Stratospheric water vapour profiling from limb-sounding instruments

Limb-sounding instruments observe the atmosphere tangentially, resolving water vapour mixing ratios from roughly 15 to 80 km altitude at 1–3 km vertical resolution. This page covers microwave emission and solar occultation techniques, their role in radiative forcing and ozone chemistry, and what the 2022 Hunga Tonga eruption revealed about the limits and value of the method.

Sensors

Aura MLS (Microwave Limb Sounder, NASA, launched 2004): Observes thermal emission from water vapour rotational lines near 183 GHz and 190 GHz. Profiles from roughly 316 hPa (~8 km) to 0.002 hPa (~90 km), with vertical resolution of approximately 2.5–3 km in the stratosphere. Global coverage every ~25 hours. Precision on individual profiles is roughly 4–9% in the middle stratosphere. Archive runs from August 2004 to present.
ACE-FTS on SCISAT (Canadian Space Agency, launched 2003): Solar occultation Fourier-transform spectrometer covering 2.2–13.3 µm. Retrieves water vapour mixing ratios from roughly 5 to 95 km with vertical resolution near 3–4 km. Provides roughly 15 occultation events per day, concentrated at high latitudes due to the 74° inclination orbit. Precision is typically better than 5% in the 15–50 km range.
MIPAS on Envisat (ESA, 2002–2012, heritage): Emission-based Fourier-transform spectrometer in the mid-infrared (4.15–14.6 µm). Retrieved water vapour profiles from roughly 6 to 68 km with vertical resolution of 3–8 km depending on observation mode. Full global coverage every three days. Archive is a key reference for pre-2012 stratospheric variability and volcanic events.
SMR on Odin (Swedish-led, launched 2001): Sub-millimetre radiometer observing near 544 GHz for stratospheric water vapour. Vertical resolution approximately 1.5–2 km, altitude range roughly 20–70 km. Operates in shared astronomy/atmosphere mode, so atmospheric observation time is limited to roughly 40% of each orbit. Provides useful cross-validation against MLS.
SAGE III on ISS (NASA, 2017–present): Solar and lunar occultation spectrometer covering 280–1040 nm. Retrieves water vapour in the upper troposphere and stratosphere, with vertical resolution near 0.5 km. Occultation geometry limits daily profiles to roughly 15 solar events. Particularly useful for detecting thin volcanic or convective injection layers.

Why the stratosphere holds its water so tightly, and why that matters

Air entering the stratosphere passes through the cold-point tropopause, typically near 17 km over the tropics, where temperatures can fall below -80 °C. That cold trap freeze-dries the ascending air to mixing ratios of roughly 3–5 parts per million by volume (ppmv). The stratosphere above is consequently very dry, and that dryness is not incidental: water vapour is the dominant greenhouse gas in the stratosphere, and even small fractional changes in its concentration alter the longwave radiative balance measurably. A sustained increase of 1 ppmv in the lower stratosphere has been estimated to produce a radiative forcing on the order of 0.1–0.3 W m⁻², comparable to a decade of CO₂ accumulation at recent rates.

Water vapour also participates directly in stratospheric ozone chemistry. It is the primary source of the hydroxyl radical (OH) in the middle stratosphere, which drives catalytic ozone destruction cycles. Monitoring its vertical distribution is therefore relevant to both climate forcing assessments and ozone-layer science, making it a target of sustained international measurement effort.

How limb sounding sees what nadir instruments cannot

A nadir-looking instrument stares straight down through the full atmospheric column. Its vertical resolution in the stratosphere is poor, typically tens of kilometres, because the signal is dominated by the dense lower atmosphere. A limb sounder points horizontally, tangent to the Earth's surface, so its line of sight passes through a thin shell of atmosphere at the tangent altitude. Move the tangent point up or down and you sample a different altitude layer. The geometry converts horizontal path length into vertical sensitivity, achieving 1–3 km vertical resolution in the stratosphere.

