The Spread of the Hunga Tonga H2O Plume in the Middle Atmosphere Over the First Two Years Since Eruption

The eruption of Hunga in January 2022 injected a large amount of water into the stratosphere. Satellite measurements from Aura Microwave Limb Sounder (MLS) show that this water vapor (H2O) has now spread throughout the stratosphere and into the lower mesosphere, resulting in an increase of >1 ppmv throughout most of this region. Measurements from three ground‐based Water Vapor Millimeter Wave Spectrometer (WVMS) instruments and MLS are in good agreement, and show that in 2023 there was more H2O in the lower mesosphere than at any time since the WVMS measurements began in the 1990's. At Table Mountain, California all WVMS H2O measurements at 54 km since June 2023, and all of the measurements from Mauna Loa, Hawaii, since the resumption of measurements in September 2023, show larger mixing ratios than any previous measurements. At 70 km several recent WVMS retrievals since September 2023 show the largest anomalies ever measured. The MLS measurements show that maximum H2O anomalies over the 2004–2023 record have occurred throughout almost all of the stratosphere and lower mesosphere since the eruption. As of November 2023, almost all of the ∼140 Tg of water originally injected into the stratosphere by the Hunga eruption remains in the middle atmosphere at pressures below 83 hPa (altitudes above ∼17 km). The eruption occurred during a period when stratospheric H2O was already slightly elevated above the 2004–2021 MLS average, and the November 2023 anomaly of ∼160 Tg represents ∼15% of the total mass of H2O in this region.


Introduction
On 15 January 2022, the eruption of the Hunga undersea volcano injected huge amounts of water vapor (H 2 O) into the atmosphere.Measurements showed that a small amount of this H 2 O was injected directly into the lower mesosphere (Carr et al., 2022;Millan et al., 2022); however, a much larger quantity was injected into the stratosphere.The early evolution of this stratospheric H 2 O plume has been documented in several studies (Khaykin et al., 2022;Millan et al., 2022;Legras et al., 2022;Nedoluha et al., 2023a;Schoeberl et al., 2022).Manney et al. (2023) showed that the H 2 O plume did not affect the 2022 Antarctic vortex, and Santee et al. (2023) showed that while the sulfate aerosol did result in some heterogeneous processing, it did not lead to appreciable chemical ozone loss.Yu et al. (2023) showed 2022 temperature anomalies from SABER in the stratosphere and mesosphere and concluded, based on WACCM6 calculations, that these could, to some extent be explained by changes in stratospheric winds, possibly driven by changes in H 2 O and SO 2 from the Hunga plume.Fleming  2024) have presented some calculations of the effect of this additional H 2 O on O 3 and temperature over the next decade.Nedoluha et al. (2023b) (hereafter N23) documented an increase in mesospheric H 2 O in 2022.Measurements from MLS and three ground-based microwave instruments often recorded the highest H 2 O mixing ratio anomalies ever seen at particular locations.N23 concluded that a portion of this increase was caused by some of the H 2 Oenriched air from Hunga that had ascended through the stratosphere and into the lower mesosphere.But it was also noted that the observed H 2 O increases in 2022 in the lower mesosphere were caused, at least in part, by the unusually slow ascent that allowed for an anomalous amount of additional methane (CH 4 ) oxidation and hence H 2 O production in the upper stratosphere and lower mesosphere.Nedoluha et al. (2023a) showed that the phase of the quasi-biennial oscillation (QBO), and the MLS measurements of N 2 O which provided the evidence of the unusually slow ascent, were similar in 2022 and 2015.
N23 also showed a large positive H 2 O anomaly in the Southern Hemisphere (SH) upper mesosphere in 2022.Using MLS CO measurements and again making use of comparisons between the similarly-phased QBO years of 2015 and 2022, N23 concluded that this upper mesosphere H 2 O anomaly was caused, at least in part, by dynamical variations.We do note that, in the upper mesosphere, H 2 O mixing ratios are affected by the ∼11 years solar cycle (Nedoluha et al., 2009;Remsberg et al., 2018), but the phase of the solar cycle is such that increased photodissociation of H 2 O occurred during the years 2022-2023, to that lower than average H 2 O mixing ratios would be expected in the upper mesosphere.
In this study we document the further spread of the Hunga H 2 O plume in the stratosphere and the increase in mesospheric H 2 O observed in 2023.While the high mesospheric H 2 O mixing ratios observed in 2022 were unprecedented, the increase observed in 2023 was much larger and more widespread.We also document the timescales over which the injected H 2 O mass spread throughout the middle atmosphere from the January 2022 eruption until November 2023.

