12 — Aethalometer¶
Theme: wavelength-resolved black carbon, and what the extra wavelengths buy you.
An aethalometer draws air through a filter and measures how much light the collected particles absorb. Absorption is converted to a black-carbon equivalent mass concentration — the mass of pure soot that would have absorbed that much.
[1]:
import matplotlib.pyplot as plt
import aerosoltools as at
aeth = at.load_aethalometer_file("../../tests/data/Sample_Aethalometer.csv")
print("instrument:", aeth.instrument)
print("columns :", list(aeth.data))
instrument: Aethalometer
columns : ['IR BCc', 'UV BCc', 'Blue BCc', 'Green BCc', 'Red BCc', 'Biomass BCc', 'Fossil fuel BCc', 'AAE', 'All data']
/opt/hostedtoolcache/Python/3.11.15/x64/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
from .autonotebook import tqdm as notebook_tqdm
Why there is a BC value per wavelength¶
There is no single “BC” channel. The instrument measures attenuation independently at each of its five wavelengths, and converts each to its own BC-equivalent mass using a mass absorption cross-section for that wavelength. The raw export makes the chain visible: per colour it carries ATN1/ATN2 (attenuation on two filter spots), BC1/BC2 (the mass from each spot) and BCc — the loading-corrected value combining them, which is what is loaded here.
So all five are BC estimates of the same air, and if only soot were present they would agree. They do not, and that disagreement is the useful part: soot absorbs across the whole spectrum roughly evenly, while brown carbon from wood smoke and organics absorbs far more strongly in the ultraviolet. UV therefore reads high whenever something other than pure soot is present.
By convention IR is reported as “the” black carbon, because it is the channel least contaminated by those other absorbers.
[2]:
for name in ["uv_bcc", "blue_bcc", "green_bcc", "red_bcc", "ir_bcc"]:
channel = getattr(aeth, name)
print(f"{name:10s} mean {channel.mean():9.1f} max {channel.max():9.1f}")
print()
print(f"UV / IR ratio: {aeth.uv_bcc.mean() / aeth.ir_bcc.mean():.3f}")
uv_bcc mean 2781.8 max 31895.0
blue_bcc mean 2850.0 max 32115.0
green_bcc mean 2736.6 max 31068.0
red_bcc mean 2726.7 max 30463.0
ir_bcc mean 2692.8 max 30476.0
UV / IR ratio: 1.033
A ratio above 1 means the UV channel sees absorption the IR channel does not — the signature of non-soot absorbers.
[3]:
fig, ax = plt.subplots(figsize=(10, 4.5))
for name, colour in [("uv_bcc", "violet"), ("blue_bcc", "tab:blue"),
("green_bcc", "tab:green"), ("red_bcc", "tab:red"),
("ir_bcc", "black")]:
ax.plot(aeth.time, getattr(aeth, name),
label=name.replace("_bcc", "").upper(), color=colour, lw=1)
ax.set_ylabel("BC equivalent [ng/m3]")
ax.set_title("The same air, measured at five wavelengths")
ax.legend(ncol=5, loc="lower center", bbox_to_anchor=(0.5, 1.06), frameon=False)
plt.tight_layout()
Per-channel units¶
The channels are not all the same quantity, so each carries its own unit and dtype — the mass concentrations in ng/m³, the Ångström exponent dimensionless.
[4]:
aeth.column_units
[4]:
{'GPS speed': 'km/h',
'Timebase': 's',
'Battery remaining': '%',
'Flow setpoint': 'l/min',
'Flow total': 'l/min',
'Flow1': 'l/min',
'Flow2': 'l/min',
'Sample temp': '°C',
'Sample RH': '%',
'Sample dewpoint': '°C',
'Internal pressure': 'hPa',
'Internal temp': '°C',
'UV BC1 smoothed': 'µg/m³',
'UV BCc smoothed': 'µg/m³',
'Blue BC1 smoothed': 'µg/m³',
'Blue BCc smoothed': 'µg/m³',
'IR BC1 smoothed': 'µg/m³',
'IR BCc smoothed': 'µg/m³',
'BB': '%',
'Delta-C': 'µg/m³',
'Reporting Temp': '°C',
'Reporting Pressure': 'hPa'}
The Angstrom absorption exponent¶
The AAE condenses the spectral shape into one number: how steeply absorption falls off with wavelength. Values near 1 indicate fresh fossil-fuel soot; higher values indicate wood smoke or other organic-rich aerosol.
