Significant ozone loss during winter 2019/2020 observed from ground based microwave radiometer

2026-05-21 - Kiruna, Sweden

Richard Johansson

Todays talk

  1. Scope of the study and instruments
  2. How chemical ozone loss happens
  3. Significance of NH winter 2019/2020
  4. Data and processing
  5. MIRA2 and MLS comparison
  6. Chemically induced ozone depletion
  7. Conclusions and future

Scope of study

  1. Compare ozone obtained with MIRA2 with satellite observations
  2. Use MIRA2 ozone observations to calculate chemical ozone depletion of winter 2019/2020

MIRA2 and MLS

MIRA2

  • Ground-based MWR situated at the Swedish Institute of Space Physics in Kiruna
  • Measure ozone from 273.05 GHz emission line
  • Continous measurement since 2014

Microwave Limb Sounder (MLS)

  • MLS onboard Aura satellite
  • MLS measures atmospheric trace gases, including ozone.
  • Covers altitudes from 5 to 120 km.
  • Provides near-global coverage every three days.
  • Serves as a widely used reference dataset for atmospheric research.
  • Continous measurements since 2004

Polar vortex

  • Pressure system over polar regions during winters
  • Airmass isolation
  • Low temperatures
  • Polar stratospheric clouds (PSC)

Polar vortex edge

  1. Potential vorticity: $Epv$
  2. Edge at maximum of $\nabla Epv$ (Fig. 1b)

Figure 1: (a) Ertel's potentialvorticity ($Epv$), (b) the gradient of $Epv$, and (c) the average wind speed along $Epv$ isolines versus equivalent latitude for the 450 K isentropic surface

PSC Chemistry & Ozone Depletion

Chlorine activation (on PSC surfaces)

  • $T < 195$ K $\rightarrow$ Type I PSC
  • $T < 186$ K $\rightarrow$ Type II PSC

  • Chlorine in active form: $\mathrm{Cl_2}$

Sunlight triggers ozone depletion:

  1. $ \mathrm{Cl_2 + h\nu \rightarrow 2Cl} $
  2. $ \mathrm{Cl + O_3 \rightarrow ClO + O_2} $
  3. $ \mathrm{ClO + O \rightarrow Cl + O_2} $
Net: $ \mathrm{O_3 + O \rightarrow 2O_2} $

Figure 2: ECMWF reanalysis of polar vortex. Gray shading indicate polar vortex edge from Nash criteria. Outer blue countours show isoterms for $T \leq T_{NAT}$, and inner blue contours for $T \leq T_{ICE}$

Winter 2019/2020

Figure 3: Daily 50-hPa minimum temperatures north of 40$^\circ$N (a) and the fraction of the lower-stratospheric vortex below the NAT PSC threshold, $V_{\mathrm{NAT}}/V_{\mathrm{vort}}$ (c). Annual summaries show the number of days with $T < T_{\mathrm{NAT}}$ (b) and the November--March mean $V_{\mathrm{NAT}}/V_{\mathrm{vort}}$ (d). Black lines in (a) indicate approximate NAT and ice PSC thresholds; whiskers in (b,d) show the range from a $\pm 1$ K uncertainty in $T_{\mathrm{NAT}}$ .

Data selection

  • Data quality control
  • 400 km radius from Kiruna
  • $\pm$ 4 h from MIRA2 measurement
  • 456 coincident measurements
  • 145 when obtained inside polar vortex

Figure 4: Spatial coincidence criteria with all MLS observations (black) and within the polar vortex (red cross)

MLS processing

Vertical resolution

  • MLS $\Delta z$ is better than MIRA2
  • No meaningful comparison without smoothing

  1. Regrid MLS to MIRA2
  2. Match AVK
  3. Smooth MLS:

$\mathbf{x}_d = \mathbf{x_a}+\mathbf{A}(\mathbf{x_h - x_a})$

Figure 5: a) show $O_3$ profile comparison between MIRA2 and MLS, while b) show MIRA2 vertical resolution and c) show MIRA2 averaging kernel and measurement response

MIRA2 and MLS timeseries

  • Daily mean of all coincident meansurements
  • Comparisons on four pressure levels:
  1. 74 hPa
  2. 56 hPa
  3. 10 hPa
  4. 1 hPa
  • Generally in good agreement
  • Mostly within $\pm \sigma$

Figure 6: Comparison of coincident $\mathrm{O_3}$ measurements from MIRA2 and MLS between 1 October 2019 and 30 April 2020 at four pressure levels: 1, 10, 56, and 74 hPa (panels a–d).

Vertical coordinate

  • Evaluated on isentropic surface of potential temperature:

$\theta = T\begin{pmatrix}\dfrac{P_0}{P}\end{pmatrix}^{R/c_p}$

  • Easier comparison with previous studies
  • Easier air-parcel tracking

Change in ozone

  • Calculated from daily means
  • Measurements obtained within the polar vortex
  • Change from two sources:
  1. Dynamics
  2. Chemistry

$\Delta O_3 = O_{3P} - O_{3M}$

Dynamics: $O_{3P}$

Dynamics + Chemistry: $O_{3M}$

Tracer relation

  • Use chemically inert for tracking dynamics
  • Nitrous oxide ($N_2O$)
  • Workflow to calculate $\Delta O_3$:
  1. Get $N_2O$ / $O_3$ reference function
  2. Calculate $O_{3P}$ from $N_2O$
  3. Get $O_{3M}$ from MIRA2
  4. Calculate chemically induced loss $\Delta O_3$

Figure 7: $N_2O-O_3$ reference function (black) with observations from MLS and MIRA2 (red scatter). Data is obtained from 2019-12-09 within the polar vortex and between 410 and 600 $K$

Ozone loss during winter 2019/2020

  • $\Delta O_3$ evaluated from early December 2019 to mid April 2020
  • Evaluated on isentropic surface of 475 K

Figure 8: Cumulative chemically induced ozone loss ($\Delta O_3=O_{3P} - O_{3M}$) at the 475 K isentropic level derived from MIRA2 observations. Shaded region indicate total retrieval uncertainty $\sigma$ .

  • Maximum loss observed early April 2020
  • Chemical induced loss of $2.04\pm 0.91$ ppmv

Conclusions

  • MIRA2 and MLS agree well
  • Observed ozone loss ($2.04 \pm 0.91$) from MIRA2 also agree with previous studies:
  1. MLS studies show maximum loss of $2.8$ ppm at 460 K
  2. Transport and chemistry models show loss of $2.3-2.6$ ppm

Future work

This study

  • Under peer review
  • First comments are positive
  • Revision and publication during summer

Water vapour retrievals

  • Guest instrument: WASPAM
  • Measure $\mathrm{H_2O}$ at 22.235 GHz

Figure 9: WASPAM spectra of water vapour line centered at 22.235 GHz. Spectra is obtained from 10h integration centered at midnight 2025-02-04

References