Before I start….. Just to let you know, in case you don’t have automatic updates activated, that my friend Jerry has just released an update of his UV4You smartphone app. There are just a few minor changes, and it should definitely look better for people like my little sister, Joanne, 😊 who’ve set their devices to a text size larger than the default.

I found a bunch of plots I’d been working on that don’t yet seem to have found a permanent home, though I did use a version of the first one in a posting from a couple of years back. I thought of them again after my most recent posting (now retitled), where I showed that the peak UVI at southern latitudes is much greater than at corresponding northern latitudes.

Here I expand on a variation of that theme, and show that the cumulative UV risk for descendants (like me) of British and Celtic immigrants to New Zealand also far outweighs that of their forebears and descendants who stayed put. It’s not just another ozone story.

In contrast to those peak values, the plots below show the average daily doses of sunburning UV throughout the year at about 80 different sites around the world from a variety of sources. That ‘ery’ label in the title comes the word ‘erythemal’ which means ‘reddening’ (i.e., sunburning). The plot therefore shows the latitudinal pattern of sunburning UV. The units are ‘Standard Erythemal Doses (SEDs)’, where a dose of just 2 SED is enough to cause a just-discernable skin-reddening in people with the most sensitive skin types. Most of the data points are from a compilation passed on to me a long time ago by my old Australian colleague Peter Gies. The latitudinal gradients are much steeper than for the peak doses I talked about last time because at low latitudes there’s little difference between summer and winter UV whereas high latitudes there’s essentially none during the winter months.

In clean air, the UV increases by about 5 percent for kilometre increase in altitude, so the six sites at altitudes higher than 1 km are labelled. For all other sites, any increases due to their higher altitude will be less than ~5 percent. The two curves show corresponding calculated values based on a climatology of ozone. The solid curve is for clear, unpolluted skies, and the dashed curve assumes an attenuation by clouds and aerosols of 50 percent.

Most of the data points lie between the curves. Exceptions are:

• Three high-altitude sites of South Pole, Mauna Loa Observatory in Hawaii, and ‘Summit’ station in Greenland.

• The other Antarctic sites for which the UV is are higher than calculated because the surface reflectivity is much higher than assumed in the calculation. They can also be affected by the Antarctic ozone hole, so the the ozone climatology is not always appropriate in spring.

• The UK sites, for which measured values are persistently less than 50 percent of those calculated for clear skies. This is a probably a consequence of their cloudy skies. Everybody knows that the UK is not famous for its sunny weather ☹

The daily skin-damaging limit is far exceeded at New Zealand and Australia’s latitudes, where the daily dose is typically around 30 SED/day. Even in the UK, with its average daily dose ~10 SED/day, there’s potential for skin damage.

The next plot shows the same data (excluding the high altitude sites), but plotted as a function of absolute latitude to more easily compare the northern and southern hemispheres. Note that for these annual mean daily doses (unlike peak values I spoke of last time) there’s only a small difference in the calculated values between hemispheres - as shown by the red and blue curves. But the data points for the south are generally much closer the clear sky curves. At mid-latitudes, the average transmission by clouds and aerosols is around 70 percent at most southern hemisphere sites, but closer to only 50 percent at the northern sites.

The next plot shows the daily dose as a function of southern latitude, but just for the range of latitudes encompassing New Zealand and Australia. For most sites, the cloud transmission is close to 0.75. It is significantly higher than that only at Alice Springs. It’s a bit lower at five sites, most significantly at (cloudy) Invercargill, near the southernmost tip of New Zealand.

Finally, the doses experienced at these cities are compared with the mean for UK sites. Most sites receive two or three times as much, but at the two lowest latitude sites (Alice Springs and Darwin), annual doses exceed those in the UK by a factor of more than 4.5. Most site and lie approximately on the curve labelled ‘1.4 x Mean Clim/UK’, which means that the transmission by clouds (and possibly aerosols) is typically 40 percent greater than for the UK sites. Not surprising. Everybody knows that Australia is sunnier than the UK!!! 😊

Interestingly, given that the population-weighted mean latitudes of Australia and New Zealand are not very different (35°S and 39°S respectively), the annual UV doses of UV in each country are quite similar, and around three times more than the UKs (54°N). Thus it’s perhaps not unreasonable that skin cancer rates are similar in New Zealand and Australia, and two or three times as high as in the UK. As noted in my last post, discrepancies in UV between Australasia and the UK will be even larger for peak UV values rather than for these average daily doses considered here.

Oops. I’ve just noticed that there’s no data in those plot from our own instruments at Wellington and Christchurch. I’ll make sure they’re included in a followup post.

This is part of a story I might eventually get around to publishing, but the end-line for that seems to be getting further and further way.

About twenty years ago, my colleague, Ben Liley, made a striking colour-coded map of the world like the one below. It shows the peak UVI at each point of the globe as calculated from daily satellite-derived measurements of ozone collected between 1997 and 2003. Those peaks probably still apply today because that period postdates that of greatest ozone depletion.

The purpose of the exercise was to identify the place in the world where the UV is highest. I’d been asked the question out of the blue in an email from somebody in the USA. I knew the answer would be somewhere in the South American tropical Altiplano because you need high altitude, low ozone, and overhead Sun at the same time as its closest approach to Earth. The Altiplano is the only place that ticks all those boxes. After finding that the enquirer represented a US arms manufacturer, I told him he’d have to pay us a lot of money to find out. He never came back to me after that, but we did the work anyway because it was an interesting question. A good example of ‘blue-sky’ research. 😊

The answer is near the centre of that circled white region in the map below, which is indeed within the South American tropical Altiplano. The peak UVI of 25 occurred near Cusco, Peru (3400 m above sea level), at latitude 13.5°S.

But when drawn with this different colour palette, the map reveals a lot more. It might even go a long way towards explaining why skin cancer rates in New Zealand and Australia are so high compared with the everywhere else in the world. That is, if it’s the peak UV rather than the annual dose that matters. There’s some evidence, for example reading between the lines here (and elsewhere), that this might be the case.

As we’ll see, the map clearly shows that peak UV values here ‘down under’ are rather high. Despite that, our annual doses aren’t, because outside the tropics there’s much less UV in winter (due to the long path-length through the atmosphere when the sun is lower in the sky). At mid-latitudes, the UV at the winter solstice is typically less than a tenth of that at the summer solstice.

The Kiwis and Aussies most at risk from UV-induced skin damage are those with skin types like as their British and Celtic forebears, where lighter skin-colours evolved to ensure they could make still enough vitamin D over the winter months when the UV is too low to initiate the chemical reactions in the skin that lead to its production.

I’ve circled New Zealand and the British Isles to illustrate the starkly higher peak UV levels experienced here in New Zealand by British immigrants. Referring to the colour scale at the bottom, you can see that’s about twice as large in New Zealand. I’ve often remarked that we’re at risk because we have the wrong skin types for our latitude. The latitude of north Auckland is similar to Gibraltar’s, which is far closer to the equator than the UK. Even the southernmost tip of New Zealand is closer to the equator than any part of the British Isles. The population-weighted mean latitudes for each country are 39°S and 53°N respectively.

Even those big differences in latitude don’t tell the whole story. With the colour scale chosen for this version of the map, the boundary between the reds and purples is for UVI = 13. In the southern hemisphere that contour runs straight through latitude 45°S (where I live). In the northern hemisphere, the pattern of UVI is more complicated. It’s higher in mountainous areas like the Tibetan Plateau because in clean air the UVI increases by about 5 percent per kilometre gain in altitude. Along with contrasts between land and sea surface temperatures, these mountain ranges also drive longitudinal differences in ozone, which in turn lead to further changes in UV. But on average, instead of the red-purple contour being near 45°N, it’s about 10 degrees closer to the equator in the northern hemisphere. The peak UVI that I’m exposed to in southern New Zealand is similar to that in North Africa. The skin-types of the people from there are much better adapted than mine!

In North America, the only places that receive similar peak UVIs to ours are the Mountainous interior, Florida, and Hawaii! The most heavily populated areas receive far less. The population-weighted mean latitude for the USA is about 35°N. That’s about 4 degrees closer to the equator than ours, but most of its population never sees UVI values as high as those in even in southern New Zealand.

The map was drawn a long time ago. Outside Antarctica, ozone levels haven’t changed much since then, but the map it still has its limitations. It’s hard to get precise numbers from it, just overall patterns. More importantly, it may not fully account for the effects of pollution, which were harder to quantify in those days. It’s not too bad at rural locations because when I compare ground-based measurements of the peak UVI between rural sites New Zealand and USA, I find that values in New Zealand are higher by around 35 percent. That agrees pretty well with the map data. For example, at 45°N in the Eastern USA near the Canadian border, the peak UVI is close to 10 (which isn’t at all surprising, because when the UVI was first introduced in Canada, it was to give a peak value of 10 at their southernmost US border). At 45°S it’s again about 30 percent higher. I’m sure the comparisons wouldn’t be as good in supercities like Mexico City or Shanghai, where the peak UVI will be lower than shown here. More recent data from NZ and USA show that the previously reported UV differences still seem to apply.

Given this huge difference in the peak UVI levels experienced in New Zealand and Australia compared with northern hemisphere locations inhabited by people of European descent, our high rates of skin cancer aren’t at all surprising. UV levels are also high in South Africa and South America, but their darker skin types result in lower rates of skin cancer.

Check out our new smartphone app, UV4You, which helps to illustrate these differences. In addition to showing the UVI at your current location and time, it also shows the typical peak summer value at your location and allows a comparison with anywhere else in the world. Because of day-to-day differences in ozone, those ‘typical’ summer values are typically 20 percent less than the all-time peaks shown in the above map.

A few months back we announced an update of our old GlobalUV app, which we’d modified to use average monthly ozone amounts rather than daily forecasts from satellites. This had the advantage that it would work off-line. Unfortunately, those climatological values sometimes differ too much from reality, so my friend Jerry Burke has bitten the bullet and written an improved app that can use both. We call it UV4You. I’m chuffed that he agreed to give it a name that’s so closely related to my posts here at UV&You. He wrote it in a new (to me) programming language called Flutter, which is compatible across both iPhone and Android devices. As a result, unlike earlier apps, this one has exactly the same functionality across those platforms.

Our previous apps used global ozone forecasts from NIWA via NOAA in the USA. But, with tRump-induced uncertainties in that chain we’re steering clear of that route. We now use global ozone forecasts from the European Space Agency’s ‘TEMIS’ website that’s maintained by the (Royal) Dutch Meteorological Institute (KNMI). We’re grateful to Jos van Geffen and Michiel van Weele there for their assistance. If no WiFi is available, the app reverts to its off-line ‘climatological’ mode.

We’re no longer using cloud forecasts, as they sometimes over-predict their blocking effect, and that can lead to potentially harmful under-predictions of the UVI. The required downloads to represent the entire world any more than for one time each day would be prohibitively large anyway (but we do still use higher-resolution cloud forecasts in our UVNZ app, which is limited to the New Zealand region). In UV4You we leave it to the users to specify their current cloud conditions.

The basic UV information you need is all available on the startup Map screen, which defaults to your current location. But the interactive Graph screen is much more interesting and contains a wealth of extra information. In both, you can get more details about the calculation parameters by touching the coloured boxes. Their colours match the internationally agreed standards for UV levels.

A big improvement in the new app is that a touch of the ‘Graph’ screen allows you to compare UV on the current day with that for a typical summer day at the same site. The original idea there was to fill the empty space on the graph in winter when the UV is boringly low, and to remind you much it changes with the seasons. In the summer it also allows you to see how much it changes due to day-to-day changes in ozone. Usefully, such comparisons can be made for any location in the world, either by clicking on different places on the map, or selecting specific sites from a drop-down menu. The ‘typical’ summer values you see there are calculated from a table of monthly mean ozone amounts as a function of latitude. Because of ozone’s day-to-day variability and its longitudinal patterns, absolute peaks can sometimes exceed these by more than 20 percent. For example, at Lauder, it gives a summer UV of 10.7 whereas in reality, the all-time absolute peak can exceed 13. To calculate the effects of extinctions by aerosols, the app makes use of a satellite-derived 10-year climatology of aerosol optical depths at 1-degree steps in latitude and longitude. Because of day-to-day differences in air quality, uncertainties are larger at polluted sites.

Behavioural advice and information on how much skin damaging UV accrues over the course of the day are also provided. That dose is expressed as the number of MEDs (Minimum Erythemal Doses). That’s the number of doses that can cause a perceptible reddening of the skin, an indicator of skin damage. The scale for that is on the right. The numbers get quite scary, sometimes exceeding 30 MEDs in summer, which means that if you were outdoors all day with a correctly applied sunscreen of SPF30, you’d still receive a ‘harmful’ dose of UV. The specified times for skin-damage all depend on your skin type, which is specified from a drop-down menu. On touching the graph screen, the app also gives the (very approximate) number of minutes of outdoor exposure that would be required to maintain healthy levels of vitamin D for previously specified attire.

Messages for other times of the day can be seen by sliding your finger over the graph. Fortunately, when the UV is high, the time for vitamin-D sufficiency is much shorter than the time for skin damage.

Some examples are shown below, all from late June 2026.

The first is from Lauder, New Zealand, where it’s a few days short of the winter solstice.

The second example below is from De Bilt in the Netherlands. That’s where our helpful colleagues at KNMI are based. For them, it’s close to the summer solstice. Their peak UVI is close to their expected summer maximum, and over the course of the day the dose will be about 13 MEDs. Also shown on this is the solar elevation and the calculated number of minutes of exposure for a selected time that would be needed to receive enough UV to maintain healthy levels of vitamin D. That’s about 4 minutes in this example (for skin type 2, assuming you’re wearing a t-shirt and shorts). In comparison, the calculated time for skin damage is about 30 minutes.

The third example, below, is from the high-altitude Mauna Loa Observatory in Hawaii, where the peak UVI and daily dose are much higher.

The app has a couple of cute educational features too.

First, there’s an interactive ‘UV Map’ page that shows the current distribution of UVI over the entire globe. You can also see how that pattern changes with month of the year and time of day. It’s pretty cool. I sometimes think the phone might get overheated with the number of calculations involved, but it doesn’t seem to overly drain the battery 😊.

Finally, there is an aptly named ‘Geeks’ screen, written by geeks for the use of geeks. You geekily can play around there to your heart’s content to learn more about what makes the UV tick.

Thinking of taking a holiday abroad or moving to another region? Wherever that might be, if you’re concerned about skin damage, the app lets you to check out the UVI there today, its expected summer peak, and how it varies with seasons. What more could you ask for?

If you want to know more about the physics behind the app, just follow this link.

Enjoy.

As always, the app is free. It’s available for iPhone here and for Android here.

Tell your friends.

PS. Sorry about the incorrect link to Greg Bodeker’s 1996 paper in my last post. The correct link has now been added.

In response to a reader who wondered why so little was known about the subject when I waxed on last time about how difficult it is to assess the effects of clouds on UV radiation. Thanks for getting me thinking, Michael.

The key question was how strongly clouds affect UV compared with visible light. The general view has been that clouds reduce UV less than our visual experience suggests—but by how much? After speaking with a colleague at Lauder, I realised we already knew the answer: we’d published the results way back in the 1990s!

A version of the key figure from that paper is shown below. It shows the effects of clouds at UV wavelengths as a function of their effect at visible wavelengths, where the cloud effects are deduced from ratios of measured-to-calculated values (called Cloud Modification Factors, CMFs).

It uses 1-minute measurements from Lauder of the sunburning UV, and total solar radiation which is proportional to the amount of light we see. Because both of these quantities change rapidly with sun elevation angle, we looked at a few specific ranges of sun angles.

Just to confuse you when describing these sun angles, we atmospheric physicists usually express them in terms of their angle from the overhead zenith (i.e., the Solar Zenith Angle, SZA), rather than their elevation above the horizon that you’re probably more used to. The SZA is just the complement of the elevation, so is 0° for overhead sun, and 90° at twilight.

The plot uses data obtained over the 1- degree interval in SZA from 25° to 26°, which is the smallest SZA-range for which there was enough data. SZAs that small occur on only 50 days each year at this site (from 21 November to 10 January). But on most days it will occur twice: once in the ‘morning’ and once in the ‘afternoon’.

The diagonal line is a quadratic polynomial fit to the data points. For this SZA, it’s pretty close to a straight line.

You can see that both the UV and the total radiation sometimes exceed the clear sky values, occasionally by 20 percent or more. The enhancements tend to be less severe at UV wavelengths, but at times it means that clouds sometimes increase the risk of UV damage! Significant ‘cloud enhancement events’ occur about 10 percent of the time in this example. Luckily, they are usually short-lived, but their occurrence is why we don’t call the ratios cloud ‘transmission’ or ‘attenuation’ factors and instead use the term Cloud modification Factor (CMF).

A CMF of 1 corresponds to clear skies, and the heavier the cloud cover, the further below one the CMFs go. In some case the CMF is less than 0.1, which means that more than 90 percent of the radiation is blocked by the clouds and less than 10 percent is transmitted. In that case, the protection from clouds is the same as an SPF10 sunscreen. At this relatively sunny site, that protection factor (SPF) occurs less than 10 percent of the time, and for most of the time the effective SPF of the clouds is less than 2!

If the cloud effects were the same for both wavelength ranges, the points would fall on a straight line between the origin and the intersection of the two dotted lines. Although the fit isn’t quite a straight line, it’s clear that, to a good approximation, the attenuation by clouds is similar at both wavelengths, with only a slight tendency for points to lie above it. This means that cloud reductions in the UV are not quite as pronounced as at longer wavelengths. But the difference is usually quite small. An extreme difference occurs when the CMF for visible wavelengths is 0.5, the CMF for UV wavelengths is sometimes closer to 0.6. or 0.7. At other times it is lower.

The corresponding plot for SZA = 45°, an angle that is reached on about 180 days of the year at this site (for all days between the spring and autumn equinoxes), is shown below. Note the larger number of data points.

For this SZA, there’s slightly more curvature in the fit line, and UV transmissions are slightly higher than for the visible.

But that curvature is less important from a health perspective because the UVI rarely exceeds 5 for this SZA, compared with the extreme UVI values greater than 10 that occur for SZA= 25°.

At larger SZAs the picture is less clear, but also less important from a health perspective. For example, at SZA=65° (shown below), the UVI is always less than 3. At this larger SZA, the curvature of the fit line is much more pronounced, and the CMFs for UV wavelengths now tend to be significantly greater than for visible wavelengths. The results are much more variable too.

At still larger SZAs the UVI quickly reduces. Results for SZA =85°, when the UVI is close to zero, are also included in the paper but I haven’t bothered to show them here. They’re less reliable because even over very small changes in SZA there can be large percentage differences in UV. You can check them out yourself here if you want to know more.

The bottom line is that cloud do protect you from UV exposure, though by not quite as much as you might think based on what your eyes tell you. At this site at least, their sun protection factor is usually nowhere as good as most commercial sunscreen lotions. You should continue to apply them even in cloudy weather.

In contrast to these results, a couple years before our own paper, my old Austrian colleague, Mario Blumthaler, reported cloud attenuations in the Bavarian Alps that were significantly smaller in the UV than the visible. I guess it’s a question of what is ‘significant’. In any case, we should (definitely, if money allows!) extend our study using multi-year data to see if the same conclusions apply for other locations and to see if anything has changed. The study described used only a single year of data from a single site more than thirty years ago.

I hope this helps to clarify the cloud situation 😊. Apologies if it’s just served to muddy the waters further …

You can see the effects of those early morning clouds. There was also a small cloud near midday. Note the small increases in radiation either side of that decrease.

I recently gave a rule of thumb for cloud transmission. It’s about 50 percent for visible light. But there’s a huge amount of variability. Their presence sometimes even results in short term increases in radiation, as illustrated above. I call these ‘Cloud Enhancement’ events (where it’s the radiation that’s enhanced, not the clouds)! They happen when the cloud doesn’t block direct sunlight, and light reflected off its surface - which can be much brighter than the background skylight - is redirected towards you. At the same time, nearby places blocked by the same cloud get less radiation.

But, what is their effect on UV? I’m interested in this question right now because my friend Jerry Burke and I are putting the finishing touches on a new smartphone app that predicts UV risks and benefits. In previous versions, we’ve used forecasts of cloud effects, but they unfortunately sometimes get it wrong. The last thing you want to do is to tell people it’s safe when it isn’t. With our newer apps, we rely on the user to specify the sky conditions. The most important factor is whether or not the sun is obscured by clouds. If it isn’t, the average UV over periods longer than a few minutes will be essentially the same as for clear sky conditions.

The effects of clouds on UV radiation aren’t quite as large as for visible radiation because at UVB wavelengths at least half of the total you receive is from sunlight that’s already been scattered from air molecules. Their effects at Lauder can be seen from the histogram below, which shows the relative frequency of occurrence of different values of what I’ve called the ‘Cloud Modification Factor’ (CMF). By that, I mean the factor by which clouds have changed the clear sky radiation. The CMF is, therefore, just the ratio of the measured UV divided by its calculated value for clear skies.

About 50,000 spectral measurements for sun elevations greater than 45 degrees were used to generate the curves. Results are shown for three different wavelength ranges: UVA, UVB, and the UV index (UVI) which is a measure of sunburning radiation.

That reducing tail at CMFs greater than one corresponds to measurements with cloud enhancements. For the three weightings, the peak values ranged from1.27 to 1.39. For about 1 percent of the sun-clear spectra the presence of nearby clouds that didn’t block the direct beam increased the UV by about 25 percent. Fortunately, these enhancements are usually less than a few minutes in duration and are often sandwiched between periods of lower UV when the same clouds obscure direct sunlight.

The sharp peak near CMF = 1 corresponds to clear-sky conditions. The broader secondary peak to its left corresponds to cloudy conditions when the sun is obscured. It reaches a maximum near CMF = 0.5 but extends over a range from about 0.2 to 0.7. Similar patterns are seen at other sites, though at sunny sites, the peak near CMF =1 dominates.

In the app, we’ll be hedging our bets for cloud effects. Although, the mean is 0.5, we don’t want to use that. As you can see from the plot, it would sometimes seriously underestimate the UVI. We therefore provide the user with three easy choices: the default ‘Clear Sky’, or ‘Sun Obscured, partly cloudy’, or ‘Heavy Overcast’. The corresponding cloud modification factors we’ll be applying are 1.0, 0.7, and 0.3. I stress that these factors are only approximate (and somewhat arbitrary). If in doubt, use the default Clear Sky forecast.

Like I said, clouds are the bane of my life. In the real world, their effects on radiation are difficult to predict.

Oops. I lost the last version of this which didn’t publish properly. Fingers crossed this time

This one’s for you, Lars Olof.😊 It’s just a quick off-topic rant. To add a modicum of relevance, I should mention that we’ll soon be releasing an update to our GlobalUV app. Today it tells me that the peak UVI in Stockholm, Oslo, Helsinki and Copenhagen will be around 4 tomorrow, about twice as high as here in southern New Zealand as we head into winter.

Why is it that all the good policies in the world seem to come from Scandinavia? It starts with a much flatter wage structure and higher taxes which get progressively higher for higher income workers. That narrows the obscene gap between rich and poor. It helps that Norway has plenty of oil, but all the Scandinavian counties also seem to keep their assets, unlike us in New Zealand. We always sell our family jewels so others can take the profits. It’s no coincidence that Scandinavians have the highest happiness index in the world, with Finland leading the pack.

It’s a while since I’ve been to that part of the world, but from this vantage point here and now it seems a bit like New Zealand used to be in the good old days before the mantras of rogernomics and the half-baked ‘user pays’ that kicked off in the 1980s. One consequence is the exorbitant price I now pay for my electricity. Even though I live only 10 km from one of New Zealand’s largest hydroelectric power stations, hardly anybody in the country pays more for their power than me. The reasons are the same as in my last rants on the topic 6 years ago, or 6 years before that. The only thing that’s changed is the price. My peak rates are now over $0.61 per unit.

The citizens of Norway must be laughing into their lattes as the price of oil rises. With their much-better-insulated homes and wages 50 to 70 percent higher than ours, the unit price of their electricity is about half that.

As the plot below shows they’ve also made some great decisions about their transport fleet, even though they’re awash with oil. look at that impressive uptake of electric cars over the last two decades!

As recently as 2010, all their new cars were entirely powered by fossil fuels, with about 35 percent from diesel and the remaining 65 percent from petrol. The contribution from all other sources was less than 1 percent. Then, things began to change. A couple of years later, battery electric cars had just started to get a measurable share of the market. Just a decade later their share was about 30 percent, and by the end of last year about 98 percent of new purchases were electrics, with the share of diesel and petrol being less than 2 percent.

That’s what good policy drives. Good on you Norway. Denmark isn’t far behind with 82 percent of new cars now being battery electric vehicles. Finland and Sweden aren’t bad either, with about 47 and 41 percent respectively, just ahead of China.

Contrast that with the sad situation here in New Zealand. After a promising start by the previous government, all their good work was undone by spiteful moves by the incoming coalition in October 2023. The threat of climate change wasn’t enough for the morons to help with the transition. Like many of their environmental decisions, they’ve moved in the opposite direction. Currently, only 2 percent of our fleet is electric. With tRump’s crisis in Iran, we’re now paying the price at the pump for that negligence. The public have started to vote. Since this last oil shock, more than10 percent of new purchases are electric.

Even that energy crisis hasn’t persuaded our government to get their heads out of their sand buckets. There’s now an obvious need to get away from fossil fuels for security reasons, but the decision makers can’t shake the lead out of their gumboots. Last week they loudly proclaimed that 13 percent of the fleet will be electric by 2035, rising to 46 percent by 2045. That’s pathetic.

I suspect the problem children are minor coalition parties led by David Seymour and Winston Peters. They’re also hell-bent on using more imported fossil fuels instead of our abundantly cheaper and more secure renewables to prop up our energy grid. A grid that will need to supply much more soon to keep the proposed new data centres running.

The stampede to mine gold is a case in point for the crazy times we live in here. It seems that everybody wants to dig it up and put it in a vault somewhere else. The only reason for doing that is for a small number of people to make a lot of money. My guess is that the (mostly overseas) profiteers already have plenty.

We’re not as bad as tRump’s America but are heading down the wrong path with a lack of principled leadership. Hopefully we’ll get a new government soon so we can reclaim our place in the world as champions for the environment.

In the meantime, bring on those Scandinavian policies - including their tax regimes!

Aaah. that feels much better … If you got this far, well done and thank you. Feel free to add a ‘like’.

What a relief on 8 April NZ time, when tRump’s most recent ‘deadline’ for the ‘decimation’ of Iran (proving he doesn’t know the historical meaning of the word) passed without escalation. In my humble opinion, the best off-ramp for all concerned, would be to implement the 25th amendment. Please do it soon, America. The new administration could explain that the late-night rantings of the nutter no longer apply, and could we all please start again?

But there was also a sad moment later that night for me while I was watching an episode of Steve Colbert’s Late Show on YouTube. After his usual rant about bad-news tRump, Colbert moved on to the unrelated good-news story of the night: the Artemis circumnavigation of the moon. He included an image of planet earth compared with the famous Blue Marble image taken from a similar vantage point during the Apollo mission in 1972. The comparison was a stark reminder of a rosier past. I found this version of the pair at Google Images (with apologies for that spelling of Artemis).
Note the updated figure caption.

Sadly, it’s not hard to spot the difference between the new image and the one from 54 years earlier. Even Colbert alluded to it.

Everything’s got a lot murkier. Even the cloud-tops are no longer white. It all looks kind-of smudged now. At left, it’s easy to discern the continental outline of Africa, but for the life of me I can’t identify the large shrouded land mass in the recent image. It’s too dirty to see properly. Our previously pristine planet of 1972 now looks vibrant in its youth compared with the decaying image at right. We have tRump and his ilk to thank for the change. Like him, our planet seems far past its prime.

While fossil fuel companies were being enriched at the rate of nearly US$3 billion dollars per day for every day over those intervening years, the waste products were building up in the atmosphere leaving others to pay for the consequences. Surely some of that 50 trillion dollars could have been spent on their clean-up! Instead, their focus has been enriching themselves and lobbying politicians like tRump to ensure their desecration of our atmosphere can continue unabated. That, of course, is what the Iran conflict is all about. It’s the oil. As usual, the aggressor just wants to get his hands on some of it for himself.

Since 1972 the world’s population has more than doubled. Our emissions of carbon dioxide have also more than doubled, and its concentration in the atmosphere has increased by 30 percent. Concentrations of other man-made greenhouse gases have also increased: methane by nearly 30 percent and nitrous oxide by 14 percent. These aren’t small changes, and we’re already seeing their consequences.

Temperatures have increased by more than 1.1°C over those 5 decades and the resultant costs of storm damage (and our insurance premiums) have increased by a factor of ten. As I noted above, it’s hard to tell what countries underlie that brown atmosphere, but I guess the pollution is possibly worse than usual if it overlies the current spheres of tRump-sponsored conflict. That brown smudging may be from associated smoke and aerosols from fires. It could also be from another indirect product of combustion, nitrogen dioxide - which is one of the few gases that absorb at visible wavelengths. It absorbs blue light leaving a preponderance of red and yellows (brown). The brown tinge in the image at right is very reminiscent of the colour of a calibration cell of nitrogen dioxide we made in the early 1980s as part of an effort to measure its concentration in the lower atmosphere. We used those absorptions of blue light to measure some of its lowest concentrations ever seen. At times it was less than 20 parts per trillion in our pristine part of the world. I’ve no doubt that the baseline would have moved up over the intervening years. But by how much? Just a few percent? Or perhaps a factor of 2, or even more? Could that be playing a part in the browning of our planet as viewed from space? Probably not. Even though the reflected sunlight we see has traversed earth’s atmosphere twice, the sun would have to be close to the horizon to cause that much absorption. Still, it’s time we repeated those old measurements.

On the subject of the ‘war’ in Iran, although the human and economic cost is high, tRump may have inadvertently done the world a huge favour and his friends in the fossil fuel industry a disfavour by accelerating the move from fossil fuels. In New Zealand, for example, the sale of electric cars has nearly quadrupled since the beginning of the war. That will help. We’ll have to transition anyway because our supplies of fossil fuels will run out in a few decades, but any steps to accelerate the phase-out are much appreciated and much needed.

You won’t find tRump talking about the difference between those images. He would have us believe that climate change is a hoax and that he and his cronies can continue to use the atmosphere as a garbage dump without consequences. To help push that narrative, he’s doing his best to neuter agencies like NASA, NOAA and NCAR who tell the truth about our atmosphere.

But, there’s glimmer of hope. There’s already been some push-back from within his own party. The camera doesn’t lie, and with this pair of images, previous supporters may begin to see through his lies. Richard Nixon was the president in December 1972 when the image was captured from Apollo, and just 20 months after that moment of triumph he was gone. Can we hope for the same with the even more corrupt incumbent? Fingers crossed.

We’re still holding our breaths on the ‘war’ front where tRump has painted himself into a corner. Some light relief on the subject was provided by local cartoonist Shaun Yeo, whose work is worth a look. His offering below appeared in our local newspaper on 20 March. Nothing’s changed since then.

Sorry about the hiatus. It’s been over two weeks since my last offering. I must be running out of ammo! If you have any suggestions, please send them in.

About 20 years ago, I led work on a widely-cited paper which concluded the peak UV values in New Zealand were 30 to 40 percent greater than at corresponding latitudes in the northern hemisphere. It was based on data from a newly established network of UV sensors that had been set up by the US Department of Agriculture (USDA). Because of our interest in UV, they’d decided to include Lauder New Zealand (45°S) as a reference site.

The comparison covered the first 3 years of operation of the network: 2002, 2003, and 2004. In addition to the data from Lauder, it used data from the North American sites within the range 40 to 50 degrees of latitude, as shown at the left below (blue and red symbols). On the right you can see the much higher peak UV values at Lauder. They are much higher even than those 5-degrees closer to the equator at 40°N, and are more comparable with those from those at lower latitudes in the mountains of Colorado and Utah.

I’m working on an update that confirms these large differences and also shows similarly large differences in daily doses of UV at latitude 45°S compared with at 45°N.

In the meantime, let’s look at what causes them …

I start by looking more closely at ozone differences between those two latitudes to see what part they play in the story, and how they might have changed over time. The ozone data I used were monthly means from a paper published by my NASA colleague Jay Herman in 2023. Because the peak UV occurs close to the summer solstice, I focused just on data from December at 45°S and June at 45°N. In each case, the data represent the average over a 10-degree range from latitude 40 to 50 degrees.

A comparison - over the period since satellite measurements began to when that paper was published - is shown below. Early in the period, you can see the ozone decline that occurred because of the release of CFCs into the atmosphere. The eruption of Mt Pinatubo in 1991 exacerbated the decline in the northern hemisphere. But by late last century, the effects of controls on the release of CFCs required under the Montreal Protocol were already starting to kick in. As a result, ozone levels at these latitudes have been pretty stable since 1995, as shown by the dashed lines. But, as you can see, the average amount at latitude 45°N (red line) is much greater than at 45°S (blue line), and by an amount greater than any reductions over the period of ozone depletion. The second panel below shows the ratio of ozone at 45°N compared with that at 45°S. On average, in summer there’s about 13 percent more in the north for this latitude band.

The importance to peak UVI of that difference in ozone is illustrated below, where I used a simple formula developed by my colleague, Sasha Madronich, to calculate the UVI for pristine conditions. The upper panel shows the peak noon UVIs near the summer solstice calculated for the ozone amounts shown above. Again, the lower panel shows the ratios (but this time inverted to the ratio of 45°S/45°N). Notice the similarity of that year-to-year pattern.

Most of North/South difference is due to the lower ozone amounts in the south, but the calculations also include corrections for the closer Sun-Earth separation in December compared with June, and for the northern-hemisphere’s higher percentage of ozone in the lower atmosphere where it more efficiently blocks UV.

The bottom line from these effects is that in clean air we should expect the peak UVI at 45°S to be about 27 percent more that at 45°N. But, with 90 percent of the world’s population in the northern hemisphere, it’s not surprising that the air there is much dirtier. Those larger aerosol optical depths apparently contribute significantly to the observed higher UV intensities at Lauder compared with sites at corresponding latitudes in the USA.

I’ll finish writing all this stuff up properly one day … 😊.

The question is: when is noon not noon?

which reminds me of the old chestnut riddle:
Q. When is a duck not a duck?
A. When one of its legs is both the same.

The answer (to the noon question) is, hardly ever. In fact, in the New Zealand summer high noon is never at clock noon. Our clocks are referenced to longitude 180°, which is well to the East of New Zealand. By the time you add in the one hour of summer daylight saving, high noon occurs much later.

The next question is: how much later?

It’s an important question because the time of highest sun elevation (or lowest solar zenith angle) is what determines the time of maximum UV radiation, which follows because the UV will be most intense when its path-length through the atmosphere is minimised.

With a bit of prompting after a few false starts, my free AI assistant (Claude, Sonnet 4.6) helped produce a graph which at first sight, it looked plausible but unfortunately was not correct. The range of offsets in time looked about right, but there were 4 cycles over the year instead of the two I expected. In case you’re interested, I’ve included a suitably annotated version of Claude’s plot at the end of this post. When I questioned Claude further, it came back with the not-very-helpful message “Claude can make mistakes. Please double-check responses”.

After double-checking it out with my colleague, Ben Liley, and finding it was indeed in error, I persuaded him to provide his correct version of the plot. He’d just made a start on it when I decided to try asking the same question of the paid version of Microsoft’s Copilot AI tool, that’s available through their Edge web-browser on my work laptop. Ben’s a highly proficient programmer, but sadly for him (he’s a mere mortal after all), the AI tool won the race.

The answer to the question is, a LOT later!!

Here’s Copilot’s result (which agreed perfectly with Ben’s). It shows the time of highest sun (high noon - or solar noon) each day for Invercargill (168.4°E, blue) and Gisborne (170.8°E, red), the Westernmost and Easternmost cities in New Zealand.

The time scale used is the same as on your clock or watch, so the dominating features are the 1-hour steps around day 90 and day 270 which correspond to the switches from New Zealand Daylight-savings Time (NZDT) at the end of summer to New Zealand Standard Time (NZST), and the switch back the summer time in spring.

The largest offsets of solar noon compared with clock-time noon occur at Invercargill around day 50 (around 20 February), when overhead sun occurs just after 2 pm: a full two hours after noon on your clock or watch. On no summer day is it earlier than 1:30 pm at this site.

It takes 24 hours for Earth to spin though 360 degrees (or 15 degrees/hour), so solar noon at Gisborne, our Easternmost city, occurs about 40 minutes earlier (corresponding to that 12.4-degree difference in longitude). There, the corresponding solar-noon times range from12:50 to 1:20 pm in summer. In New Zealand, it’s only at these Easternmost areas for a short period just before the switch to summer time (around day 270) that the time of highest sun ever corresponds to clock noon.

The smoother variations in those curves arise because of the ellipticity of Earth’s orbit about the Sun and the tilt of its spin axis. They’re called the ‘equation of time’ variations, with have a range of about ±16 minutes.

Gisborne’s latitude is 38.7°S compared with Invercargill’s 46.4°S. That difference means that its highest daily sun elevation are 7.7 degrees higher than Invercargill’s. The shorter resulting path-length of the Sun’s rays though the atmosphere means that, as well as the UVI peaking earlier in Gisborne, its maximum for clear-sky conditions is also greater (despite its slightly higher ozone amount that day), as shown below.

On 1 January, the calculated peak UVI at Gisborne was 11 percent higher, reaching 11.4 at 1:10 pm compared with Invercargill’s 10.3 at 1:50 pm . Interestingly though, because of Invercargill’s longer daylight hours, its daily dose of UV was only 6 percent less (with 63 SED of sunburning UV at Invercargill compared with 67 SED at Gisborne).

As I’ve harped on about before, the danger time in the NZ summer is therefore from around 11:30 am to 4 pm. That’s from two hours before solar noon in the East to two hours after solar noon in the West.

Just for completeness, in the version below I’ve included New Zealand’s Northernmost city, Kaitaia (35°S, 170°E). Since Invercargill is also the Southernmost city (in addition to being the Westernmost), this shows the full range of clear-sky variability of UVI throughout the country this day.

The time offsets between clock noon and solar noon are relatively large for New Zealand, but they’re hardly ever zero anywhere in the world.

Bonus (or should that be bogus?) material

Claude AI’s wrong answer. Luckily the error make little difference in practical terms.

Seeing this error made me suspicious about Claude’s previous calculation of the elevation angle of the centre of the Milky Way. That too seemed to have too many turning points for my liking (which is why I ran it past a couple of experts before posting). Here’s the update using Copilot. I haven’t plotted exactly the same quantities, but the overall result looks consistent with the earlier plots. The main point was that southern mid-latitudes - including New Zealand - are the best place to view the Milky Way. Patterns are essentially the same each year.

I said I’d check. I’m glad I did. It was an interesting exercise.

I started by looking at the variability of daily ozone over Lauder (45°S) during the two-month December-January period for each summer since satellite data became available. The version used here isn’t the raw ozone data from the satellite community. It includes corrections using a method developed by my old Lauder colleague, Greg Bodeker, to bring satellite data into line with ground-based measurements.

The results are shown below. With nearly fifty years of daily measurements, it’s difficult to see the variability for any particular year. To give you the idea, I’ve highlighted it in red and blue for the first and last summer respectively. You can immediately see that the ozone amounts last summer were generally lower than they were in in 1979, despite the slight uptick in the most recent summer, which will be shown soon.

That mass of lines shows that it’s highly variable too. In addition to its day-to-day fluctuations, the ozone also tends to decrease over this period. That’s just part of its annual cycle which has a maximum in spring and a minimum in autumn. Linear approximations to those seasonal decreases are shown for the first and last summer by the red and blue dashed lines. Strictly, if you want to quantify the day-to-day variability, you should look at residual values after removal of those trends.

Annual summaries of the year-to-year changes for all these parameters are shown below for Lauder. The upper panel shows the minimum, maximum, and mean ozone amounts each summer. There’s significant year-to-year variability, but no real evidence of any long-term changes apart from a decrease of about 10 percent over the period 1980 to 2000. Since then it’s been remarkably flat, and may even be showing signs of improvement.

Over the entire period, the mean summer ozone amount was about 288 DU (grey dashed line). The spread of ozone each summer can be deduced simply from the difference between the red and black curves. Typically, that range is around 100 DU, but that includes the most extreme outlier days. A better measure of spread is what’s called the ‘Standard Deviation’ (which, for the budding statisticians among you, is the square root of the variance about the mean, as defined here). Normally, about 70 percent of all data falls within 1 standard deviation of the mean.

The middle panel above shows the seasonal trends in ozone that I mentioned when discussing the first figure. Typically, the ozone decreases by about 7 DU per month (i.e., about -2.3 percent per month) over this period, but in some years the trend is close to zero, or twice that, showing that the day-to-day variability is sometimes large enough to swamp the seasonal trend.

The lowermost panel shows the variability of ozone, this time expressed as standard deviations. The black line shows the raw standard deviations. Typically, they are around 20 DU, but, as noted above these are too large because some of the variability is due to the seasonal decrease. A better estimate of the variability is to calculate the standard deviation after the linear trends have been removed, as shown by the red curve. The average value is about 15 DU, which means that for 70 percent of the time, the day-to-day variations in ozone are less than 5 percent of the mean. Therefore, for most days (but not all, as discussed here), a climatological estimate of the ozone will be sufficient. There’s a lot of year-to-year variability in these standard deviations, but no real evidence of any long-term changes.

The slightly higher ozone amounts expected due to its correlation with the 11-year cycle in solar output (with peaks around 1980, 1991, 2002, 2014, and 2025) are barely discernable. The strongest El Niños (in 1982-83, 1997-98 and 2015-16) don’t seem to affect things much either, and nor do any long-term changes due to climate change. In fact, the year-to-year variability looks quite random.

I’ve checked a few other sites (some shown below), and found that the patterns shown above are quite typical for other mid-latitude sites.

The short answer to the question is, therefore, no. It’s not getting more variable.

That’s good news. If ozone were becoming more variable, there would be more days now with higher than usual ozone and - more importantly - more days with lower ozone. Those low-ozone days would continue to be associated with settled sunny weather (i.e., during anti-cyclones with high atmospheric pressure) when our risk of skin damage is already highest.

It was an interesting exercise as far as spatial differences rather than temporal differences are concerned. Below, I compare the numbers for New Zealand with those from Bordeaux, France, which at 45°N is the closest city to Lauder’s antipodes.

As you can see, the year-to-year variability in ozone is quite similar, as are the seasonal slopes and the standard deviations shown in the lower panels. But, if you closely compare the top panels for each site, you can see that there’s a lot more ozone over Bordeaux. It’s mean is 332 DU compared with only 288 DU at Lauder, a relative difference about 15 percent, which corresponds to an increase in UVI at Lauder of about 21 percent. There’s an additional increase of about 7 percent due to the elliptical nature of of Earth’s orbit about the Sun (with closest approach in the NZ summer). Based on those two factors, if the air clarity were comparable, the UVI in summer at Lauder would be about 28 percent more than at Bordeaux. But the air quality isn’t comparable. It’s probably much cleaner at Lauder. Because of Europe’s more polluted air, actual differences are probably at least 30 percent, as I plan to verify in an upcoming post.

And it’s not just Bordeaux that had much higher ozone amounts than Lauder. I checked the corresponding ozone amounts for 5 other cities close to latitude 45°N. Four of them (Milan, Portland, Montreal and Ottawa) were very similar to Bordeaux, ranging from 332 DU to 338 DU, and one (Minneapolis St Paul) wasn’t much less, at 320 DU.

Finally, to show its not just a fluke for Lauder, I also compared the results for our capital city, Wellington (41°S), with those from Thessaloniki which is at a similar northern latitude in Greece.

First, Wellington

Now, Thessaloniki, at the same latitude in the northern hemisphere

As you can see, the the NH-SH asymmetry is similar . The ozone above Thessaloniki is 324 DU, compared with 288 DU at Wellington, a difference of more than 12 percent. Interestingly, the N-S differences in these corrected ozone seem to be nearly twice as large as mean latitudinal differences as in the ozone fields we’d used in a previous study.

A closer look is warranted ….

As part of another project, I happened to have an ozone climatology showing the monthly mean ozone amounts as a function of latitude for the 10-year period from 2003 to 2012. A comparison of mean ozone amounts for the 3 summer months (June, July, and August in the northern hemisphere; and December, January and February in the Southern Hemisphere) are summarised below.

Over this 10 period, ozone differences between the Northern and Southern Hemispheres are largest in the latitude range 40-50 degrees. They are slightly smaller than for the sites selected the 5 decade period in the plots above, but the main cause of the larger inter-site discrepancies is the larger ozone difference for the summer equinox month (Jun/Dec) - when the highest UVIs usually occur - compared with the other summer months.

So you can mull the shorter-term differences, I’m including the plot (below) for Bordeaux that’s equivalent to the first shown above. The scale of the y-axis is the same but ozone values at Boulder are confined to the upper half of the plot, and frequently exceed the axis bounds, especially early in the summer.

These ozone differences - along with our cleaner air and seasonal differences in Sun-Earth separation - contribute to peak levels of sunburning UV in New Zealand being 30 to 40 percent greater than at comparable northern latitudes.

In the New Zealand summer, UV levels are dangerous for fair-skinned people.

Be aware. Take protection.

The only good news is that its not getting worse. Thanks for the question Michael 😊

The day with the most UV over that period was Saturday December 13, about one week short of the summer solstice. The plot below shows in detail how the UVI varied over the course of that day.

As you can see, there were a few scattered clouds in the morning between 9 and 11 am, but symmetry shows that they had only a slight effect on the UVI. It looks like the sky was entirely cloud-free from 11 am until sunset. The peak occurred shortly before 2 pm New Zealand summer time (which includes an hour of ‘daylight saving’). Daily maxima occur up to 40 minutes earlier in more eastern parts of the country.

The maximum UVI this day was about 11.5, but it’s the dose of sunburning UV that matters. A high UVI for an instant won’t hurt you. It’s the accumulation over time that does. That dose is usually expressed in terms of the number of Standard Erythemal Doses (shortened to SEDs), which means the standard dose for the- reddening of the skin by sunlight.

The number of SEDs to cause the first signs of skin reddening - which indicates skin damage - depends on your skin type. For the fairest skins it might take only 2 SEDs, while for the darkest it will be more than 15 SEDs. If you’re somewhere in between but think you’d look better with a tan, the number is probably less than 3 SEDs.

The plot below shows a close-up of how the skin damaging dose of UV dose built up over that single day. As you can see, it far exceeds that threshold for safe exposure. Over the full day, there’s 30 times the safe exposure limit for the most sensitive skin type and at least 20 times too much for most would-be sunbathers. Clearly, protection would have be needed to avoid skin damage.

There are some simple useful take-home messages there.

The periods with the most rapid increase in UV dose are the ones to be avoided. That’s the period centred near 2 pm this day. For those UVI levels - near 12 - you can get about 10 SEDs per hour. That’s five skin-reddening doses each hour for fair skin, or one dose every 12 minutes! A good rule of thumb is to avoid exposure within a couple of hours either side of that peak time. In other words, avoid sun exposure between noon and 4 pm. If you did that on this day, the total exposure would be no more than 30 SED. Still a dangerous amount, but a lot better than the 68 SED for the entire day. At the other end of the scale, it shows that you’re pretty safe before 8 am and after 8 pm. Not many get sunburnt at night! (or during twilight, or even if the sun is less than thirty degrees above the horizon).

In retrospect, I’m glad I chose not to play golf that day. The Saturday competition times at our local golf club - with tee-offs starting around noon - couldn’t be worse as far as UV exposure is concerned. In any case, I’m much happier playing any outdoor sport in the early mornings, or late evenings. To keep the dose below 10 SEDs that day, I’d have to be finished by 11 am, or start after 4 pm. At other times of the year, those limits can be relaxed. In either case, there are always plenty of daylight hours available here to complete the round under low UV conditions. My personal preference is to play in the mornings. Even though the changing dew conditions may make putting more difficult, there’s usually less wind and it’s much cooler too.

And I still wear a hat and apply sunscreen.

‘What about me?’, you might say. That’s just one place and time.

The answer’s at your fingertips. You can get personalised advice to plan your outdoor activities wherever you live for any day of the year by consulting one of our free smartphone apps:

• GlobalUV (which works anywhere in the world), or

• UVNZ (for New Zealand only).

Go get one!

Extra, in case you’re still craving for more …

The above plots were for the day this summer with the highest daily dose of skin damaging. Below are the corresponding plots for the day with the highest peak UVI, 1 Jan 2026. As expected, it’s also close to the summer solstice; this time 10 days after it.

Although the peak UVI was nearly 20 percent greater this day, the daily dose was 25 percent less. By noon, the doses were similar both days, but it clouded over in the afternoon this day. Those highest values are greater than for clear skies because they include radiation scattered from the edges of clouds. Note that all the data shown here are 10-minute means, rather than the hourly means plotted in a previous post. The longer averaging period makes a big difference to the UVI, but little difference to the daily dose.

As a general rule, the largest daily doses of UV tend to occur on cloudless days when the sun is never obscured. On partly cloudy days the sun tends to be obscured some of the time, and if that happens around noon, about half the UV is blocked.

Forecasts of the diurnal variability of UVI for the current day at any site, including behavioural advice tailored to the user, are now available from smartphone apps like GlobalUV or UVNZ. The iPhone weather app also includes estimates of its coarse daily variability. Forecasts of the peak UVI at any site are also available from the internet (e.g., The Netherlands TEMIS site). And, if can put up with the intrusive advertisements from Amazon’s Alexa and ask her nicely, she sometimes - but not always - tells you what the peak UVI will be at your location. Secretly though, I think she’s more interested in the advertising revenue.

But how well do these forecasts of UVI perform? They aren’t automatically logged, so each day over the peak summer UV period I manually recorded the peak UVI predicted from several different sources. For the first half of the period (to 4 Jan), I recorded data for Alexandra. I showed a comparison between different forecast options here. But, to check their absolute accuracy, I needed to compare their results with measurements. There are none at Alexandra, so from 4 January I also included forecasts for the nearby Lauder UV measurement site (about 30 km to the northeast), whenever possible. Unfortunately, data from Amazon’s Alexa were available to me only from Alexandra, and not from Lauder. They were also missing for two short periods when I was away from home.

Results for clear skies

There was an extremely high correlation in the clear-sky forecasts from the two sites over the latter period, as you can see in the below. The two curve almost over-lie.

That high correlation enabled me to use Alexandra data to fill in clear-sky forecasts at Lauder for the earlier period prior to 4 Jan (as shown for the first 3 days in Jan above), where the Alexandra data were scaled up 2.5 percent to account for differences in latitude and altitude. Below, I’ve compared the peak UVI measurements (dashed curve) with three clear-sky predictions.

As expected, there’s good agreement between the upper envelope of measurements and the clear sky forecasts. But, as also expected, on cloudy days the measured UVI is much lower than predicted. Sometimes they’re lower by more than a factor of 5 (in cases where the cloud transmission is less than 20 percent. Predictions with the TEMIS model, which is tuned for European conditions, are lower than the UVNZ model which is tuned to the cleaner New Zealand conditions. Errors of a similar magnitude can be seen in the GlobalUV model, which uses a climatology of ozone values that therefore doesn’t include effects of day-to day differences in ozone.

Results for all skies

The correlation between sites was still quite strong, as shown below for all sky conditions, but differences between sites are larger than for the clear skies forecasts shown earlier. Larger differences are of expected here because of local differences cloud patterns.

I therefore have less confidence in the extrapolation of Alexandra data to Lauder for the earlier period, and have included only the Amazon forecasts for the period prior to 4 Jan 2026 in the comparison below.

For the latter period, the agreement between measurements and forecasts at Lauder is encouragingly better for these all-sky forecast than for the clearsky forecasts. The periods of cloudier weather with lower UVIs are generally well captured in the forecasts. Unfortunately though, the actual peak UVI sometimes exceed the forecast by more than a factor of two. The Amazon forecasts seem to have a persistently low bias, especially on the cloudy days, which is particularly troublesome. In the earlier period prior to 4 January, the clear sky bias seems smaller, but the cloudy sky bias seems worse. To me this suggests that the effects of aerosols have been seriously overestimated in their model. Although we’re not quite comparing apples with apples there, the previous plot does show that differences in cloud effects between the two nearby locations is rarely more than one UVI unit, whereas measured values in one case (17 December) differed from the Amazon forecast by more than six UVI units - corresponding to an error greater than a factor of four.

In case you want more details, here are a few statistics from those graphs. The first line in table shows the mean UVI for each method compared with the measurements, while the second shows the corresponding correlation coefficient.

On average over this period, the mean of the peak daily UV measured was 7.1 (white). All three clear sky models (blue) overestimate the measured values, while both cloudy models (brown) underestimate them. Correlation coefficients are better for the latter, but are never greater than 0.74, which isn’t very impressive. It means that at best only half of the expected variability due to clouds has been captured.

The take-home messages are:

1. Clear sky predictions agree well with measurements and can be significantly improved by including day-to-day variabilities in ozone, which as I showed last time can be large. The GlobalUV app will be updated to allow both that option.

2. Inclusion of loud effects improves the correlation, but (dangerously), the UVI is sometimes seriously underestimated. If sun-avoidance behaviour is influenced by such predictions, there could be serious health effects.

3. With the current (limited) skill in forecasting cloud effects, it may therefore be preferable to provide the public with predictions for clear skies, and with estimates of the mean cloud reduction factor for cloudy conditions, ranging from typically 0.5 for partly cloudy conditions, to around 0.3 for heavily overcast conditions, as is currently done in GlobalUV app.

Updated 5 Feb 2025 to better explain the plots of daily doses.

My colleague James Renwick’s recent article in The Conversation explains why the weather has been so unsettled in New Zealand this summer, causing a trail of destruction including several deaths. It’s due to a range of factors including Climate Change - so more of the same can be expected in the future.

The Central Otago region - where I live - was less affected, but still abnormally unsettled. Despite that, there was still plenty of sunburning UV about. Plenty to cause a lifetime of skin damage and even death. Apparently, more than half of New Zealand’s GenZeders (aged less than 30) deliberately go out on high UV days to get a tan. It’s so stupid.

On those sunniest days, the UVI can be more than 12, and the daily dose of sunburning UV sometimes exceeds 70 ‘Standard Erythemal Doses’ (SED), which is thirty times more than enough to cause skin damage for people who think they need to tan. To put it another way, even if a sunscreen of SPF 30 were correctly applied all day long, there’d still be measurable skin damage.

The hourly-averaged measured UVI at Lauder over the two-month (62 day) period from 1 December 2025 to 31 January 2026 is shown below. The year roll-over is marked by highest value this summer, UVI = 12, which occurred around noon on the first day of 2026. At that time, skin damage could occur in less than 15 minutes. On this partly cloudy day, the peak 10-minute value reached 13.5, which is higher than the expected clear-sky value. But that wasn’t sustained over the full hour and was balanced by lower values when clouds blocked the sun. There were fewer than 10 perfectly cloudless days over the entire period - recognizable by the symmetrical diurnal pattern. There were also weekly intrusions of bad weather with a lot of cloudy days, but only three when the UVI didn’t exceed the ‘Alert’ level of 3. It was potentially damaging just about every day. The peak hourly UVI exceeded the ‘Extreme’ level (UVI > 10.5) on about a third of of the days.

But it’s the UV dose over minutes or hours that matters. The corresponding daily doses of sunburning UV - as it accumulates each day - are shown below. As you can see, they range from 11 to 68 SED per day, with an an average over the period of 45 SED. Note that the peak dose on (13 December) was much greater than on the day with the peak UVI (1 January), which was cloudier in the afternoon. For fair skinned people, just 2 SED is enough to cause skin damage (as evidenced by skin reddening), so even the lowest UV day (10 January), the dose was more than enough to cause sunburn.

The accumulated sunburning dose over the entire period reached a staggering total of 2775 SED. That’s enough for nearly 1400 harmful doses.

The New Zealand summer school holiday period runs from mid-December to the end of Jan. It varies a bit from school to school, but for my grand-kids I think it was from Saturday 13 December 2025 ‘til Sun 1 Feb 2026. Over that 7-week summer period, the total UV dose at Lauder (45S) was 2200 SED, enough for more than 1000 bouts of skin damage for fair-skinned people should they choose to spend most of the daylight hours outdoors without protection.

Although the summer was cooler and less settled than usual, in terms of its ability to burn it could still be described as a summer from hell. Those youthful sun worshippers would be well advised to take note. Although they might feel bulletproof at their stage of life, things look rather different from my end of the life cycle. I personally know many who’ve had melanomas removed; and knew several who’ve died from it. Off the top of my head the latter include my old school friend Paul Ackerley, my old boss and colleague Gordon Keys, and a couple of rellies Mike Hattie (my brother-in-law) and Wally Lynch (my wife’s uncle).

UV levels are highest when the sun is highest in the sky, so they show a strong seasonal variation, with a maximum near the summer solstice. At the winter solstice, UV levels in the north of New Zealand levels are less than 20 percent of their summer peaks, while in the south they are less than 10 percent of their summer peaks. On sunny days, the UVI is strongly peaked to the solar noon period, which is between 1 and 2 pm in NZ summer time, when it tends to be about 20 percent greater in the far north of the country compared with the far south. For similar sun elevation angles, the UVI is typically larger in autumn when ozone amounts are smaller than average compared with in spring when ozone amounts are larger.

Some knowledge of the potential damage from UV is needed. And I don’t think so-called ‘Influencers’ will be the ones to provide it. You can’t directly sense UV either. Although it tends to be more intense on warm days, it’s not proportional to temperature. The UV apps that my friend Jerry Burke and I developed were designed to fill that void in knowledge for the great unwashed. You can also get UV forecasts from the internet (e.g., TEMIS), Amazon’s Alexa, or from the iPhone weather app.

But how well do they perform? To answer that, I manually recorded the peak UVI predicted for Lauder from several different sources over the summer holiday period and compared results with the measurements shown above. I’ll show the results next time.

A great short movie. Fantastic plot(s) with vibrant colour appeal, though it could be argued that the character development leaves a bit to be desired. Oh, and there’s no audio.
(I hope it works OK for you. It’s a bit large, and I got a (non-fatal) error message because no audio is included. Fingers crossed 😊).

In recent posts, I’ve talked about the variability in the amounts of protective ozone above us in the atmosphere, and how an ozone climatology fails to capture the full picture. I persuaded my colleague, Ben Liley to help illustrate the problem. He did it by making this movie loop, which shows the patterns of change in ozone over a full calendar year. The particular year he selected is 2021, but it’s quite typical for recent years. It uses ozone values derived from satellite, but adjusted to match ground-based data using a procedure developed at Lauder by another colleague, Greg Bodeker.

Here Ben has provided us with two views of the world: one from outer space directly above the international dateline, just to the east the New Zealand (at right), and one from its antipodes near Spain (at left). The amount of ozone overhead is illustrated by the colours, ranging from purple for very low ozone amounts through to red for the highest amounts.

It’s neat. I’m sure you’ll love it! There’s at lot in it too. To get the full benefit, you’ll need to play the 15-second loop a few times. Try focussing on different locations each time to compare the seasonal and day-to-day variability. You can stop it at any point if you want to see the pattern for any particular time of year.

The grey area that appears the winter months in both hemispheres and reaches its maximum at the winter solstice, shows the extent of the polar night when no ozone data are available from this satellite instrument. It uses sunlight that’s been back-scattered from the atmosphere to make the measurements.

The first thing you probably notice is the fast-moving patterns of change. These are driven by changing weather patterns in the atmosphere below. Ozone amounts in the stratosphere above are inversely proportional to barometric pressure at the earth’s surface, which means that there tends to be less ozone overhead lower during settled sunny condition. A double whammy for UV risk. These changes are most prominent at higher latitudes, where they propagate eastwards because of the strong ‘westerly’ winds at the altitudes near 20 km where most of the ozone can be found. Typically, the wind speeds are around 50 km/hour, but they can be much faster in the polar regions. This is the cause of the day-to-day (inverse) variability in UVI as well. If you look closely, you can see occasional discontinuities near the international dateline. These are inevitable in displays of this kind when there are large ozone changes over the 24-hour period of data assimilation.

Notice the strong seasonal pattern too, with a maximum in spring and a minimum in autumn outside the tropics. The notable exception to that rule is the development of the Antarctic Ozone Hole, which was quite typical that year. It’s that dark blue region that starts developing in August (around the time sunlight returns to the south Pole), reaching its maximum size in early October, and recovering in December. As usual, it’s nicely confined to high latitudes largely devoid of people, far to the south of New Zealand, where I live. In fact, when its at its deepest, ozone amounts here are much more than average.

Overall, you can clearly see that there’s less ozone on tropics. There’s also a notable asymmetry between the hemispheres. In the northern hemisphere, the reddish colours in spring show that ozone levels sometimes exceed 500 DU, but in the south the colour scale only occasionally goes beyond yellow (about 450 DU).

It’s a great example of every picture being ‘worth a thousand words’. I know I’m well below that here, but I could go on (and on) … 😊

Thanks Ben. Nice work!

One day I’ll try to persuade Ben to make an equivalent movie for noon UVI, using the same method as my friend Jerry Burke did for the UV Apps (as described here). I doubt if he’ll able to resist the challenge! As he’s shown before (in a static display), if the calculation includes topography, you see that the world’s highest values, sometimes exceeding UVI = 20 occurring over the South American ltiplano region.

Welcome to 2026 everybody. As you’ll see, I was busy in the last days of 2025 😊.

In my last stack here at UV & You, I trumpeted the return of our GlobalUV app, with a couple of caveats. Today we’ll take a closer look at those caveats. It’s highly appropriate at this time of year because of the high risk of excessive UV exposure during the summer holidays

The new version uses a climatology of monthly mean ozone amounts, as depicted below, where data for all longitudes have been averaged over the decade from 2003 to 2012. Don’t take too much notice of the polar winter periods at latitudes greater than 70 degrees. They look a bit ragged because no ozone data are available from the satellite sensors used, and they’ve been replaced by zeros. But the overall picture between 65N and 65S, which includes 99.95 percent of the world’s population, is a good representation of the mean monthly values. Notice that there tends to be less in the tropics, and at mid-latitudes there’s less in autumn and more in spring. There’s also more ozone in the northern hemisphere than the south.

But how good are these for estimating the UVI at your place on a day-to-day basis?

To help answer that question, I compared the mean ozone column amounts (for all longitudes within 2.5 degrees latitude from 45S with measured amounts over Lauder New Zealand (45S, 170E). The results are shown below, where the red points are the daily data for Lauder, and the red line is the mean averaged over a three month period. The blue line is the monthly mean for all latitudes at 45S.

Concentrating first on the solid lines you can see that, although any long-term trends in ozone have been small over this period (due to the success of the Montreal Protocol on Protection of the Ozone Layer), there is a large seasonal pattern and also a large seemingly random inter-annual variability in the size of those seasonal swings. This is most noticeable in the spring maxima, which were smallest in 2006 and 2015 and largest in 2009 and 2017. A similar year-to-year variability is seen in both the satellite data (blue), and averaged ground-based data (red).

Any climatology will inevitably fail to take such differences into account. Fortunately though, during the peak summer period (near the grid lines in the plot), when the UVI is most intense, these year-to-year differences are much smaller. Notice that even with the longer averaging period in the data from Lauder (3 months compared with one month for the satellite data), its seasonal variability is larger. This is not unexpected given the extensive averaging in the satellite data (encompassing 180 degrees of longitude, and 5 degrees of latitude in this case). A study we undertook in the early 1990s showed that the mean ozone at Lauder and its seasonal swings were quite typical for the latitude. But something might have changed in the last 3 decades, so we’ll be having a closer look.

Of greater concern are the large day-to-day differences in ozone, illustrated by the scatter of red points. These lead to large day-to-day differences in the UVI, but again, the variability is smaller in the most dangerous summer period.

These differences are nicely illustrated on one of the regularly-updated plots on the NIWA/ESNZ UV/ozone web site. A recent example is shown below. Ground-based ozone data (red and black symbols) are updated daily, but satellite data (solid red line) are updated less frequently. There’s good agreement between the two ground-based measurements, which also agree well with satellite data (up to the last update in August). The plot shows that ozone amounts tended to be larger than average (blue line) this year, especially in spring.

Again though, the variability is much smaller in summer (December-January). Despite its shortcomings, the climatology therefore provides a useful assessment of UV risk, especially during the summer when the risk is greatest. It’s also a great educational tool.

There are several other sources of UV information out there too. I took a note of the predicted daily maxima from a few of them over the last couple of weeks of 2026 for my hometown, Alexandra (45.25S, 169.4E). Over that period, the weather was rather unsettled. Measurements from nearby Lauder, about 30 km to the northeast, showed there was only one cloudless day over the entire period. That was on 22 December, as shown by the smooth curve that day in the plot below. On other days, there were rapid changes between measurements as clouds passed overhead, sometimes obscuring the sun. Much of that variability - but not all- disappeared when the averaging period was increased from 10-minutes (blue line) to 60-minutes (red line).

In the next graph (below), I compare the peak measured UVI at Lauder with corresponding predictions for Alexandra from several sources. These include smartphone apps (UVNZ, GlobalUV), an on-line tool (‘Temis’), and weather forecasts from Apple (iPhone) and Amazon (Alexa). Some, as labelled, include cloud effects, while others don’t.

As you can see, there’s reasonable agreement in the peak UVI for clear-skies (solid lines), which also agree with the data from Lauder on the clearest day (22 December). Over this short solstice period, any variability in solar noon UVI forecasts from the GlobalUV app are small. They boringly remain clamped to UVI=10.4. In contrast, the clear-sky values from the UVNZ app, which (uses ozone forecasts rather than the climatology) swing about that mean value in response to day-to-day changes in ozone. Despite that, the health messaging is not significantly different. The Dutch-hosted ‘TEMIS’ web-based product shows a similar day-to-day variability to UVNZ, but with a bias to lower values probably because it over-estimates the attenuation by atmospheric aerosols in this pristine part of the world. For New Zealand, UVNZ will be the most accurate because earlier versions of it have been shown to agree well with measurements at Lauder for clear sky conditions (e.g., see here and here). Notably though, the same papers show that cloud effects are poorly characterised.

For the cloudy-sky forecasts in the figure above (dashed lines), there’s a lot of variability between apps. And all of them seriously underestimate the UVI on the clearest day, 22 December. The Amazon prediction that day is (dangerously) less than 40 percent of the true value. On other days, the cloudy sky forecasts sometimes differ from each other by more than a factor of three. As I mentioned in my previous post, in view of such uncertainties it’s probably safer to use a clear-sky forecast, and use your own judgment to modify that by a factor of 0.5 for sun-obscured partly cloudy conditions, or 0.3 for heavy overcast conditions. While Lauder is in close enough proximity for clear sky comparisons, it’s probably too far away to evaluate cloudy-sky forecasts at Alexandra because cloud effects can be highly localised.

Either way, it appears that you need to take the forecast of cloud effects with a large grain of salt. Unless there’s a heavy overcast (in which case you don’t need to worry about UV), my advice would be to use the clear-sky forecast and a bit of common sense.

For antipodeans, consider this your early Christmas present for sun safety in the southern summer holidays. For others, find just how hard it will be to sustain your levels of vitamin D in the northern winter.

Some of you may have noticed that our tried and trusted UV smartphone app, GlobalUV, has been misbehaving lately. That’s because some of the data for it came via NIWA, and the IT sections within the (now superseded) organisation imposed new rules that made it too hard for old codgers like us to deal with.

The good news is that my friend Jerry Burke has made an update which will be immune to future shenanigans like that. The new version makes use of a monthly climatology of mean ozone amounts as a function of latitude, rather than using the most current satellite data. As a result, the app works even if you don’t have internet access, such as when flying or hiking. Because of its global coverage, the previous version required the download of several megabytes of data at each use. This version needs none because all the necessary data is stored within the app. Significantly for these crazy times, that also makes it tRump-proof. Using the climatology is justified because global patterns of variability of ozone amounts have been remarkably stable for the last two decades.

The downside is that day-to-day variations are now no longer included. For most of the globe (latitudes less than 30 degrees from the equator) this doesn’t matter because day to day variations in ozone are small. They become increasingly large as you head to higher latitudes. In Polar regions, uncertainties in UVI arising from this will be large (especially in the Antarctic Spring), but for most inhabited areas, they will usually be less than ten percent.

The other thing we’ve done differently is in the way we allow for cloud effects. In the previous version we used satellite estimates, but because the amounts of data were so large, we downloaded only the noon values and presumed they would apply throughout the day. In reality (of course), cloud effects are highly variable and generally too complex to allow their effects to be calculated accurately, even if cloud fields are downloaded multiple times per day. For example, it’s currently impossible to know from satellite data products if a small puff of cloud is currently obscuring the sun at your precise location. Your eye is a much better indicator of that.

The most reliable indicators of cloud effects on UV - namely, whether the sun is unobscured or obscured by clouds or whether the sky is completely overcast - are easily identifiable by the user of the app. We therefore provide the user with those three choices. The cloud corrections are only approximate but are probably more reliable in many cases than our previous method. We assume that whenever the sun isn’t obscured by clouds, the UVI will be similar to the clear-sky value. And, following a study that I reported on previously, we apply corrections to the clear-sky forecast by a factor of 0.5 for situations when the sun-obscured, or by a factor of 0.3 for fully overcast conditions. Like the previous app (and all others), this doesn’t allow for UV enhancements due to reflections from clouds under partly cloudy conditions, but in practice these aren’t so important because the enhancements are balanced soon afterward by reductions when those clouds move to obscure the sun.

The app is freely available for both iPhone and Android smartphones. It’s pretty nifty in its interactions with the user. In use, the app defaults to the last-used location, but you can change that at any time by touching the map screen, or by selecting from a list of locations specified by country and city.

As previously, users can input their skin type and clothing attire, to enable the app to provide appropriate health messaging, including estimates of the exposure times both for skin damage and for vitamin D sufficiency at any time throughout the day. In the example below, the number below the dot for the present time indicates that at that time, an exposure period of just 2 minutes is sufficient for daily vitamin D sufficiency (for a person of skin type 2 wearing shorts and tee shirt). The time-of-day defaults to the present time, but that can niftily be altered by on the Graph screen (below) by sliding your finger across the screen. This tool is therefore a useful aid to schedule outdoor activities and optimise health outcomes.

It’s also a handy educational aid. In the ‘Geeks’ screen, see the example shown below, you can investigate seasonal changes at for any site in the world by changing the month. There’s also an option to see by how much the UVI changes with ozone amount. The multi-coloured coloured blob on the lower map shows the current geographical pattern of UVI, which moves to the left as the day progresses.

This cut down version of the original app is not quite as accurate as the previous version because it doesn’t include longitudinal variability in ozone. It’s good enough for practical purposes, but, as you can see from the satellite-derived image below, these can lead to significant errors, especially at higher latitudes. The pattern shown in the image below typically moves from west to east from day to day, leading to the daily variability mentioned above.

The image thus implies that the daily variability in ozone tends to be smaller in summer and at lower latitudes. In this example, the expected variation over the following few days above southern New Zealand (where I live) would be in the range 325 to 375 DU, leading to maximum deviations in the UVI from the mean of about 8 percent. But errors can occasionally be much larger. For example, the lowest climatology ozone amount for latitude 45S is just over 290 DU, whereas the lowest ever recorded at Lauder (45S) is about 220 DU. On that extreme day, the true value would have exceeded the app’s estimate by 40 percent! In summer, an error of that magnitude would be very troublesome, but fortunately, this extreme outlier occurred well outside the peak summer period, with a true UVI that day of only ~5. It does, however, show that there’s a price to be paid for tRump-proofing the product. One day we may revert to using current satellite-derived ozone fields, especially if we find there’s are potential problems due to clear-sky biases in ozone amounts. But it would probably be from agencies in Europe rather than the USA unless things change there. If we did re-travel that route, we’d need a web site to host the ozone data, which would be accessed by users of the app. Please let me know if you’d be able to help with that. 😊 I’ll let you know here if and when any further updates become available.

In the meantime, download the (free) GlobalUV app and get to know it. Use it to know when the risk is greatest so you can plan safe outdoor activities. Explore its educational benefits and learn - for your own personal circumstances - how many minutes of sun exposure should be needed anywhere and at any time to maintain healthy blood-levels of vitamin D without damaging your skin. Go to the Geek’s screen to see how the risk varies by month, or by ozone amount. The answer will always be at your fingertips.

Finally, I have to apologise for including the incorrect final image in my previous post. The corrected version can be seen in the update here.

Updated 9 Dec 2:33 pm after noticing that I’d used the wrong image for the ozone minimum.

A recent Ozone News update from the United Nations (UNEP) made a lot of noise about the ozone hole over Antarctica this year being much smaller than usual. That’s good news, indeed. But it’s a bit of a beat-up. In the longer term, a good-news outcome is not yet assured.

I agree that its encouraging news. But, when they (carefully) say the hole is “among the smallest in decades”, that doesn’t really give the right impression. It’s nowhere near the smallest. As the historical plots below from the NASA ‘ozone watch’ website show, it’s not really even an outlier. It ranked as the fifth smallest in the last 33 years (since 1992), and about the 13th smallest on record since satellite measurements became available in 1979. And we shouldn’t necessarily expect a continuation of the trend next spring because it’s imposed on a lot of year-to-year variability. Similar variabilities and trends are seen in both the maximum area of the ozone hole (upper plot), and the minimum ozone amount recorded (lower plot).

The images illustrate how fickle the recovery is. We’re clearly on a slow path towards recovery, but more ups and downs are expected along the way, due to differences in Antarctic wind patterns. In September this year, unusual air movements in the stratosphere about 25 km above Antarctica resulted in a premature warming, which contributed to the lower ozone loss.

The report quotes my NASA colleague Paul Newman as saying

“We’re seeing ozone holes form later, break up earlier, and trend smaller than in previous decades, …. (but we) still have a long way to go before we recover to 1980s levels.”