Ground-Based and Satellite Data Sets
The WVMS instruments have been making nearly continuous measurements of H 2 O in the middle atmosphere since the early 1990's.Measurements are made from the Network for the Detection of Atmospheric Composition Change (NDACC) sites at Table Mountain,California (34.4°N,242.3°E),Mauna Loa,Hawaii (19.5°N,204.4°E),and Lauder,New Zealand (45.0°S,169.7°E).These instruments make spectrally resolved measurements of the 22 GHz H 2 O emission line to obtain a vertical profile of H 2 O.The vertical resolution in the mesosphere is ∼16 km (full width at half maximum).
The standard WVMS measurement product, which will be used in this study, is retrieved from a ∼1 week integration of the spectrum within ±30 MHz of the H 2 O emission peak at 22 GHz.The precise time period is determined by variations in conditions at each site.Results from these retrievals from 1992 to 2021 were presented in Nedoluha et al. (2022), where H 2 O vertical profiles were shown from 45 to 80 km.
The measurements from Mauna Loa were interrupted on 27 November 2022, by a lava flow which cut power, communications, and road access to the site.Solar panels were flown in and installed by helicopter in September 2023, and these provide sufficient power to operate a single WVMS instrument.The WVMS6, which had been providing the long-term data set was initially brought back into operation, but the initial performance was unsatisfactory.Since 26 September 2023 the H 2 O measurements are therefore being made with the WVMS5 instrument, which had previously been used primarily for experimentation and to validate WVMS6 measurements.The WVMS5 instruments makes use of a new corrugated Gaussian horn antenna design developed specifically for WVMS5 by Dr. Jorge Teniente and the Anteral Antenna group (Teniente et al., 2011).This antenna went through many years of development on WVMS5 to improve the reflectivity (S11) and the H and E plane differences in order to reduce the baseline created by the use of both antenna planes and an absorber bar in the signalreference measurement (Gomez et al., 2012).
There has been a slight adjustment to the WVMS H 2 O retrievals from Lauder that are presented here relative to those shown in N23 which particularly affected the H 2 O mixing ratios in July and August 2022.The WVMS retrievals rely upon a background temperature, and in N23 the temperatures for that period were calculated using an MLS climatology of coincident measurements covering the days of the ∼1 week WVMS retrieval.Here we have made use of the average of the contemporaneous MLS temperatures taken during the ∼1 week WVMS.The difference between climatological and measured temperature is usually small, however the year-to-year mesospheric temperatures near Lauder from June through August are particularly variable.In 2022 these differed by >10 K from climatology over several weeks.An incorrect temperature background of 10 K causes an error in the retrieved H 2 O of ∼0.5 ppmv in the mesosphere, and the new WVMS retrievals show variations that are in slightly better agreement with MLS.
The Aura MLS H 2 O product is retrieved from the radiances measured by the radiometers centered near 190 GHz.The MLS v4 H 2 O retrievals were used in Millan et al. (2022) because of poor fits in v5 retrievals in regions of extremely enhanced H 2 O.In this study we use v5 retrievals, but, except in direct comparisons with WVMS measurements, we show monthly zonal medians in order to limit the effect of any problems with MLS measurements when the H 2 O values are extremely enhanced.The v5 retrievals are generally recommended by the MLS Team.Livesey et al. (2021) showed that the v5 retrievals remove an upward drift in the MLS v4 H 2 O measurements of ∼2%-4%/decade from ∼50 hPa to 0.1 hPa since 2010 relative to the Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) (Bernath, et al., 2005).Those profile comparisons showed a kink in the relative drift near 10 hPa, a feature that is apparent in plots showing MLS H 2 O anomalies relative to climatology near this level.
UARS MLS H 2 O measurements in the middle atmosphere were only available from September 1991 to April 1993, at which time the 183 GHz radiometer failed.We show H 2 O variations from 1991 to 2005 as measured by the Halogen Occultation Experiment (HALOE).HALOE observed between 2.45 and 10.0 μm using a solar occultation technique which provided measurements in two separate latitude bands on any day (one in sunrise mode and one in sunset mode).A full description of the design and operation is given by Russell et al. (1993).The results shown here use the HALOE third public release v19 retrievals.