Two physical principles underpin the operational instruments. Microwave emission sounders such as Aura MLS detect thermal emission from water vapour rotational transitions. Because the atmosphere emits at these frequencies regardless of solar illumination, the method works day and night and is unaffected by clouds below the tangent point. Solar occultation instruments such as ACE-FTS and SAGE III measure the attenuation of direct sunlight as the satellite rises or sets through the atmosphere. Occultation is self-calibrating (the unattenuated solar spectrum is measured above the atmosphere on each event), which gives it excellent long-term stability but restricts measurements to sunrise and sunset geometry, roughly 15–30 events per day per instrument.

What the Hunga Tonga eruption demonstrated about anomalous injection

On 15 January 2022 the Hunga Tonga-Hunga Ha'apai submarine volcano erupted with exceptional violence, injecting an estimated 146–150 Tg of water vapour directly into the stratosphere, according to analyses of Aura MLS observations published in the peer-reviewed literature. That single event increased the total stratospheric water vapour burden by roughly 10%, a perturbation with no clear precedent in the satellite record.

MLS detected the plume within days, tracking its ascent to above 30 km and its subsequent spread across both hemispheres over the following months. The vertical profile structure was essential: a nadir sounder would have seen an integrated column anomaly but could not have distinguished whether the water was in the upper troposphere or the middle stratosphere, which matters enormously for its radiative lifetime and chemical impact. The event also exposed a genuine limit of occultation instruments. ACE-FTS and SAGE III, with their restricted daily event counts, sampled the plume intermittently; MLS, with its near-daily global coverage, provided the continuous monitoring needed to track dispersion. Neither technique alone was sufficient.

Honest limits: what limb sounding cannot do

The technique provides no information below roughly 10–15 km. The lower troposphere, where most weather occurs and where water vapour concentrations are orders of magnitude higher, is entirely outside the measurement envelope. Attempting to retrieve tropospheric profiles from limb geometry produces unreliable results because the tangent path through the lower atmosphere is obscured by clouds and the signal is saturated at the relevant frequencies.

Horizontal resolution is poor by the standards of imaging instruments. A single MLS profile integrates over roughly 200–500 km along the flight track, and the across-track footprint is similarly broad. Fine-scale horizontal structure in water vapour, such as thin filaments associated with stratospheric intrusions, is smoothed out. Occultation instruments are worse in this respect: each profile represents a single ray path through the limb, with no across-track sampling at all. The result is that limb sounders excel at vertical structure and long-term trend detection but cannot map horizontal gradients at the resolution a meteorologist might want.

Archive continuity is also a concern. MIPAS ceased operations in 2012. Odin's atmospheric observation time is shared with astronomy. If Aura MLS were to fail without a replacement in orbit, there would be no microwave limb sounder providing continuous global stratospheric water vapour profiles. The gap risk is real and has been noted repeatedly in community assessments.

From profiles to products: what analysis can extract

Raw Level 2 profiles from MLS, ACE-FTS or SMR are useful to atmospheric scientists but require substantial processing before they become decision-relevant. The standard analytical chain involves quality-filtering profiles against published precision and accuracy thresholds (each instrument team publishes version-specific validation papers), constructing zonal-mean climatologies, and computing anomalies against a baseline period. For event detection, such as a volcanic injection, the relevant product is a time series of mixing ratio at a specified altitude range, compared against the multi-year mean and its variability.

Trend analysis demands particular care. Instrument drifts, changes in retrieval algorithm version, and gaps between missions can introduce spurious trends that mimic real atmospheric change. Best practice merges multiple instruments using bias-correction methods tied to overlap periods, a technique well documented in the SPARC (Stratosphere-troposphere Processes And their Role in Climate) data initiative. Satellize applies this published methodology when constructing stratospheric water vapour trend products for clients who need defensible long-term records rather than single-instrument snapshots.

For operational monitoring of anomalous events, the most useful deliverable is a near-real-time alert triggered when MLS profiles in a defined latitude-altitude box exceed a threshold anomaly. MLS data are publicly available with latency of roughly 1–2 days after observation, which is adequate for tracking slow-evolving stratospheric features but not for rapid-response applications.

