RealClimate logo

Sweatin’ the Mediterranean Heat

Filed under: — group @ 22 October 2007

Guest Commentary from Figen Mekik

This quote from Drew Shindell (NASA Goddard Institute for Space Studies, New York) hit me very close to home: “Much of the Mediterranean area, North Africa and the Middle East rapidly are becoming drier. If the trend continues as expected, the consequences may be severe in only a couple of decades. These changes could pose significant water resource challenges to large segments of the population” (February, 2007-NASA, Science Daily).

I live in Michigan, but Turkey is my home where I go for vacation on the Med. This year’s drought was especially noteworthy, so I would like to share some of my observations with you, and then explore the links between the North Atlantic Oscillation (NAO), Mediterranean drought and anthropogenic global warming (AGW).

The 10-hour flight from Chicago to Istanbul often inspires passengers to romanticize about Istanbul, both tourists and natives alike. Istanbul is the city of legends, forests, and the Bosphorus. It is an open museum of millennia of history with archeological and cultural remnants surrounded by green lush gardens. It is the place where east meets west; where blue meets green; where the great Mevlâna’s inviting words whisper in the wind “Come, come again, whoever you are, come!”

So you can imagine our collective horror as the plane started circling Istanbul and we saw a dry, desolate, dusty city without even a hint of green anywhere.

The Marmara region of Turkey where Istanbul is located received 34% less precipitation than average this past winter; and the Aegean Region, which includes the city of Izmir, received 43% less precipitation than average since October of 2006. Precipitation this low was unprecedented in these regions in the last three decades. So is this a freak year, you may ask?

This next graph shows observational spring precipitation data between 1950 and 2007 (shown with green dots) for Istanbul and multi-centennial (1635-2000) spring precipitation reconstruction from tree ring data for the NW Black Sea region in Turkey (pink and red dots show a 5-year running average for precipitation data from tree rings) (data). Multi-centennial tree ring precipitation reconstructions for the Marmara region are not available. Although the observational precipitation record for Istanbul is generally lower than the spring precipitation reconstruction for the NW Black Sea region, both observational and reconstructed data follow a similar trend, except for two unusually rainy springs in 1998 and 2000 in Istanbul. Also note in the tree ring precipitation data that although the amount of precipitation has fluctuated throughout centuries, it has not consistently dropped for more than 3-4 years until the 1970’s. Since then there has been a steady decline until 2000 (the red part of the graph).

In the summer of 2007, temperatures rose over 46°C in many parts of Turkey as well as the entire Mediterranean region. This heat combined with aridity is estimated to have cost Turkish farmers ~$3.9 billion. Turkey’s wheat crop dropped by about 15%. The Turkish Aegean region alone suffered from 30% lower harvest yields in cotton, corn and tobacco and a 50% drop in fig production.

By mid-summer, the drought started to affect major cities. Ankara (~4 million), the capital, suffered serious water rationing this summer (two days on, two days off). Car washing and lawn watering were outlawed within city limits. With the unforeseen burst of the main water pipe feeding the metropolis, the whole city was left without running water for an entire week. Hospitals had to be issued groundwater in tankers, and city officials started to debate whether to delay public school openings until mid-October to contain potential spread of disease. By mid-August, Ankara had only 5% of total capacity in its reservoirs and dams.

Tuz Gölü, a large salt lake in central Turkey within the Konya Basin, lost half of its water volume in the last four decades. The Konya Basin itself, which hosts a third of all groundwater reserves in Turkey and is the home to eight wetland bird species on the brink of extinction, lost 1,300,000 acres of wetland and witnessed a water table drop of 1-2 meters, also in the last 40 years.
It looks grim.

But not just for Turkey. Even popular vacation areas around the Aegean were hellish this year. Greece had to declare “state of emergency” at least twice this summer: once for forest fires killing over 60 people, burning half a million acres of land, and costing $1.6 billion; and once for drought on the Cyclades Islands due to water shortages.