[5]:
print(f"AAE mean : {aeth.aae.mean():.3f}")
print(f"AAE median : {aeth.aae.median():.3f}")
print(f"AAE range : {aeth.aae.min():.3f} to {aeth.aae.max():.3f}")
AAE mean : 0.990
AAE median : 1.040
AAE range : 0.130 to 1.430
[6]:
fig, ax = plt.subplots(figsize=(10, 3.5))
ax.plot(aeth.time, aeth.aae, color="tab:purple", lw=1)
ax.axhline(1.0, ls="--", color="gray", label="~1: fossil-fuel soot")
ax.axhline(2.0, ls=":", color="saddlebrown", label="~2: biomass burning")
ax.set_ylabel("AAE [-]")
ax.legend(loc="upper right")
ax.set_title("Angstrom absorption exponent over time")
plt.tight_layout()
Source apportionment¶
Because soot and wood smoke have different spectral slopes, the measured AAE can be used to split the signal into a fossil-fuel and a biomass-burning contribution. The instrument reports both directly.
[7]:
print(f"fossil fuel : {aeth.fossil_bcc.mean():9.1f} ng/m3")
print(f"biomass : {aeth.biomass_bcc.mean():9.1f} ng/m3")
print(f"IR total : {aeth.ir_bcc.mean():9.1f} ng/m3")
fossil fuel : 2458.0 ng/m3
biomass : 266.1 ng/m3
IR total : 2692.8 ng/m3
[8]:
fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(aeth.time, aeth.fossil_bcc, label="fossil fuel", color="dimgray", lw=1)
ax.plot(aeth.time, aeth.biomass_bcc, label="biomass burning", color="darkorange", lw=1)
ax.set_ylabel("BC equivalent [ng/m3]")
ax.legend(loc="upper right")
ax.set_title("Apportioned black carbon")
plt.tight_layout()
The apportionment inherits the assumptions behind it — fixed Ångström exponents for the two source types — so treat it as an indication of the mix rather than a measured quantity.
Everything else still applies¶
An Aethalometer is an ordinary time series, so activities, cropping and summaries work as usual. Because the channels are separate columns rather than a size distribution, summaries are per channel.
[9]:
midpoint = len(aeth.time) // 2
aeth.mark_activities({
"First half": [(str(aeth.time[0]), str(aeth.time[midpoint]))],
"Second half": [(str(aeth.time[midpoint + 1]), str(aeth.time[-1]))],
})
aeth.summarize_activities()
Summary of aerosol metrics per activity:
+---------------------+----------+------------+-------------+
| | All data | First half | Second half |
+---------------------+----------+------------+-------------+
| Duration (HH:MM) | 06:37 | 03:19 | 03:18 |
| IR BCc [ng/m³] mean | 2692.791 | 481.422 | 4917.318 |
| IR BCc [ng/m³] std | 4439.713 | 249.864 | 5443.902 |
+---------------------+----------+------------+-------------+
[9]:
| Segment | Duration (HH:MM) | IR BCc [ng/m³] mean | IR BCc [ng/m³] std | |
|---|---|---|---|---|
| 0 | All data | 06:37 | 2692.791 | 4439.713 |
| 1 | First half | 03:19 | 481.422 | 249.864 |
| 2 | Second half | 03:18 | 4917.318 | 5443.902 |
Next: 13 — APS.