Stratospheric ozone column monitoring for treaty verification

Total ozone column and vertical ozone profiles retrieved from UV backscatter and occultation instruments provide the long-term, calibrated record needed to verify Montreal Protocol compliance and track the stratosphere's slow recovery.

Sensors

TROPOMI (Sentinel-5P): UV/visible/near-IR/SWIR spectrometer; total ozone column at 3.5 × 5.5 km nadir pixel (degraded from original 3.5 × 7 km after 2019 upgrade), daily global coverage, latency roughly 3 hours for near-real-time product and 5 days for offline reprocessed product. Retrieves total column via DOAS and provides vertical sensitivity through the stratosphere.
OMPS Nadir Mapper + Nadir Profiler (Suomi-NPP / NOAA-20): Nadir Mapper measures total ozone column at 50 × 50 km resolution with daily global coverage. Nadir Profiler adds crude vertical information (roughly 5 km vertical resolution in the stratosphere) using wavelength-dependent UV backscatter. Limb Profiler on Suomi-NPP provides 1–2 km vertical resolution profiles from 10–60 km altitude.
MLS (Aura): Microwave Limb Sounder; measures ozone mixing ratio profiles from the upper troposphere to the mesosphere with roughly 3 km vertical resolution in the stratosphere. Particularly valuable in the polar vortex where UV backscatter methods degrade at high solar zenith angles. In continuous operation since 2004.
GOME-2 (Metop-A/B/C): UV/visible spectrometer, total ozone column via DOAS at 40 × 80 km (Metop-A/B) or 40 × 40 km (Metop-C). Three satellites in sequence extend the long-term record back through GOME on ERS-2 (1995), essential for trend analysis. Morning equatorial crossing time complements Sentinel-5P's afternoon orbit.

What a Dobson unit actually measures

The Dobson unit (DU) is the thickness, in hundredths of a millimetre, that the atmosphere's entire ozone column would occupy if compressed to standard temperature and pressure at the Earth's surface. One DU equals 2.687 × 10¹⁶ molecules per square centimetre. A healthy mid-latitude column runs around 300 DU; the Antarctic ozone hole, at its most severe, has dipped below 100 DU in October. That difference, roughly 200 DU, is what the Montreal Protocol is trying to reverse.

The unit matters for treaty verification because it is instrument-independent in principle. In practice, inter-calibration between satellite sensors, and against the Brewer and Dobson ground-based spectrophotometer network, is the unglamorous work that makes any long-term trend credible. A drift of even 1% per decade in a satellite instrument can mimic or mask a real atmospheric signal of the same magnitude.

How UV backscatter retrieves the column, and where it fails

Ozone absorbs solar UV strongly in the Hartley band (200–310 nm) and more weakly in the Huggins bands (310–340 nm). Backscatter instruments measure the ratio of radiance at wavelengths with different ozone absorption cross-sections. The ratio cancels out most surface and cloud effects, leaving a signal dominated by the ozone column above the cloud top. The retrieval is well-understood and has been applied continuously since the Nimbus-7 TOMS instrument in 1978.

The method has a structural weakness at high solar zenith angles, which is precisely where it matters most. Poleward of roughly 80° latitude in winter, the sun stays near or below the horizon, and the backscattered UV signal becomes too weak or geometrically distorted for reliable nadir retrieval. This is not a calibration problem; it is geometry. During the Antarctic winter, when the polar vortex is building and ozone depletion is actively occurring, UV backscatter instruments are effectively blind. Microwave limb sounding (MLS on Aura) and occultation methods fill this gap because they use emission or transmission geometry that does not depend on solar illumination of the nadir scene.

The lowermost stratosphere: where profiles become ambiguous

Vertical ozone profiling adds information that total column measurements cannot provide. The stratospheric ozone layer peaks near 20–25 km altitude, but roughly 10% of the total column sits in the lowermost stratosphere and upper troposphere, between roughly 8 and 15 km. This region is the hardest to profile from space.

UV backscatter profiling relies on the fact that shorter wavelengths are absorbed higher in the atmosphere, so different wavelengths probe different altitude layers. But in the lowermost stratosphere, the weighting functions of adjacent wavelength channels overlap substantially with the troposphere, and tropospheric ozone, aerosols and clouds all contaminate the signal. OMPS Nadir Profiler achieves roughly 5 km vertical resolution there, which means it cannot cleanly separate a 2 km thick ozone anomaly at the tropopause from tropospheric background. MLS has better vertical resolution in the stratosphere but loses sensitivity below roughly 100 hPa (approximately 16 km). No single current instrument resolves this layer cleanly. Assimilation into chemical transport models is the standard approach to bridging the gap.

Building the record that treaty verification demands

The Montreal Protocol entered into force in 1989. Detecting a statistically significant recovery trend in the ozone column against natural variability (the quasi-biennial oscillation, solar cycle, volcanic aerosol loading) requires time series measured in decades, not years. The current consensus, based on merged satellite records and ground station data, is that total column ozone in the upper stratosphere has been recovering at roughly 1–3% per decade since the late 1990s, depending on latitude and season. That signal is real but modest relative to interannual variability.