I’m happy with the second part of what he said. Those plots above show that we’re only at the beginning of the path to a full recovery.

But I disagree with his assertion that the ozone hole is breaking up earlier. While that’s true for this particular year, the opposite is true more generally, as has been clearly illustrated in recent Ozone Hole Reports compiled by my colleague, Paul Krummel, at Australia’s CSIRO (using similar data provided by NASA). Unfortunately, these emailed reports don’t appear to be available online, so I’ve taken the liberty of copying a couple of images from his last one. They show the development (and decline) of the ozone hole this year and the previous 5 years, compared with the long term means and percentiles.

The first, below, shows that early in its development, this year’s ozone hole (black line) was quite normal, until a rapid decrease in mid-September, from which it never quite recovered. It then collapsed much more rapidly than usual in November. By mid-November its area this year was in the bottom 10 percent of all previous ozone holes. But this early collapse is unusual compared with recent years. In the latter part of the ozone hole period (especially in December), the area of the ozone hole has usually been within the top 10 percent in recent years (coloured lines).

Similarly, although this year’s ozone minima were less severe after mid-September, especially from November onwards, in other recent years the low values persisted longer than the average, as shown below.

By both criteria - hole area and ozone minimum - the breakup of this year’s ozone hole was indeed unusually early. But, by the same two criteria, the breakup of recent ozone holes has been later than usual (coloured curves) in the previous 5 years. In 2020, the ozone hole was the most persistent on record, with 2021 and 2023 being not far behind. Although the ozone holes in 2022 and 2024 weren’t particularly strong, they were relatively persistent compared with the long-term mean.

This changing pattern over the decades is also borne out by the ‘skew’ in the statistical distributions seen the plots. For ‘normal’ distribution patterns, which are symmetric about the mean, the mean values - as represented by the white curves - would overlie the centre of the dark grey area. While this is approximately true for the ozone minima (lower plot), for the ozone hole areas (upper plot) the mean value in December is far larger than the expected limits.

From the viewpoint of UV radiation that persistence into the month of December is important because of the higher sun elevations at that time of year. As I’ve noted before, that can already lead to extremely high levels of UV over Antarctica, where the peak UVI already sometimes exceeds that in San Diego USA. Any extension of the period of ozone depletion is therefore concerning.

That’s not likely to be helped by attempts to curb climate change through the stratospheric aerosol injections (SAI) that I mentioned in my last post. High latitudes are preferred for such injections. If injected at heights lower than the stratosphere (ozone layer), they’d be quickly rained out, and only at polar latitudes is the stratosphere low enough to be ‘easily’ accessible for seeding from available aircraft. A recent paper calculates that the aerosol injections required to maintain temperatures no more than 1.5C above pre-industrial levels could reduce the already low amounts of ozone over Antarctica by another 30 DU (which would correspond to a decrease of around 30 percent if it persisted during the period of the springtime ozone hole).

The UNEP Report notes that “Scientists attribute this encouraging trend to the ongoing decline of ozone-depleting chemicals in the atmosphere - a direct result of global action under the Montreal Protocol”. That good news needs to be sung from the rafters. If we could do the same to curb Climate Change by controlling the emission of greenhouse gases, there would be no need for those pesky aerosol injections.

Thanks everybody for the nice comments on my Hijacked! post. Sorry, its a bit more prosaic (but happily, shorter!) this time. The title isn’t very informative, but I just couldn’t resist that deliciously alluring alliteration. 😊 The title slide below is closer to the mark, but my point of view is a precautionary note to one of the suggested fixes.

Climate guru James Hansen’s excellent and provocative recent 30-minute talk in Helsinki is well worth watching. But my concern about some aspects of it motivated me to contact him for clarification.

Among other things, he was advocating the use of a suggested climate modification tool called Stratospheric Aerosol Injection (SAI). That could help cool the planet in the same way that aerosols from the Pinatubo eruption did in the 1990s. He argued for injecting the aerosols into the stratosphere at high southern latitudes (far too close to where I live for comfort), rather than over the more densely populated latitudes of the northern hemisphere, where the people who caused most of the problem live. As I’ve noted before, I’d prefer not to have aerosols introduced anywhere in the atmosphere, and especially not in my back yard, because of their effect on light quality.

I thought it might have been a case of NIMBY for him too. He assured me that it wasn’t. Rather, it would best to introduce them at those high latitudes because there it can have the largest effect. He made the point that one of the most consequential effects of climate change is its interaction with deep ocean circulation. Changes in it could accelerate ice melt in Antarctica, which will (eventually) lead to disastrous increases in sea level. We’re already in the situation where water temperatures near the surface have risen enough to initiate the melting of Antarctic ice. That melting is accelerated by a feedback mechanism where the surface water becomes less salty and dense. It therefore acts as a ‘lid’ that inhibits its normal replacement with water with deep cold water from the depths of ocean, and instead diverts away to lower latitudes (where it warms) or greater depths (where it cools). The conceptual diagram below, which I’ve filched from Hansen’s slides, helps the explanation. The patterns of movement shown by the arrows are part of the normal oceanic circulation that has a turnover timescale of around a century. Plenty of time for the water to cool off substantially at depth. The upwelling colder water (green arrows at left) is diverted away from the Antarctic ice.

Modelling studies show that aerosol injection at those high latitudes will cause much larger temperature decreases in that part of the ocean compared with injections over other latitude ranges, as illustrated in the next couple of slides from his talk.

The slide below maps the 4 possible aerosol emission scenarios (E1 to E4) that he considered. Note that in each case the same total amount of aerosol is delivered to the stratosphere.

Finally, the slide below shows that the most effective strategy for cooling the critically important part of ocean is to inject over the Southern Ocean and Antarctica (scenario E3).

I’m not an oceanographer, so can’t really comment on the validity of the simulation. I defer to Hansen and the oceanographic community on that. But I still don’t like the idea of it, quite apart from the aesthetic reasons mentioned previously.

From my atmospheric science perspective, the aerosols injected there may have unexpected consequences. Past experience has shown that Antarctica is highly sensitive to change, partly because the air over the continent can remain isolated from the rest of the world for long periods, and partly because of the long period of darkness in winter which inhibit photo-chemistry and lead to very cold atmospheric temperatures.

One possibly harmful consequence is that the aerosols will provide additional surfaces for the rapid (heterogeneous) chemistry that already leads to rapid ozone depletion there on the surface of ice crystals each spring. This should be an issue only for the next few decades while chlorine levels remain elevated, and the presence of these aerosols may even accelerate ozone recovery because any solar energy they absorb will lead to heating which inhibits the formation of the ice crystals. On the other hand, the presence of the man-made aerosols means that ozone depletion will continue quite happily without those ice crystals.

Over the longer term, there’s the issue of what happens to those aerosols. Given the atmospheric circulation patterns, I’d expect the injected aerosols to be entrained towards the south pole before descending to the lower atmosphere where they would become condensation nuclei for the formation of precipitation (mainly snow). But, instead of the snow being pure water, it would contain these aerosol particles. Thus, instead of forming a perfectly white highly reflective surface, it would be slightly darker. That in turn would lead to increased absorption of sunlight, which would (perhaps disastrously) accelerate the rates of both warming and the ongoing melting of snow and ice. Especially since the time scales for ice-melt are huge, so the aerosol injections would have to continue for centuries. That heating effect would be less problematic if the aerosols were completely non-absorbing (and therefore interact with radiation only by scattering it to different angles). But such an ideal is difficult to achieve.

A final message on the topic to world leaders. Please don’t meddle with our atmosphere. There’s been enough meddling with it already, and its the only one we’ve got. If you must (also) meddle, do it in your own back yard, not mine. Fix the problem at its source, by rejecting the burning of fossil fuels.

On that subject …..

If it wasn’t so serious it would laughable that after all the pomp and circumstance of Thirty COP Climate summits, the parties still have their heads buried firmly in the sand. They can’t even agree that we have to phase out fossil fuels. No direct references to them are even allowed. What planet are they on? Unfortunately in reality, it’s our only one.

Nature had a good-cop bad-COP summary finishing with this quote “If you look at it scientifically, the conclusion must be nothing less than a failure. Because we know that now — 10 years after we signed the Paris Agreement — we’re still increasing emissions, we’re still seeing an acceleration of temperature rise and we’re seeing suffering among hundreds of millions of people across the world already, today”.

Sorry about the hiatus …. This isn’t about UV either, but having recently farewelled a couple of an old Lauder colleagues, and having reached the ripe old age of 75 myself, I see a small window of opportunity to recall noteworthy changes that have occurred over my career before they’re completely fogged over by the passage of time. I was originally going to focus just on a plane hijack, but while putting that into context, it became more of a nostalgia trip through a bygone era that’s itself being hijacked. It’s a story I could have included in ‘Saving our Skins’. It’s now 5 years since I published the book, so I’ve decided to make it an update Epilogue to the on-line version.😊

It’s exactly 46 years today since my career as an atmospheric scientist began, at the Lauder Auroral Station. That was in 1979, about two weeks before Air New Zealand’s Mount Erebus disaster in Antarctica. In those days, Lauder was still part of the New Zealand Government’s Department of Scientific and Industrial Research (DSIR). It stayed that way until 1992 when, after a government reorganisation of science, it became part of the newly formed National Institute for Water & Atmospheric Research (NIWA). Overall, that has worked out OK. At least Lauder has survived. But, last year the Government decided (have they nothing better to do?) that we needed yet another change. From July this year, NIWA would no longer exist. We at Lauder became part of Earth Sciences New Zealand (acronym yet to be decided), a conglomeration of NIWA, MetService, and Geological & Nuclear Sciences. I must admit that I’m not fully on board with the Earth Sciences name. To most, that would appear to sideline anything above ground level. But I guess we should give it a chance. It will need all the help it can get in this tRumpian world that we live in now (hopefully post-tRumpian in the near future). In a previous post, I reported that my AI aide, Claude, had suggested ‘Geo-Spheres’ as a name. That’s slightly better, but it doesn’t really cut it for me. If we can’t think of a suitable word, perhaps the grouping itself is an unholy alliance?

It’s been a bad year for Lauder in other ways too. Before we even had time to change our email addresses, we lost the service of three of our major players over the years. Andrew Matthews, who was a key player in the 1990s died after a long illness in July. He was the driving force in the establishment of Lauder as one of 5 primary atmospheric measurement sites within the international NDACC network. And our old colleague, Gordon Keys, who retired from NIWA in 1999 after 35 years at Lauder, passed away in October, about a month short of his 95^th birthday. Like Andrew, he was instrumental in putting Lauder on the map. To make matters worse, my old colleague and mentor, Paul Johnston, has been unable to contribute (much as he’d still like to) after only semi-successful spinal surgery. All three contributed to Lauder becoming the best-instrumented NDACC site in the southern hemisphere, and possibly the best in the world. In the picture below from 1991, I’m the only one still standing.

Lauder continues to evolve, with an increasing emphasis on climate change and renewable energy, but to me the place has been hijacked by change.

Outside of work there have been big changes too, both at Lauder and in the wider region. When our family arrived, there was a daily rail passenger-service connecting Dunedin and Cromwell, wending its way past Middlemarch, Ranfurly, Lauder, Alexandra, and Clyde. Now only the first 60 km of the line is used by occasional tourist-train excursions through the Taieri Gorge. The remaining 150 km section from Middlemarch to Clyde is now the popular Otago Central Rail Trail cycling path shown below.

But back in 1979, staff and families at Lauder could still catch the rail-car at the local station for a day’s shopping in Dunedin, the undisputed provincial centre of Otago and the 4^th largest city in the land. It’s still about the same size and a beautiful city but has now been relegated to 7^th on that list.

The village at Lauder had a population of less than 50, but if you included the surrounding farming area of Becks, there were perhaps a few hundred. There were about a dozen kids enrolled at Lauder Railway School, which included some from the research station and outlying farms. The social life centred around sports clubs (especially tennis and rugby), and the local pubs. The latter were the only option to buy meals. On the rare occasions we ate out, it was much more likely to be at a friend’s place; more often than not, barbecued meat washed down with Speights beer. It was a simple life. There were only two TV channels, and the conversations at morning tea – after of course dealing with the perennial topic of weather - would often revolve around what we’d watched on TV the night before.

There were shopping options in Lauder and nearby Omakau, but the road to Alexandra had recently been improved so it was only a half-hour drive down the valley. Most at Lauder, including my wife Louise, would head down there each week for their main food supplies, or to shop at Arthur Barnett’s department store.

