New scare: decline of lower stratospheric ozone

There is a new paper by William T. Ball (ETH Zürich) et al.(21 co-authors!!!)  titled “Evidence for a continuous decline in lower stratospheric ozone offsetting ozone layer recovery” published in Atmospheric Chemistry and Physics (6 Feb 2018). This paper has induced many comments by journalists which did not carefully read the paper, and produced the usual scary text about “we will all die by increased UVB radiation”. Actually the paper does not give this conclusion, but uses often well hidden statements to obscure it’s main findings (after heavy data torturing by what I think very obscure statistics):

• the Total Ozone Column (TOC) remained more or less stable since 1998 in the latitudinal band -60° to °60°
• the O3 concentration in the lower stratosphere seems to have declined by about 2 DU since 1998 (remember that the mean of this strongly varying TOC is about 300 DU!)
• the O3 concentration in the upper stratosphere is increasing, what the authors see as a fingerprint of the efficiency of the Montreal protocol
• the O3 in the lower troposphere is also increasing, which the authors see as a fingerprint of human activity

The conclusion of the paper: if the lower stratosphere O3 had not been decreasing, we would notice the efficiency of the Montreal protocol in out-phase O3 destroying gases… but alas, we do not observe any efficiency for the moment.

1. The most important figures from the paper

This is figure 1; it shows the global O3 trends according to the latitude (so every point at a certain latitude is the mean trend for that latitudinal band); red colors show an increase in TOC, blue a decrease.

Figure 4 of the Ball paper shows the tropospheric O3 column (i.e. the ground ozone) is increasing:

Don’t be fooled by the slope of the linear regression line: in 12 years the total increase is just a meager 1.5 DU !

We will compare this to the measurements done at Diekirch and at Uccle (both locations approx. at 50° lat. North, i.e. at the extreme right border of the graphs.

Here is what we measure in Diekirch:

The TOC at Diekirch seems to be slightly decreasing since 2002, even if the general trend since 1998 is positive.

but the ground ozone levels are slightly increasing since 2002 (by 0.2 ug/m3 per year, please compare to the left side scale!)

Uccle finds this for the TOC (link):

So here we see two periods: a decline from about 1980 to 1996, and then an increase!

Uccle also has a good figure with the trends of their balloon soundings (I added the comments):

Here the lower stratosphere corresponds to the yellow marked region: just below that region, we see that over the years the O3 concentration is increasing, and that the changes in the yellow region are minimal.

Conclusion: the regional behaviour at our latitudes (50° North) do not quite correspond to the global latitudinal findings of the Ball paper.

 

2. The UVB radiation measured at ground level.

Here is what we measured in Diekirch during the last 15 years:

UVB intensity remains practically constant over the whole period 2002 to 2017.

I wrote several comments and papers on the relation-ship between TOC and UVB levels at ground level: here the main figure in my paper from 2013:

This figure clearly shows that when the TOC declines, UVB radiation increases (compare the two highlighted days). But alas, things not always go such smoothly during longer periods. The next figure shows the results of measurements done by Klara Czikova et al. in the Czech republic over 50 years (“Reconstruction and analysis of erythemal UV radiation time series  from Hradec Králové (Czech Republic) over the past 50 years“),

 

Just look at the years between the two blue lines: TOC is more or less constant, cloud cover increases and, quite inexplicably the yearly UVB also increases( left scale shows daily mean dose) . This means that short time phenomena can show a different behaviour than yearly averages or totals. Note also the decreasing UVB dose from about 2008 on.

 

3. Conclusions

The findings of the Ball et al. paper may be interesting from a scientific stand-point, but they are not a cause for any panic. The important factor for health reason is the UVB dose, and that dose either remains constant or declines in our region. Does the Ball et al. paper vindicate the Montreal protocol? Yes and no: if really in the upper stratosphere both ozone depleting substances are decreasing and O3 concentrations increasing, than this should point to an efficiency. But the elephant in the room is the decreasing solar (and UVB) activity during the last years, as shown by this graph of the 10.7cm radio waves flux (a proxy for UVB activity):

Clearly solar activity is on a decline since 2000, so less ozone will be created at the lower layers of the stratosphere (even if the O3 destroying substances had remained constant…). The authors ignore this, and it might well be that the O3 depletion in the lower stratosphere is mostly a consequence of declining solar activity!