Maintaining the record means managing instrument transitions carefully. TROPOMI is the current workhorse, but it will eventually be succeeded by instruments on the Copernicus Sentinel-5 satellites (planned for the early 2030s on MetOp-SG). Each handover requires overlap periods and careful cross-calibration. The same challenge applied to the transition from TOMS to OMI to TROPOMI. Gaps or calibration offsets in that chain would corrupt trend estimates and, in a treaty context, could become contested evidence.

What a government buyer actually receives

A national environmental ministry or treaty-reporting body typically needs three things: a current-state product (this month's ozone column over its territory and adjacent polar regions), a trend product (multi-year time series against baseline), and an anomaly-alert product (notification when column values fall below a defined threshold, relevant for UV-index public health advisories as much as for scientific monitoring).

TROPOMI Level-2 total ozone column data are freely available through the Copernicus Data Space Ecosystem with near-real-time latency. The analytical work lies in merging those data with the longer GOME-2 and historical record, applying consistent quality flags, and producing country-specific or region-specific summaries with uncertainty bounds that a non-specialist can interpret. Satellize runs this class of analytics on open constellations including TROPOMI, and the same pipeline architecture used for the Kingdom of Tonga crop-estimation programme applies here: ingest, quality-filter, aggregate, deliver in formats that feed directly into a government's existing reporting workflow.

One honest caveat: satellite data alone cannot substitute for a ground-based Brewer or Dobson spectrophotometer network for absolute calibration anchoring. A government serious about treaty-grade monitoring should maintain at least one well-maintained ground instrument. The satellite record amplifies that point measurement to continental and hemispheric scale; it does not replace it.

Typical figures

Spatial resolution (total column)	3.5 × 5.5 km (TROPOMI); 50 × 50 km (OMPS Nadir Mapper); 40 × 40 km (GOME-2 Metop-C)
Vertical resolution (profiling)	~5 km (OMPS Nadir Profiler, stratosphere); ~3 km (MLS, mid-stratosphere); ~1–2 km (OMPS Limb Profiler)
Revisit / temporal sampling	Daily global coverage (TROPOMI, OMPS, GOME-2); MLS ~3,500 profiles per day at ~165 km along-track spacing
Spectral bands used	UV Hartley band 200–310 nm, Huggins bands 310–340 nm for backscatter; microwave ~240 GHz (MLS)
Measurement unit	Dobson units (DU); 1 DU = 2.687 × 10¹⁶ molecules cm⁻²
Typical retrieval uncertainty (total column)	~1–2% for TROPOMI under clear-sky, mid-latitude conditions; larger at high solar zenith angle (>80°)
NRT data latency	~3 hours (TROPOMI NRT); ~5 days (TROPOMI offline reprocessed)
Archive depth	Continuous satellite record from Nimbus-7 TOMS (1978) through TROPOMI (2017–present); merged multi-sensor records maintained by ESA, NOAA and WMO
High-latitude limitation	UV backscatter unreliable poleward of ~80° in polar night; MLS and occultation methods required for winter polar vortex monitoring
Delivery formats	NetCDF-4 (standard Level-2/3); HDF5 (MLS, OMPS); GeoTIFF and CSV aggregates available via analytics pipeline

Analytics Satellize can run

Monthly total ozone column maps by territory	TROPOMI Level-2 DOAS retrieval, quality-flagged and gridded to 0.1° × 0.1°; merged with GOME-2 for morning/afternoon bias correction	GeoTIFF map set and summary statistics table, delivered monthly
Multi-decadal trend analysis against Montreal Protocol baseline	Merged TOMS/OMI/TROPOMI/GOME-2 time series with QBO, solar-cycle and volcanic-aerosol regression following WMO/UNEP Scientific Assessment methodology	Annual trend report with uncertainty ranges, formatted for national treaty submission
Polar vortex ozone anomaly alerts	MLS stratospheric profile assimilation combined with TROPOMI column thresholding; alert triggered when 7-day mean column falls below user-defined DU threshold over defined region	Email or API alert with supporting map within 6 hours of threshold breach
Ozone profile climatology for a defined region	OMPS Nadir Profiler and MLS profile aggregation, binned by month and latitude band; lowermost stratosphere uncertainty explicitly flagged	NetCDF climatology file with documented uncertainty budget, updated annually
UV index forecast support layer	Total ozone column from TROPOMI NRT combined with surface albedo and cloud fraction to estimate erythemal UV dose; standard WHO UV Index formula	Daily GeoTIFF UV index layer, latency under 6 hours, suitable for public health dashboards
Instrument cross-calibration quality report	Simultaneous nadir overpass (SNO) comparisons between TROPOMI, GOME-2 and ground Brewer/Dobson stations; bias and drift statistics computed per quarter	Quarterly PDF report flagging any sensor drift exceeding 0.5% that could affect treaty-grade trend estimates

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.