WVMS, MLS, and HALOE Measurements of H 2 O
In Figure 1 we show the monthly zonal median H 2 O mixing ratio anomalies in the stratosphere and mesosphere for January to November 2023.A similar set of panels for 2022 is shown in N23.The anomalies are calculated relative to a 2004-2021 MLS-based climatology.The year begins with positive anomalies almost everywhere, with the largest anomalies in the lower stratosphere and in the tropical upper stratosphere.From January through April the strong tropical anomalies spread both northwards and southwards in the upper stratosphere.During this period the anomaly in the tropical lower stratosphere decreases as younger stratospheric air, unperturbed by the Hunga eruption, rises through the tropopause.Then, in June 2023 mixing ratio anomalies >1 ppmv spread into the northern midlatitude lower mesosphere (∼1 hPa to 0.1 hPa).By August, the lower mesosphere from ∼50 S to 50 N shows anomalies >1 ppmv, with actual mixing ratios >8 ppmv throughout most of this region.In October 2023, almost the entire upper stratosphere (10 to 1 hPa) and lower mesosphere shows anomalies >1 ppmv, with the sole exception being the Northern Hemisphere (NH) upper stratosphere.There is a discontinuity in the anomaly at 10 hPa.This is caused by a discontinuity in the v5 MLS retrieval at this level, as is shown in comparisons with ACE-FTS by Livesey et al. (2021) and is discussed in Millan et al. (2024).
In Figure 2 we show MLS nitrous oxide (N 2 O) measurements (not anomalies) with contours of H 2 O anomalies superimposed.Like N 2 O, CH 4 has a sharply decreasing vertical gradient in the upper stratosphere, and CH 4 measurements from the HALOE instrument in the early 1990's showed a similar "rabbit ear" structure (caused by the Brewer-Dobson circulation) in latitudinal variation in April (Randel et al., 1998).High N 2 O values are indicative of younger air, and Figure 2 shows that in 2023 this rising younger air brings with it H 2 O-enriched air from the Hunga plume.In April both the H 2 O anomaly and the N 2 O mixing ratios in the upper stratosphere are slightly higher in the SH than NH midlatitudes.From May through June, with the beginning of Boreal summer, this hemispheric asymmetry is reversed, and the H 2 O anomalies >1.75 ppmv begin to enter the NH lower mesosphere.
In Figure 3 we show WVMS and coincident MLS H 2 O measurements near the WVMS sites from January 2022 through November 2023.The MLS measurements are an average over a coincident time period (usually ∼1 week) within ±2°latitude, ±30°longitude of the WVMS sites, and are convolved with WVMS averaging kernels appropriate to each site (e.g., Nedoluha et al., 2022).The H 2 O increase at 54 km is first observed at Table Mountain and Mauna Loa, and then slightly later at Lauder.The gap in the WVMS data from Mauna Loa is the result of the cut in power and communications caused by the eruption of Mauna Loa.As noted in Section 2, there are two WVMS instruments at Mauna Loa, with the primary timeseries provided by the WVMS6 instrument.However, since there was some difficulty with the restart of WVMS6, and since the solar panels only provide sufficient power for one instrument, the post-Mauna Loa eruption measurements are currently being provided by WVMS5.Before the Mauna Loa eruption WVMS5 and WVMS6 were in good agreement, as can be seen in Figure 3.
Figure 3 shows the large mid-2023 increase in H 2 O in the lower mesosphere observed by WVMS and coincident MLS measurements at Lauder, Mauna Loa, and Table Mountain.Some of the WVMS measurements are biased slightly high relative to the MLS measurements, but the magnitude of the observed increase in lower mesospheric H 2 O is in good agreement.As was noted in N23 the H 2 O mixing ratios at 54 km were already at record-high levels by the end of 2022.However, based on an analysis of coincident N 2 O measurements and the phase of QBO, it was shown that a significant component of the observed increase in the lower mesosphere was probably caused by dynamical variations (slower ascent, and hence increased CH 4 oxidation to form H 2 O) unrelated to Hunga Tonga.In 2023, however, there was a much larger increase in lower mesospheric H 2 O, and the MLS N 2 O measurements do not indicate that this is a period of unusually slow ascent that would lead to an increase in H 2 O.This increase, which is far outside the range of historical variability, can therefore only be ascribed to the Hunga plume.
The H 2 O anomalies at 70 km in Figure 3 also show an increase throughout much of 2023, but dynamical variations cause much larger anomaly variations at this altitude than in the lower mesosphere.As was noted in N23, the large increase in anomalous H 2 O mixing ratio observed over Lauder in 2022 was coincident with a large decrease in anomalous carbon monoxide (CO).Similar variations in H 2 O and CO occurred in 2015 during a similar phase of the QBO.N23 therefore concluded that a significant portion of the 2022 upper mesospheric increase over Lauder was probably not caused by Hunga Tonga.In late 2022 and early 2023 there was a decrease in H 2 O at 70 km at both Mauna Loa and Table Mountain, followed in subsequent months by an increase of comparable magnitude.At Lauder there was a sharp increase in 70 km H 2 O at the end of 2023.The two measurements of H 2 O anomalies >1 ppmv at 70 km at Mauna Loa in November 2023 are the only such measurements since those WVMS observations began in 1996.Similarly, the 16-23 September 2023 WVMS measurements at Table Mountain which shows an anomaly of >1.5 ppmv is the only such measurement since observations began in 1992, while the second largest anomaly (1.2 ppmv) occurred during the 31 October to 6 November 2023 retrieval.Thus, the MLS and WVMS measurements suggest that increased H 2 O from Hunga is affecting H 2 O values in the upper mesosphere as well.This increase in H 2 O occurs during a period when solar irradiance is increasing and thus causing increased photodissociation of H 2 O (and hence lower H 2 O mixing ratios) in the mesosphere.The Lyman-α solar irradiance in 2023 is higher than at any time since the maxima during solar cycle 23 which peaked in 2001-2002 (Machol et al., 2019).
To put the recent lower mesospheric increase into a multidecadal context we show in Figure 4 the H 2 O timeseries at 54 km at the three WVMS sites since 1991.At Table Mountain all of the ∼weekly 54 km WVMS H 2 O measurements since June 2023, and all of the measurements from Mauna Loa since the resumption of measurements in September 2023, show larger anomalies relative to the MLS climatology, and also larger absolute mixing ratio values, than any previous measurements.At this altitude Lauder is often near the vortex edge from June through August (Harvey et al., 2018) and there are strong latitudinal gradients in mesospheric H 2 O (coincident with the temperature gradients mentioned in Section 2).This large gradient in H 2 O can cause a large weekly H 2 O anomaly that is comparable to the increase caused by Hunga Tonga.Also, the WVMS measurements from Lauder are inherently noisier due to the high tropospheric opacity at the site.But the presence in both WVMS and MLS measurements of anomalies consistently more than 0.9 ppmv above the instrumental average since mid-July 2023 is unprecedented.
The timeseries shown in Figure 4 allows for comparison of the H 2 O variations following the Hunga eruption with the gradual increase seen in the early 1990's in measurements from HALOE, and from the WVMS instruments at Lauder and Table Mountain (Evans et al., 1998;Nedoluha, Bevilacqua, et al., 1998).Nedoluha et al. (2003) showed an increase of ∼0.4 ppmv in upper stratospheric and lower mesospheric measurements from HALOE from 1991 to 1995, with a significant fraction of that increase driven by increased CH 4 oxidation caused by slower tropical upwelling (Nedoluha, Siskind, et al., 1998).When compared to the 1990's H 2 O increase near the stratopause, the change in H 2 O from the Hunga plume is both much larger and occurs over a much shorter period.

The Spread and Persistence of the Hunga H 2 O Plume
In Figure 5 we show the variation of the mass of H 2 O in the stratosphere from 83 hPa to 0.1 hPa since the beginning of 2021 (latitudes from 75°S to 75°N represent ∼97% of the global area).As in previous plots, the values shown are calculated from monthly zonal mixing ratio anomalies from the 2004-2021 climatology, but for Figure 5 these values are integrated over the indicated pressure range and area to provide a mass.The typical annual variation measured by MLS over this entire pressure range and latitude region, which comes almost entirely from the lowermost altitudes, is ∼30-40 Tg.As was shown in Nedoluha et al. (2023a) the 365-day mean 100 hPa tropopause temperature was ∼0.5-0.9K above the 42-year mean in 2021.The high average tropopause temperature during 2021 is probably the primary reason why the mass of H 2 O in the 83 hPa to 31 hPa level is ∼25 Tg above the 2004-2021 climatological value.The value in this level in 2021 is not unusual, but is comparable to the highest levels measured since 2004.
The average mass of H 2 O from 83 hPa to 0.  are nearly identical except in the months immediately following the eruption when the plume had not yet spread evenly over all longitudes.Wilmouth et al. (2023), using 2021 as a background, found that, through the end of 2022, the global H 2 O enhancement from 100 hPa to 1.2 hPa, referenced to 2021, had decreased slightly, from 145 Tg to 135 Tg. Figure 5 also shows that there has been only a slight decrease in the total amount of H 2 O above 83 hPa during the almost 2 years after the eruption.While the anomaly in H 2 O in the 31 hPa to 10 hPa region is now only ∼1/2 as large as immediately after the eruption, a significant fraction of this mass has risen to altitudes with pressures below 10 hPa.
The excess H 2 O in the 83 to 31 hPa layer has also increased since early 2023.Figure 1 suggests that there is some increase in this layer from descending air from the Hunga plume, but about ½ of the increase since March 2023 occurs in the 20°S-20°N region, so this suggests an increase H 2 O entering the stratosphere through the tropical tropopause.Dessler et al. (2014), using trajectory calculations, showed a close correlation between MLS H 2 O anomalies and Modern-Era Retrospective analysis for Research and Applications (MERRA) temperatures in this region (Randel and Park (2019) showed a similar close correlation using Global Positioning System and radiosonde data).Figure 6 shows MERRA2 temperature anomalies at 100 hPa from 20°S to 20°N.These are consistently well above the long-term average from the beginning of 2021 onwards, and there is a further increase in 2023, suggesting that, while descending air from Hunga certainly causes an increase in the lower layer shown in Figure 5, some of this increase could be caused by an increase in H 2 O entering the stratosphere.The H 2 O anomaly above 83 hPa in November 2023 was ∼160 Tg, which represents an excess of ∼15% over the average amount of H 2 O measured in this region by MLS from 2004 to 2021.
In Figure 7 we show a measure of the rate of spread of the H 2 O plume.As in Figure 1, a zonal monthly median has been calculated from monthly MLS data in 2°latitude bins, and a monthly climatology has been subtracted.Figure 7 then shows the month in which the H 2 O mixing ratio anomaly at each latitude and pressure reaches a maximum value for the 2004-2023 MLS time series.The unprecedented global effect of the Hunga eruption is apparent in that, for nearly every location up to 0.1 hPa, the monthly maximum zonal median H 2 O anomaly has  occurred in the months since January 2022.This statement is also true for the maximum absolute H 2 O mixing ratio.
During the first 2 months after the eruption, the increased H 2 O in the plume causes sufficient radiative cooling to significantly perturb the vertical velocity (Coy et al., 2022;Niemeier et al., 2023;Randel et al., 2023).Niemeier et al. (2023) show that this velocity perturbation decreases in subsequent months.There is also gradual CH 4 oxidation that occurs in the stratosphere which increases H 2 O mixing ratios.H 2 O is therefore not a perfectly non-interacting tracer.Nevertheless, we show in Figure 7 the timescale for the spread of the maximum anomaly from the Hunga injection site at ∼20-30 hPa and 20.5°S.The maximum anomaly at (10 S-10 N and 10 hPa rises to 1 hPa an ascent of ∼20 km) in ∼7 months.This is comparable to the timescale calculated from the residual vertical velocity for the tropical pipe region in the Goddard Earth Observing System Chemistry-Climate Model shown in Li et al. (2012), but is somewhat shorter than the timescales calculated for the age-of-air in that study.
Figure 7 shows that the plume spread quickly in the tropics in the low and mid-stratosphere, as was previously noted by Schoeberl et al. (2022).It was first clearly observed in early April by WVMS measurements at Mauna Loa (19.5°N) at 28 km, using a non-standard retrieval technique that provided sensitivity in the mid-stratosphere (Nedoluha et al., 2023a).But the ascent into the mid-stratosphere took much longer, with the maximum H 2 O anomalies in the mid-stratosphere in the tropics not occurring until almost a year after the eruption.
The asymmetry in the evolution of the SH and NH tropics shown in Figure 2, is also apparent in Figure 7, and follows the cross-equatorial variations expected from the Brewer-Dobson circulation.There is a maximum in SH tropical upper stratosphere at 20°S at 1.8 hPa in March 2023, but not until September does the maximum reach 1 hPa.By contrast, the maximum in the NH tropics at 1.8 hPa occurs in May 2023, but it reaches 1 hPa and enters the mesosphere soon thereafter.
There are at least two regions where the month of maximum H 2 O occurs during this time period, but is unrelated to the Hunga plume.The maxima in H 2 O in the SH upper mesosphere in August 2022, and at the highest northern latitudes in February and March 2022, are primarily caused by unusual dynamical conditions, as is evidenced by unusually low CO mixing ratios, as was shown in N23.