 

 

 

 

Sea-level budget along US Atlantic North-West coast

An important new paper has been published by Thomas Frederikse et al. (Delft University) on the sea-level changes along the northern part of the US Atlantic West coast (between latitudes 35° and 45°, from Sewells Point to Halifax). The authors try to check if a budget involving changes of salinity and ground level variations would agree with the tide-gauges observations for the last 50 years. I confess that I have a positive bias for sea-level research done by Dutch scientists, as opposed to the scary stories told by people like Stefan Rahmsdorf from the PKI. The Dutch have a centuries long experience with measuring, battling and mitigating a harsh sea that always tries to invade the many regions below sea-level (an area which amounts to about a third of the country). So Dutch research on this topic usually is much more sober and not agenda driven. As a start you may read this paper by Rovere et al. (2016) as a very good and clear introduction into the subject of sea level change.

1. Sea-level changes at different regions are vastly different

The following figure shows how different the sea-levels measured by tide-gauges can be; remember that these gauges are installed on the ground, and strictly speaking measure the relative sea-level. At Stockholm the ground is rising due to the post-glacial rebound (GIA, Glacial Isostatic Adjustment), whereas in Galveston (Texas, USA) there is a big ground subsidence (ground sinking) mostly due to excessive ground water pumping (see here), so that local tide gauges report an alarming (relative) sea-level rise..

For Dutch people the recordings at the Maassluis (a lock on the outlet of the Maas river to the North Sea) are reassuring: in 165 years the relative sea-level rise is only 1.5 mm/year, and shows no sign of acceleration. As the globe is leaving a Little Ice Age since 1850, such a rise has essentially a natural cause, and is not (much) caused by human activity or greenhouse gas emissions! What is surprising is that despite the big differences in the amplitude and sign of the changes, the trends are practically linear i.e. persistent!

The figure also tells us that a global sea-level may be an interesting scientific curiosity, but this modeled “virtual” level has no significance at all for local mitigation policies.

2. What are the main contributors to sea-level change?

Steric changes are changes related to density changes; the sea water density can change for instance by warming (often also called eustatic changes when given relative to a fixed point as the center of the globe) and/or inflow of less saltier water. Lower density means more volume for a given mass i.e. rising sea level if  the geological tub for the oceans remains unchanged (which is not the case!). The following picture shows that the density changes are far from uniform over the globe. As a consequence local steric sea-level changes are quite different, from -2 to + 2 mm/year.

Isostatic changes are related to local vertical ground movements, caused for instance by excessive pumping of ground-water, increased pressure by new buildings or heavy infrastructure, but most importantly by glacial isostatic adjustment (GIA): GIA is the rebound of the earth crust (both positive and negative) caused by the disappearing ice mass that accumulated during the last great glacial period (which ended about 10000 years ago). This is a very slow process, with big regional differences. The Baltic coast for instance is rising at Stockholm by more than 4 mm/year, by 12 mm/year around Greenland (see here); this paper shows that the New Zealand coast has both uplift and subsidence parts with changes from -1 to + 1 mm/year.

The next picture from the paper shows that practically all of the 14 US stations used show negative vertical land movements i.e. subsidence (look at the grey bars: only 4 stations have uplift, mostly negligible except at station #3)

3. Lessons learnt

The major aim of the Frederiksen paper was to establish a model for local sea-level, i.e. making a budget of the different contributions and comparing the effect of this budget to the observations by the tide-gauges. The results are quite good:

As this figure shows, the observations of the tide gauges (grey bars) are very (or at least reasonably) close to the results of the budget (orange bars). Especially interesting is the comparison of the contributions of ice melt (glaciers, Arctic and Antarctic) with the GIA: I have highlighted these on the next table:

The sum of ice-melt is 0.57 mm/yr, that of the GIA (here subsidence) is 1.75 mm/yr, about three times higher! So if we believe that all ice melt is due to the human emissions of greenhouse gases, this anthropogenic “sin” pales in comparison to the natural geological influence.

The acceleration (supposed constant) of the ice-melt caused sea-level would cause in 80 years a sea-level rise of 0.5*(0.009+0.003+0.015)*80**2 = 86.4 mm, less than 10 cm ! This must be compared to the linear geological caused increase of  1.75*80 = 140 mm.

4. Conclusion

The Dutch study does not point to any  human caused rise in sea-level that would present a big problem around 2100. Changes in local (relative) sea-level at the West Atlantic US coast are real, but come predominantly from natural factors. This does not mean that no protection work will have to be done in a far future, but it puts the contribution of human GHG emissions into perspective.

 

PS1: the first two figures are from a slide-show by Frederikse. I lost the link.

PS2: A paper by Paul Sterlini (Koninklijk Nederlands Meteorologisch Instituut) et al. published in GRL July 2017 comes to similar conclusions. The title is “Understanding the spatial variation of sea level rise in the North Sea using satellite altimetry” (paywalled, free access through Researchgate). This paper finds that meteorological effects account for most of the observed regional variation in local SLR. The contribution of ice melt (glaciers + Greenland) around the Dutch coast is shown here at being less than 0.25 mm/yr for the period 1993 to 2014: