing the source geotransform/EPSG — so pyramids can colour-map and georeference it via Dataset.plot().

In [3]:

Copied!





def plot_arr(arr, like, ax, **kw):
    """Colour-map a single-band derived grid (the NO2 column) with pyramids Dataset.plot."""
    return Dataset.create_from_array(
        arr=arr[None, :, :].astype("float32"), geo=like.geotransform, epsg=like.epsg
    ).plot(band=0, ax=ax, **kw)
def plot_arr(arr, like, ax, **kw):
    """Colour-map a single-band derived grid (the NO2 column) with pyramids Dataset.plot."""
    return Dataset.create_from_array(
        arr=arr[None, :, :].astype("float32"), geo=like.geotransform, epsg=like.epsg
    ).plot(band=0, ax=ax, **kw)

1 · Maps: Feb–Mar 2019 vs 2020¶

Sentinel-5P NO₂ comes from Google Earth Engine, so earthlens treats it as the GEE backend. We pin the dataset, band, and AOI over eastern China up front so the per-year fetches below differ only by their date window.

In [4]:

Copied!

AOI = [108.0, 28.0, 122.5, 41.0]
DATASET = "COPERNICUS/S5P/OFFL/L3_NO2"
NO2 = "tropospheric_NO2_column_number_density"
AOI = [108.0, 28.0, 122.5, 41.0]
DATASET = "COPERNICUS/S5P/OFFL/L3_NO2"
NO2 = "tropospheric_NO2_column_number_density"

Pre-lockdown composite (Feb–Mar 2019)¶

Build the GEE request, authenticate, then load the single mean composite raster. The construct → authenticate → load steps are kept on separate lines, and load() returns a list so we take the first raster.

In [5]:

Copied!





scene_2019 = EarthLens(
    data_source="gee",
    cadence="raw",
    dataset=DATASET,
    variables=[NO2],
    aoi=AOI,
    start="2019-02-01",
    end="2019-03-15",
    scale=7000,
    reducer="mean",
)
scene_2019.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
rasters_2019 = scene_2019.load()
a2019 = rasters_2019[0]
scene_2019 = EarthLens(
    data_source="gee",
    cadence="raw",
    dataset=DATASET,
    variables=[NO2],
    aoi=AOI,
    start="2019-02-01",
    end="2019-03-15",
    scale=7000,
    reducer="mean",
)
scene_2019.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
rasters_2019 = scene_2019.load()
a2019 = rasters_2019[0]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.20s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.20s/img]

Lockdown composite (Feb–Mar 2020)¶

The same request shifted to the 2020 window — the lockdown period.

In [6]:

Copied!





scene_2020 = EarthLens(
    data_source="gee",
    cadence="raw",
    dataset=DATASET,
    variables=[NO2],
    aoi=AOI,
    start="2020-02-01",
    end="2020-03-15",
    scale=7000,
    reducer="mean",
)
scene_2020.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
rasters_2020 = scene_2020.load()
a2020 = rasters_2020[0]
print("composites:", bool(a2019), bool(a2020))
scene_2020 = EarthLens(
    data_source="gee",
    cadence="raw",
    dataset=DATASET,
    variables=[NO2],
    aoi=AOI,
    start="2020-02-01",
    end="2020-03-15",
    scale=7000,
    reducer="mean",
)
scene_2020.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
rasters_2020 = scene_2020.load()
a2020 = rasters_2020[0]
print("composites:", bool(a2019), bool(a2020))

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.01s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.01s/img]

composites: True True

Read the columns onto a shared scale¶

Pull both composites into arrays, convert to µmol/m² for readable numbers, and pick a single 98th-percentile vmax shared across the two years so the maps are directly comparable.

In [7]:

Copied!





u = 1e6
n19 = np.asarray(a2019.read_array(), dtype="float32")
n20 = np.asarray(a2020.read_array(), dtype="float32")
n19 = (n19 if n19.ndim == 2 else n19[0]) * u
n20 = (n20 if n20.ndim == 2 else n20[0]) * u
vmax = float(np.nanpercentile(np.concatenate([n19.ravel(), n20.ravel()]), 98))
u = 1e6
n19 = np.asarray(a2019.read_array(), dtype="float32")
n20 = np.asarray(a2020.read_array(), dtype="float32")
n19 = (n19 if n19.ndim == 2 else n19[0]) * u
n20 = (n20 if n20.ndim == 2 else n20[0]) * u
vmax = float(np.nanpercentile(np.concatenate([n19.ravel(), n20.ravel()]), 98))

Map the two years and their difference¶

The two years are drawn on the shared colour scale; the difference (a derived array) is wrapped back into a Dataset via plot_arr() so it too renders through Dataset.plot().