Summary
While the high mesospheric H 2 O mixing ratios observed in 2022 were unprecedented, the increase observed in 2023 was much larger and more widespread.By October 2023 large (>1 ppmv) anomalies were observed throughout the lower mesosphere by Aura MLS and by three WVMS instruments.At Table Mountain all WVMS H 2 O measurements at 54 km since June 2023, and all of the measurements from Mauna Loa since the resumption of measurements in September 2023, show larger mixing ratios than any previous measurements.At 70 km several WVMS measurements in the September-November 2023 time period at Table Mountain and Mauna Loa show the largest anomalies ever measured at these sites.
The MLS measurements allow for the tracking of the spread of the Hunga plume throughout the middle atmosphere, and maximum H 2 O anomaly values over the entire 2004-2023 data record have occurred throughout almost all of the stratosphere and lower mesosphere since the eruption.These measurements also show that, as the water vapor spread, the total mass anomaly in the stratosphere and lower mesosphere (82 hPa to 0.1 hPa) remained nearly constant, so that, 22 months after the eruption, almost all of the water injected into the middle atmosphere during the Hunga eruption remains in the middle atmosphere.
confirm the Microwave Limb Sounder (MLS) observation of a >1 ppmv increase in H 2 O in the lower mesosphere in 2023 • Since the eruption the MLS H 2 O monthly zonal mass anomaly from 76 S to 76 N and 83-0.1 hPa has been larger than any other time in the MLS record • Almost all of the mass of H 2 O injected et al. (

Figure 1 .
Figure 1.The monthly zonal-median H 2 O anomaly relative to a 2004-2021 microwave limb sounder (MLS) climatology for MLS measurements from January-November 2023 for latitudes 75°S to 75°N.Data is shown on the native MLS pressure levels.Indicated altitudes are approximate.The arrows indicate the latitudes of the WVMS sites.Contours are in 0.5 ppmv intervals up to 2 ppmv, and in 1 ppmv intervals for larger H 2 O mixing ratio anomalies.The left-hand color bar was, until the eruption of Hunga Tonga, adequate for all but a very small number of MLS monthly zonal-median H 2 O anomalies near the poles and in the upper mesosphere.The righthand color bar has been added to allow for the differentiation of mixing ratios within the plume.

Figure 2 .
Figure 2. The monthly zonal-median N 2 O mixing ratios (colors) and H 2 O mixing ratio anomalies (lines) as measured by Microwave Limb Sounder from April through July 2023 for 75°S to 75°N.The indicated altitudes are approximate.

Figure 3 .
Figure 3. WVMS H 2 O anomalies (blue and cyan; see text) and coincident Microwave Limb Sounder (MLS) measurements (red) at three water vapor millimeter wave spectrometer (WVMS) sites.The MLS measurements have been convolved with WVMS averaging kernels.The anomalies are calculated relative to an MLS 2004-2021 H 2 O climatology.

Figure 4 .
Figure 4. H 2 O anomalies at 54 km from water vapor millimeter wave spectrometer (WVMS) (blue and cyan; see text), halogen occultation experiment (HALOE) (green) and microwave limb sounder (MLS) (red) at, or coincident with, the three WVMS sites.The HALOE and MLS measurements have been convolved with WVMS averaging kernels.The satellite measurements are referenced to the right-hand axis which, to prevent overplotting of data, is offset by 1 ppmv from the axis for the WVMS measurements.The measurements from HALOE have been adjusted to create an unbiased time series relative to MLS.The blue line shows the average WVMS offset from the MLS climatology.
1 hPa and 75°S to 75°N in 2021 in the MLS measurements was 1096 ± 19 Tg, where the uncertainty indicates the standard deviation of the monthly mass anomaly from 2004 through January 2022.Initial calculations of additional stratospheric H 2 O mass from the MLS measurements range from ∼130 to 150 Tg (Millan et al. (2022), Xu et al. (2022), and Khaykin et al. (2022)), consistent with the increase shown in Figure 5. Results are shown both from zonal median and zonal mean anomalies, and the results

Figure 5 .
Figure 5.The monthly 2 O mass anomaly from 82 hPa to 0.1 hPa as measured by microwave limb sounder (MLS) from 75°S to 75°N.Results are shown both as calculated from the zonal median anomalies (solid) and zonal mean anomalies (dotted).The error bars indicate the standard deviation of the monthly mass anomaly over the indicated pressure ranges for MLS measurements from 2004 through January 2022.Also shown in the labels is the mean of the 2021 H 2 O mass in each pressure layer.

Figure 6 .
Figure 6.Daily anomalies in 100 hPa temperature from 20°S to 20°N from MERRA2 relative to a 1980-present average.The blue line is from a 365-day smoothing.

Figure 7 .
Figure 7.The number of months after the eruption of Hunga that the zonal median microwave limb sounder (MLS) H 2 O anomaly reached its maximum value.The cross indicates the approximate position of much of the original injection.Small dots indicate that the maximum occurred in the last month shown (November 2023).Results are shown for all MLS pressure levels from 100 to 0.01 hPa and in 2°latitude increments.Regions with no symbol (primarily in the upper mesosphere) indicate that the maximum H 2 O anomaly in the 2004-2023 MLS time series occurred before January 2022.