Alexandra was easily the largest town in Central Otago, with a permanent population of around 4500. The numbers would swell by a factor of three or four in the summer school holidays, when the camping grounds were awash with holidaymakers from coastal Otago and Southland. They came for the hot dry weather and filled their time with swimming and picnicking in the Manuherikea River or at the nearby outdoor swimming pool. It was hugely popular in winter too, when busloads came for skating on natural ice at the nearby Lower Manorburn Dam. And each spring the Alexandra Blossom Festival, with its associated activities, would usher in the transition between winter and summer. A big step for the town was in 1984, when the Ryan family moved their old Four-Square grocery business across Centennial Avenue to open their New World Supermarket, the first in Central Otago.

That was long before the emergence of Queenstown as the tourist mecca we see today. Alexandra was the hub of the Central Otago region. In the early 1980s, Queenstown and Wanaka were still delightful little villages with less than 1000 inhabitants, few shops, and no supermarkets. The locals there came to Alexandra for their shopping and commercial activities. Only Queenstown’s airport - at Frankton - was much busier than Alexandra’s. Since 1969 there had been scheduled daily flights linking Alexandra and Christchurch, operated by Mount Cook Airlines using Britten-Norman Islander 9-seaters. Queenstown, with its nearby ski fields, was served by the same company, but using much larger Hawker Siddeley 748 turboprop aircraft (pictured below), that seated about 50 passengers.

With the beginning of work on the Clyde hydro-electric dam, business was brisk for the airlines. But, as the roads improved, passenger patronage would fall off for the rail link. Six months after our arrival in Lauder the Clyde to Cromwell section of the line was uplifted because it and the tortuously winding road that crossed it multiple times in the narrow gorge were slated to be under metres of water when the dam was filled. The main purpose of the railway became for transporting materials for the construction of the dam from Dunedin to the new rail-head at Clyde.

By 1991, the dam had been completed and the old slow road through the Cromwell gorge had been replaced with the modern highway we see today that halved the travel time between Alexandra and Cromwell. That was the death knell for the railway. The last train service to Clyde was in the same year.

That marked the beginning of a new era for Alexandra too. It was no longer the ‘destination of choice’. It became a stop-over on the way through to the lakes and ski-fields further west. With the improved road access, they began to boom.

Cromwell too would eventually surpass Alexandra. It became the base for ‘Tradies’ for both Queenstown and Wanaka, where property values were becoming prohibitively expensive. It’s strategic location also played a part, being only 60 km from both tourist centres and half that from Alexandra, as well as being on the direct route to Queenstown from both Christchurch and Dunedin.

The wine industry was still in its infancy. Most of the countryside was devoted to sheep farming, with pockets of stone and pip fruit orchards in the sheltered lower valleys. Paddocks surrounding the road through the Gibbston Valley en route to Queenstown were filled with grazing livestock rather than the manicured grapevines we see today.

After our return from Oxford in 1986, Louise persuaded me that we should purchase our own house in Alexandra so the kids could walk to the local school and be involved in after-school activities. Along with other staff, I would commute the 35 km to Lauder in the work van. We’d often pass van-loads of commuters going the opposite direction. After the financial slump in 1987, many of the farmers near Lauder needed to supplement their income, and organised themselves into vans to commute down to the Clyde Dam construction site. It was a hive of industry.

The Lauder research station was part of the DSIR’s Physics and Engineering Laboratory, with its head office in Gracefield Avenue, Petone, just north of Wellington. We’d occasionally have to travel there. That involved a plane trip to Christchurch, then on by jet to Wellington airport, followed by a bus to the station to catch a train to the Hutt Valley. It was shortly after 1987 that Louise dropped me off at the Alexandra airport for what turned out to be a memorable trip. I can’t recall the reason for that particular visit. It may have been to discuss the ill-fated instrument I was working on to measure nitric oxide in the atmosphere. The hardware was being built for me in their workshop. But I vividly remember the return journey.

My flights back home were supposed to retrace the northward route a couple of days earlier, but when I arrived in Christchurch I was advised that ‘due to operational matters’ there would be no flight to Alexandra with Mount Cook airlines. Instead, I was reticketed on their flight to Queenstown and advised that I would be put on the next flight from there to Alexandra. By this time, work on the Clyde Dam was winding down and passenger numbers to Alexandra were on the decline, so I knew perfectly well that the real reason for the change was that there weren’t enough passengers to make it worthwhile. I also knew that there were no connecting flights from Queenstown to Alexandra, so they’d have to find some other way to get me home. There were no bus services either, so it would probably mean they’d be forking out for an expensive 100 km taxi fare. That set me thinking. There had to be a better way. It was all in the timing.

If you look at a map of the South Island of New Zealand, you’ll see that Alexandra isn’t too far off the direct flight path from Christchurch to Queenstown. This particular flight wouldn’t land at Mt Cook, but the flight captain said he would take a detour so the Japanese tourists on board could catch a glimpse of its snowy peak. If a detour like that was possible, surely I thought, it might also be worth having a go at hijacking the plane to take me directly home. By then, hijacking had become quite a popular sport in the big bad world outside New Zealand. But we were still fairly relaxed in our part of the world. I decided that our resumption of travel after viewing Mt Cook would provide the perfect window of opportunity. As soon as we’d left it in our wake, I called the flight attendant. I casually asked her to speak to the captain to see if he could divert to Alexandra on the way. She was interested, but concerned about the disruption to the 50 or so tourists on board. I said they’d love the diversion, explaining that it would give them a great chance to look down on the Clyde Dam construction site as they ascended from Alexandra. I persuaded her pass on my request to the pilot. He was a big ruddy-faced bloke in his thirties, about my own age, with a shock of black hair. He must have thought it a good idea too. He checked with the powers that be, and I’ll live for the rest of my days with the image of his face beaming back to me from the cockpit with his thumb up, fist-pumping the air. He was as excited as me.

About 45 minutes later we landed at Alexandra. It was the early evening, and I didn’t really expect anybody to be there to meet us. I hadn’t got as far as worrying about how I’d travel the 6 km from the airport back to town. This was long before the arrival of cell phones. Surprisingly though, there was not one, but two vehicles parked alongside the runway. One was the white ‘Commer’ van from a local travel agent. The airline must have thought to radio ahead to arrange transport back to town for me. The other, amazingly, was the car belonging to my colleague Paul Johnston and his wife Carol. Louise must have had another appointment, and had arranged for Carol to pick me up. Coincidentally, the flight I’d hijacked had arrived in Alexandra at the same time as the scheduled one that had been cancelled, so they weren’t at all surprised to see me there. I can’t remember which ride-offer I took up, but no doubt the other person was a bit grumpy. Other than that small fly in the ointment, not many can say that they’ve had such hijacking success! I don’t think the outcome would have been so happy in today’s world. I’d love to meet that happy pilot properly and chew the event over with him. My guess is that was a highlight of his flying career. It certainly was for me.

Sadly, a whole way of life has now been hijacked. Scheduled flights to Alexandra ceased shortly after and have never resumed, and in 1991 Mount Cook Airlines was gobbled up by Air New Zealand. We now have to drive to Frankton airport, near Queenstown. The International Airport there is now the 2^nd busiest in the South Island – after Christchurch. The population of the Queenstown area is approaching ten times Alexandra’s. Instead of Queenstown people coming to Alexandra to shop, it’s now the other way round. With increased traffic densities, we now allow two hours for the journey that used to take less than an hour. It’s not just the transport infrastructure that’s under stress. What was once a pristine part of the world is quickly being degraded, with sewage and wastewater treatment facilities being unable to keep up with demand, contaminated water is often discharged into the Kawarau River that feeds the Clutha river, which runs past my hometown. That’s further stressed throughout the region by nitrates in the run-off water from dairy farms, and will be further under stress if a proposed new gold mine at Bendigo (near Cromwell) receives the fast-track consent it’s currently seeking. Air-quality too is being degraded. I’m especially concerned about the impacts on the Lauder clean-air measurement site. With the rapid developments in Queenstown, Wanaka, and Cromwell, all to its west, polluted air will inevitably encroach during the prevailing westerlies. In bygone times, the only departures from pristine air occurred when smoke from bushfires – sometimes from as far afield as Australia – occasionally encroached. Our clean green reputation is on the line.

The Clyde Dam, which was the catalyst for the demise of Alexandra as the hub for Central Otago, did have one tangible benefit for the town. Since the completion of the Roxburgh Dam 30 km downstream in 1956, silt – much of it from the Shotover River near Queenstown - had been progressively building up and clogging the Clutha River below the town. It was blamed for the devastating floods in 1995 and 1999. Now, since the construction of the Clyde Dam, the silt has become someone else’s problem. The bulk of it is now trapped further upstream.

Although my hometown is different now, it’s still a very nice place to live, especially for retired people like me. Everything I need is within walking distance. While it continues to grow steadily (the current population is about 6000), it’s growth rate has never been out of control like the locations to its west. Long may that last. If so, it’ll do me nicely. But I’d still prefer the old times before we were hijacked by ‘progress’.

I trust this wasn’t too different from my usual fare. At least it wasn’t too technical. Hopefully, I’ll still be around to write another Epilogue to celebrate the 50-year anniversary of my arrival at Lauder. I wonder if we’ll still be part of an organisation inaptly called Earth Sciences? 😊

Share

I hope my last post wasn’t too depressing. After that, I thought you might feel the need for something a bit lighter. This good-news story might just fit the bill.

The report describes how the Montreal Protocol not only ‘put the ozone layer on the path to recovery’, but also how it ‘delayed climate tipping points that would have forced earth further beyond planetary boundaries’. It shows that the earlier action is taken, the more easily the problem is solved (as for ozone). Conversely, the later the action is taken, the more difficult it is to solve (as for global warming caused by CO2).

First the ozone part of the story. We now expect ozone levels to be similar in the second half of the century to they were in the 1970s prior to ozone depletion. By contrast, the maps below show the dramatic decreases that would have occurred if we’d done nothing. As I’ve discussed before, those huge ozone reductions would have led to huge increases in UV radiation and skin cancer.

But that’s not where the good news ends ..

That early action on CFCs was serendipitously great news for Climate change too, because the CFCs that destroy ozone are also greenhouse gases. Continued increases in their concentrations would have added to the effects of increasing CO2.

The black line in the plot below shows the warming effect of Ozone-Depleting-Substances (ODSs, which are mainly CFCs). Because of the Montreal Protocol, the current warming from them is less than 20 percent of what it was in 1990. The green band shows the ever-increasing effect that would have occurred if CFCs had never been suspected as a danger to ozone. After that discovery, many manufacturers voluntarily ceased using CFCs, so production rates declined. But they started to increase again before the Montreal Protocol came into force in 1987. The blue band shows the slower rate of increase that would have then occurred without the Montreal Protocol. Instead of their warming effect decreasing to 20 percent of their peak values, they would have more than doubled without the Montreal Protocol, and quadrupled without that early warning that curbed production rates.

The report points out that over the period between 1955–2005, ODS emissions were responsible for about 30 percent of global warming, implying that by now their effect would be comparable with that from CO2 if no actions had been taken. If ozone warnings had come even a decade later, the climate forcing from ODSs could have equalled or exceeded the forcing of CO2. Global warming would therefore have been twice as severe when compared with a 1950 baseline. Without the Montreal Protocol, temperature increases by the middle of this century could have been a further 1 degree Celsius more than the potential 3C increase that I discussed last time.

But it’s not only about ozone and climate either …

If you restrict that climate change, then you also restrict a plethora of downstream adverse effects for mankind. For example, the nine ‘Planetary Boundaries’ that are shown in the picture below (Most of them are pretty self-explanatory. I’m not sure myself what’s meant there by ‘Novel Entities’, but I’ll worry about that another time). Extensions beyond the safe planetary limit are in orange and red. Note that most of them are already well-beyond the safe range …

Of the remaining three, only the levels of ozone and aerosol are currently ‘under control’. Without meaningful action, ocean acidification will soon surpass its safe limits.

As I’ve said before, the Montreal Protocol has been a huge success. Not only for ozone, UV and skin cancer, but also for climate change and for all those ‘downstream’ effects as well. Good as that may be though, urgent action is still required to bring them all into the green and to keep them there.

The time is now. As was so well put in the forward to the article:
‘Fifty years ago, not enough was known for people to be alarmed about the consequences of up-ending Earth’s natural systems. Fifty years from now, it will be too late to do what is now within our power to change. The next five years may prove to be the sweet spot in the history of humankind.’

Sorry, this was supposed to be just a good-news story.😊 It started off optimistically enough, but turned out to be more than just ‘beer and skittles’. ☹