Morocco experienced 50% less rainfall than average this year which will likely result in half of last year’s grain harvest. And because feed prices went up as a result of this drought, Moroccan livestock was also seriously affected. Again, is this an unusual year? Actually not. This next figure, which is from Esper et al.’s (2007) paper in Geophysical Research Letters, shows a significant drop in Feb. – June PDSI (which is a standardized measure of surface moisture conditions after Palmer, 1965) since 1980 in Morocco.

Some speculate that the tragic fires in Greece may have been arson, others say better maintenance of water pipes and anticipation of the coming drought early in the previous winter would have prevented Ankara’s water shortage, and better irrigation systems in Morocco would have mitigated agricultural disaster there.

Though it is “debated” in the US, most people in Turkey consider AGW to be a given. This is generally a good attitude, of course, but it opened the door for some government and city officials to simply blame AGW for drought instead of their incompetence in dealing with it. As a result, the Turkish General Directorate of Disaster Affairs started discussing whether AGW should be listed under “natural disasters” in order to provide better risk assessment and adaptation plans, and to prohibit building new structures in “high risk” areas. Even Al Gore came to visit Istanbul this summer to give a talk at a conference called “Climate Change and its Effect on Life.”

So, is AGW to blame for Mediterranean drought? Although bad land use practices and arson share the blame, AGW is most likely the culprit. Here’s how:

The North Atlantic Oscillation is an alternation of air masses between polar and subtropical regions of the North Atlantic. When the NAO index (difference between the normalized sea level pressure anomalies in the Iceland and Azores areas, respectively) is in a positive phase, a low pressure system prevails over Iceland and a high pressure system over the Azores. This causes cooler northern seas, stronger winter storms across the Atlantic Ocean, warm wet winters in northern Europe, and cold and dry winters in Canada and Greenland. However, this also causes less rain and reduced stream flow in southeastern Europe and the Middle East. In general, when NAO is in a positive phase, the Mediterranean region receives less precipitation.

The precipitation pattern in Turkey is well correlated with the phases of the NAO index. For instance Cullen and deMenocal (2000) showed that Euphrates’ spring stream flow varies by about 50% with the NAO index. The NAO index vacillates between negative and positive phases, but its oscillations had a more annual pattern before the 20th century, and since then has become more decadal. More importantly, since the 80’s it has remained in a prolonged positive phase. This is in keeping with the precipitation drop I described for Turkey and Morocco (and the Mediterranean region) in the last several decades.

But do we know that this recent prolonged positive phase in the NAO index is not simply a part of its natural decadal variability? And is this recent positive phase actually related to global warming?

These are tough questions to answer definitively, but it is likely that AGW will continue to keep the NAO index positive because both atmospheric CO2 rise and stratospheric ozone depletion cause a strong polar night vortex. The North Pole is dark and very cold in the winter. This creates a large temperature difference between high latitudes and subtropics. The resulting large pressure contrast forces east-west winds into a stratospheric spiral. And this stratospheric vortex likely causes the NAO to prefer a positive phase. This was first shown by Shindell and colleagues in 1999, and seems to still hold true in the IPCC AR4 runs – although the average signal is smaller. And if it stays that way, southern Europe and the Middle East are likely to continue to get drier.

Thus, the Mediterranean region is at high risk for desertification. Even if 2007 were an anomalously dry year, these disastrous events show us that small perturbations in weather patterns can lead to tragic and costly outcomes. They also show what is in store for this area in the next few decades as global warming progresses. And desert makes more desert. As land is overused and scorched by the hot sun and no rain, it dries up and vegetation cover diminishes due both to drought and wildfires. Lack of vegetation leads to further loss of humidity and increases erosion rates. So, deserts expand even more.

Having said all of this, here is one last thought: if AGW is affecting nearly all latitudes negatively, then we need to wonder where global awareness of this problem stands. The Pew Center Report (2006) found that while over 90% of the population had heard of AGW in countries like the U.S., Germany, France, Britain, Spain and Japan, this dropped to about 75-80% in Russia, Turkey and China and below 50% in Jordan, Egypt, Indonesia and Pakistan (only 12%!). So the peoples in Middle Eastern countries most at risk for growing desertification are barely aware of the problem of AGW. What is even more striking is that of the percentages of people aware of AGW I listed, 70-80% are concerned about it in Indonesia, Egypt, Turkey, Jordan and India; but only 50-60% of U.S. citizens and peoples of China consider AGW to be a serious issue. So the problem isn’t just ignorance, it is also that of profound apathy.