In [8]:

Copied!





fig, ax = plt.subplots(1, 3, figsize=(16, 5))
plot_arr(
    n19, a2019, ax[0], cmap="YlOrRd", title="NO2 - Feb-Mar 2019", vmin=0, vmax=vmax
)
plot_arr(
    n20,
    a2020,
    ax[1],
    cmap="YlOrRd",
    title="NO2 - Feb-Mar 2020 (lockdown)",
    vmin=0,
    vmax=vmax,
)
plot_arr(
    n20 - n19,
    a2019,
    ax[2],
    cmap="RdBu_r",
    title="change (2020 - 2019)",
    vmin=-vmax / 2,
    vmax=vmax / 2,
)
plt.tight_layout()
plt.show()
print(
    f"mean tropospheric NO2 fell {(1 - np.nanmean(n20)/np.nanmean(n19))*100:.0f}% (Feb-Mar 2020 vs 2019)"
)
fig, ax = plt.subplots(1, 3, figsize=(16, 5))
plot_arr(
    n19, a2019, ax[0], cmap="YlOrRd", title="NO2 - Feb-Mar 2019", vmin=0, vmax=vmax
)
plot_arr(
    n20,
    a2020,
    ax[1],
    cmap="YlOrRd",
    title="NO2 - Feb-Mar 2020 (lockdown)",
    vmin=0,
    vmax=vmax,
)
plot_arr(
    n20 - n19,
    a2019,
    ax[2],
    cmap="RdBu_r",
    title="change (2020 - 2019)",
    vmin=-vmax / 2,
    vmax=vmax / 2,
)
plt.tight_layout()
plt.show()
print(
    f"mean tropospheric NO2 fell {(1 - np.nanmean(n20)/np.nanmean(n19))*100:.0f}% (Feb-Mar 2020 vs 2019)"
)

No description has been provided for this image

mean tropospheric NO2 fell 36% (Feb-Mar 2020 vs 2019)

2 · Monthly time series, 2019–2021¶

To see the seasonal cycle, the lockdown dip, and the rebound, we fetch one mean composite per month across three years and take the regional mean of each.

Fetch one composite per month¶

Loop over every month of 2019–2021, building the same GEE request with a one-month window each time. Each iteration constructs → authenticate → load, then records the regional-mean NO₂ (in µmol/m²).

In [9]:

Copied!





months, vals = [], []
for y in (2019, 2020, 2021):
    for m in range(1, 13):
        s = date(y, m, 1)
        e = date(y + (m == 12), (m % 12) + 1, 1)
        scene = EarthLens(
            data_source="gee",
            cadence="raw",
            dataset=DATASET,
            variables=[NO2],
            aoi=AOI,
            start=s.isoformat(),
            end=e.isoformat(),
            scale=7000,
            reducer="mean",
        )
        scene.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
        rasters = scene.load()
        p = rasters[0]
        if p:
            arr = np.asarray(p.read_array(), dtype="float32")
            arr = arr if arr.ndim == 2 else arr[0]
            months.append(pd.Timestamp(s))
            vals.append(np.nanmean(arr) * 1e6)
print(f"{len(vals)} monthly composites")
months, vals = [], []
for y in (2019, 2020, 2021):
    for m in range(1, 13):
        s = date(y, m, 1)
        e = date(y + (m == 12), (m % 12) + 1, 1)
        scene = EarthLens(
            data_source="gee",
            cadence="raw",
            dataset=DATASET,
            variables=[NO2],
            aoi=AOI,
            start=s.isoformat(),
            end=e.isoformat(),
            scale=7000,
            reducer="mean",
        )
        scene.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
        rasters = scene.load()
        p = rasters[0]
        if p:
            arr = np.asarray(p.read_array(), dtype="float32")
            arr = arr if arr.ndim == 2 else arr[0]
            months.append(pd.Timestamp(s))
            vals.append(np.nanmean(arr) * 1e6)
print(f"{len(vals)} monthly composites")

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.40s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.40s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.66s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.66s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.19s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.19s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.33s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.33s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.17s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.18s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.90s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.90s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.08s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.08s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.76s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.76s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.22s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.22s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.67s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.67s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.13s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.13s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.92s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.92s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.45s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.45s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.88s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.88s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.20s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.20s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.44s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.44s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.60s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.60s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.90s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.90s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.35s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.35s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.30s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.31s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.09s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.09s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.96s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.96s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.01s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.01s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.20s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:04<00:00,  4.20s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.63s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.63s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.32s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.32s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:07<00:00,  7.32s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:07<00:00,  7.32s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.65s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.65s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.68s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.68s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.09s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.10s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.94s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:02<00:00,  2.94s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.77s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.77s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.21s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.21s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.15s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:03<00:00,  3.15s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.87s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:06<00:00,  6.88s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]:   0%|          | 0/1 [00:00<?, ?img/s]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.67s/img]

COPERNICUS/S5P/OFFL/L3_NO2 [tropospheric_NO2_column_number_density]: 100%|██████████| 1/1 [00:05<00:00,  5.67s/img]

36 monthly composites

Plot the series with the lockdown window shaded¶

Draw the monthly means as a line and shade the main China lockdown window (Feb–Apr 2020) so the dip stands out against the seasonal cycle.

In [10]:

Copied!





fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(months, vals, "-o", color="tab:red")
ax.axvspan(
    pd.Timestamp("2020-02-01"),
    pd.Timestamp("2020-04-01"),
    color="0.8",
    alpha=0.6,
    label="China lockdown",
)
ax.set(ylabel="mean NO2 (umol/m^2)", title="Eastern China NO2, monthly 2019–2021")
ax.legend()
fig.autofmt_xdate()
plt.show()
fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(months, vals, "-o", color="tab:red")
ax.axvspan(
    pd.Timestamp("2020-02-01"),
    pd.Timestamp("2020-04-01"),
    color="0.8",
    alpha=0.6,
    label="China lockdown",
)
ax.set(ylabel="mean NO2 (umol/m^2)", title="Eastern China NO2, monthly 2019–2021")
ax.legend()
fig.autofmt_xdate()
plt.show()

Recap¶

The NO₂ maps (including the derived difference) are rendered with Dataset.plot(); the monthly time series is a matplotlib chart.

Try it yourself¶

Re-point AOI, or compare other year pairs.
See the GEE backend reference.