Reading the record: trend signals buried in the noise

The long-term trend in stratospheric water vapour is one of the more contested quantities in atmospheric science. MLS and its predecessor HALOE (Halogen Occultation Experiment, on UARS, 1991–2005) together provide a record spanning more than three decades. The broad picture shows an upward trend of roughly 0.5–1% per year in some altitude ranges, but the record is punctuated by large interannual variability driven by the quasi-biennial oscillation (QBO) and by episodic volcanic injections. Separating forced trend from natural variability requires the full vertical profile, not just a column integral, because the QBO signal and the trend signal have different vertical structures.

This is precisely where limb sounding's vertical resolution earns its keep. A 2–3 km vertical resolution profile can separate the QBO signature, which is concentrated near 20–30 km, from longer-term moistening in the middle and upper stratosphere. No other operational measurement technique provides this combination of altitude range and vertical resolution at global scale. The method is slow, expensive to maintain in orbit, and geographically coarse. It is also, for this specific problem, irreplaceable.

Typical figures

Vertical resolution (stratosphere)	1.5–3 km (MLS, SMR); 3–4 km (ACE-FTS); 0.5 km (SAGE III occultation)
Altitude range (water vapour)	Approximately 15–80 km (MLS); 5–95 km (ACE-FTS); 20–70 km (SMR)
Horizontal along-track resolution	200–500 km per profile (MLS); single ray path per event (occultation instruments)
Daily global coverage	Near-complete every ~25 hours (MLS); ~15–30 occultation events per day (ACE-FTS, SAGE III)
Frequency / spectral range	183 GHz and 190 GHz (MLS); 2.2–13.3 µm mid-IR (ACE-FTS); 280–1040 nm (SAGE III)
Measurement precision (mid-stratosphere)	4–9% per profile (MLS); better than 5% (ACE-FTS, 15–50 km)
Data latency (MLS)	Approximately 1–2 days after observation (public archive)
Archive depth	MLS from August 2004; ACE-FTS from February 2004; MIPAS 2002–2012; HALOE 1991–2005
Tropospheric information	None. Technique is blind below approximately 10–15 km.
Delivery formats (public data)	HDF-EOS5 (MLS); netCDF (ACE-FTS, SAGE III); accessible via NASA Earthdata and Canadian Space Agency portals

Analytics Satellize can run

Stratospheric water vapour anomaly alert	Profile quality-filtering per instrument team thresholds; zonal-mean anomaly computed against multi-year baseline; threshold exceedance detection	Near-real-time alert (email or API) when mixing ratio in a defined latitude-altitude box exceeds 2-sigma anomaly; updated daily as MLS data arrive
Multi-decadal trend profile	Multi-instrument merging with bias correction during overlap periods (HALOE/MLS, MLS/ACE-FTS); linear and piece-wise trend fitting; QBO and ENSO regression	Altitude-resolved trend report with uncertainty bounds, delivered as netCDF and annotated PDF; suitable for climate-treaty or regulatory submissions
Volcanic injection plume characterisation	Event-triggered profile extraction from MLS Level 2; altitude of maximum anomaly, total mass estimate, hemispheric spread rate derived from successive daily composites	Event report within 72 hours of data availability; includes mixing ratio maps at key pressure levels and time-series plots of plume evolution
QBO-phase stratospheric water vapour composite	QBO phase classification from ERA5 or radiosonde zonal wind; conditional compositing of MLS profiles by phase; anomaly maps at 50, 30, 10 hPa	Seasonal climatology GIS layers and summary statistics; used for seasonal forecast context in climate-sensitive planning
Instrument inter-comparison and drift assessment	Coincidence matching of MLS, ACE-FTS and SAGE III profiles within defined spatial and temporal windows; bias and drift statistics computed per altitude level	Validation report quantifying inter-instrument biases; supports decisions on which instrument record to use for a specific altitude range and time period
Radiative forcing estimate from water vapour anomaly	Anomaly profiles fed into a published offline radiative transfer model (e.g. RRTMG); instantaneous forcing computed at tropopause and top of atmosphere	Forcing estimate with sensitivity range, delivered as a technical note; relevant for post-eruption climate impact assessments

Sources

Who does the work

We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.