124 Responses to “Sweatin’ the Mediterranean Heat”

  1. 101
    Timothy Chase says:

    Just looking at the NAO, it reminds me of the temperature trend in the US for the 20th century. The NAO was positive from about 1900 to 1940 as the US trended rising with the combined effect of greenhouse gases and solar variability, near zero or negative during the aerosol era from about 1940 to 1970s as the US trended dropping or flat, then positive with the warming from the late 1970s to present.

    Maybe I am just seeing a pattern which isn’t really there – but if warming results in a stronger Polar Night Vortex and results in the NAO trending positive with milder, rainier weather in the US SE, the teleconnections would appear to be there. The fact that warming has been milder in the US than the rest of the world has been something of a puzzle and I remember Hansen suggesting that the North Atlantic Oscillation might have something to do with it. Of course right now things are rather dry in the US SE, particularly in the Atlanta area.

    In any case, it helps to be able to identify the regional trends so that people will know what is “in it for them.” I would like to think that people’s concerns would extend beyond their own little neighborhoods, but often this is not the case.

  2. 102
    J.S. McIntyre says:

    re 89, 90

    re the number…mea culpa.

    Picked it up off the news (Yahoo) yesterday. Should have linked it, as the number is back down to .5 million now.

    (Should have known better; reporting in real-time often gets it wrong, going with breaking stories unconfirmed – for example, earlier in the day they were at 500,000, then dropped back to 250,000).

    As for your comments re Meditteranean climate.

    That’s California climate to a “T”.

    When the first post-gold rush surge of population was underway, many people who came to California transported with them the idea that they could plant English Gardens. It worked on the East Coast, so why not here. The idea is now more or less ingraned in the culture of the place. (Side note: something along the same lines happened in Australia, at least in terms of the assumption by settlers that the rules in one place – England – would be the same there. Of course, this wasn’t the case.)

    While English Gardens can thrive along the fog-bound northern coasts, the majority of the state’s climate is not suited to them. But we keep trying, laying down acres of grass and plants not suited to the area which, along with the need to have swimming pools uses up a lot of water.

    Since the first long drought in the 80s some homeowners have changed they gardening philosophy to reflect the way the climate here really is. I’m assuming that as the years go by, we’ll see a lot more of this type of gardening in the state. Then again, some people tend to cling stubbornly to bad ideas, so we’ll see.

  3. 103
    Figen Mekik says:

    This is a belated response to Cobblyworlds (#1) and Adam (#65) and it isn’t a complete answer but may shed some light. And I hope that the RC editors will pitch in should my response be incorrect or if they wanted to add something to it. I am a deep sea carbonate dissolution and sea surface temperature paleoceanography person by training, and not at all an atmospheric scientist. Gavin, particularly, has contributed much to this subject in scientific lit. And I got some help from Mike Mann in formulating this response.

    I simplified the cause for the development of the polar night vortex in my post to keep things simple, but the main impact of greenhouse gas (GHG) forcing on the behavior of the NAO index is related more to the vertical and meridional profile of GHG forcing than solely the polar amplification of surface warming. In other words, increasing the concentration of greenhouse gases in the atmosphere strongly affects the development and motion of planetary waves (large scale air waves -or sometimes ocean waves – whose driving force is variations in the Coriolis effect with latitude)from the troposhpere into the stratoshpere. So vertical motion in the atmosphere is enhanced with greater rising motion at high latitudes. This causes cooling north of 55 degrees latitude and warming at lower latitudes. The net result is a stronger polar night vortex high in the atmosphere. The last few sentences of my explanation are mostly paraphrased and simplified from Shindell et al., 1999.

    I hope this is helpful.

  4. 104
    J.S. McIntyre says:

    re 97

    “I’m from Big Bear Lake, CA. Big Bear only has two escape routes left at this time and no one is allowed up the mountain, except residents. ”


    My sister, who works on the real-time logistics end of fighting fires with the USFS, has mentioned that they’ve avoided disaster in the Big Bear area at least twice in the past few years, but that with all the bug kill its inevitable that it will eventually go.

    “There have been many reports I have read that were from the national forest service and those I spoke with in Big Bear Lake, that have said that the intensity of these fires, because they are burning hotter will decrease the likelihood of life returning to those areas due to the larger destructive capacity of these fires. This will lead to increased decertification.”

    Found this re soil damage:
    Fire can cause soil damage, especially through combustion in the litter layer and organic material in the soil. This organic material helps to protect the soil from erosion. When organic material is removed by an essentially intense fire, erosion can occur. Heat from intense fires can also cause soil particles to become hydrophobic. Rainwater then tends to run off the soil rather than to infiltrate through the soul. This can also contribute to erosion. In actuality, the negative effects of fires on soils are often exaggerated, and many fairly intense fires in western United States forests cause little soil damage. There is also the potential for alien plants to become established after fire in previously uninfested areas.


    This is also interesting, if brief:

  5. 105
    Hank Roberts says:

    Hi Figen, no, I’m just an amateur (got a bit past a MA in animal behavior and was studying honeybee behavior genetics when I ran out of grad school money, many decades ago). Can’t give up poking around, and I find amateurs can get amazing help from agency people who know things their agencies can’t or won’t support.

    Like scattering branches across burned areas instead of picking hillsides completely clean, to break up sheet erosion — even when it leaves some ‘burnable’ material out of the burn piles. Like digging the little contour grooves and berms that I learned from foresters in the Mediterranean countries via Usenet. Like spraying sugar water after a fire to cheat the invasive annuals of the flush of minerals and nitrogen the first growing season by feeding the microbiota instead.

    Look at the fire maps from So. California, zoom in on the fires in Google Earth, and many of them are tiny tightly packed developments full of houses (also known as “fuel”) crammed together in canyons surrounded by ‘open space’ — the people could live spread out in fire resistant buildings, but only if they didn’t try to take over all the surrounding land, only if they were willing to live in an area that’s able to burn fairly often, fairly cool, and fairly gracefully.

    After a fire it looks devastated. But the next few years are amazingly full of life coming back out.

    I don’t know how to get people to take interest in what lives on the planet around them, except to get out into the woods early as kids. Once fascinated, always fascinated.

  6. 106
    Figen Mekik says:

    John and everyone else affected by the SoCal fires,

    My deepest sympathies go out to you. This is indeed a horrific tragedy and I wish you great strength and patience. Hopefully the worst has already gone by.

  7. 107
    Hank Roberts says:

    Can you tell us anything about fire planning in Turkey?
    Intervals between fires normally?
    Lessons learned?

  8. 108
    D Pecan says:

    As always the postings on this site are superb.

    An acquaintance observed that the social sciences have not weighed in on either gradual global warming (which seems less likely by the day) or abrupt climate change (which seems more likely).

    Forecasts of specific weather alterations are likely to fall to the simplest chaos theory models, much like forecasting the weather five days out. Forecasting social disruption is easier, but I’ve not read anything academically rigorous. This may be due to the fact that nearly every forecast is likely to be derided as apocalyptic. This is a sad commentary on the social sciences.

    We’re being put to the test as a species. We’re hard-wired to respond to the tiger but blithely indifferent and insensitive to broad patterns and to long term potentialities. To ‘weather'(sorry there’s more than one person falling into puns) this looming crisis and maintain the basis of civilization we need futurists and visionaries who can tie all the pieces together.

    This may not be directly in the topic area of RC but it’s definitely something that will be necessary to a full discussion of this topic.

    Any social scientists out there?

  9. 109
    Svend Jensen says:

    105 Hank Roberts – Keyline designs PA Yeomans..
    Soil moisture retention and surviving fires in rural landscapes..
    More suitable for rural landscapes and reasonably hilly topograpy are the well tested ideas from P A Yeomans ‘KeyLine designs’ in Australia dating back to the 1950s. Also incorporated into Permaculture designs..

    BASED on strict adherence to accurately measured contour lines, water channels and grid lines, berms and when the rain does fall on an irregular basis, it falls keeping it localised and allowing penetration..Also sometimes practice of a ‘dip rip’ ploughing to improve soil moisture retention..
    Google : Keyline Designs by P:A Yeomans – interesting homepage

    Sadly, San Diego back country area is too hilly, too densely populated for Keyline applications..

  10. 110
    Figen Mekik says:


    I’m working on your question… Just some news: my summer vacation spot in Turkey, Bodrum , right on the corner of the Aegean and the Med is having massive forest fires as we speak. So far 1100 acres of land, 200 thousand pine trees and 10,000 olive trees have burned!

  11. 111
    Pekka J. Kostamo says:

    Some observations on the Mediterranean area.

    There are some 30 countries around this sea. These have a total population of about 450 million, of which 186 million live in the European Union and 260 million outside it on the southern and eastern shores. Water is the limiting factor for ecology, agriculture and cities in the area.

    Perhaps the most important background factor is the population age structure. The EU faces an increasing ageing problem, with a disproportionate number of people approaching retirement age. Outside, on the southern and eastern shores the enormous problem is too many children and youths, their education and employment prospects. Median age on the EU side is 40 years, in the rest of the region it is 24 years.

    Also the per capita GNP is widely different: 26.500 USD in the EU and only 6.500 USD outside.

    There is a long history and a multitude of links across the Mediterranean. Of major significance is also that some 25 million “Europeans” have kinship, clan and close cultural ties with the outside.

    A shift in climate patterns, and particularly the predicted decline of precipitation, is truly the last thing the region needs.

    As an example, my very sporadic monitoring of Moroccan press has provided some details. It used to be common wisdom that one year in six would be a drought year with major loss of agricultural output. 2007 was such a year and the Government is struggling with two major issues. High world market wheat price is pushing up the subsidies budget. Rise in the street price of bread would be a direct and immediate threat to social stability in the country. Secondly, there is a shortage of some 20% of seeds for the next agricultural season – and the season is starting right now. The shortage is of local, drought resistant varieties, traditionally held by the local small farmers.

    There was also a report on declining water resources. In the 1960’s the available water per capita was 2500 cubic meters. By 2000 the corresponding volume had declined to about 1200 m3. The Government target is to stabilize the water supply at a level of 750 m3 eventually – ways and means to be identified later.

    What next? Who knows, but it is most likely that major history for the Mediterranean has not ended yet.

  12. 112
    Nick Barnes says:

    Re #1 and #33: My understanding, based on an email exchange with Bill Chapman at CT, is that the Arctic sea ice anomaly should soon reduce. The arctic winter is always very long and very cold, and will probably freeze pretty much all the open water in the arctic ocean, despite the unique boundary condition of this summer’s melt. We haven’t had the rapid freezing that we usually get in October, because the arctic ocean was unusually warm. But it is cooling and will freeze before the end of the winter. So we will enter next year’s melt season with approximately a normal area of ice, but it will be thin and may not survive long.
    Of course, it all depends on weather and currents.
    I’m thinking of holding a sweepstake on this winter’s maximum ice area. I’d go for 12 million square kilometres (last winter was ~13, long-term average is ~14). Any takers?

  13. 113
    Aaron Lewis says:

    RE 112
    Salinity also matters. Less sea ice means more storm mixing which means higher salinity of surface water which lowers the freezing point, and changes density curves. See (

  14. 114
    Nick Barnes says:

    Re 113. Yes. This was something I left out for simplicity’s sake. I think it’s fairly important to say now that we’re expecting the anomaly to decrease, rapidly and soon. Otherwise when it happens we won’t hear the end of it from the deniers. Arctic sea ice has, rightly, become something of a poster child for the reality-based community.

  15. 115
    J.S. McIntyre says:

    Regarding the Southern California fires (and forest fires as they relate the AGW) it appears they served to keep the following news item more or less out of the public eye. disturbing example of more “business as usual” from the administration that says it “gets” Global Warming:


    Forest-fire warning cut from testimony

    As wildfires raged through Southern California early this week, the nation’s chief health official was prepared to tell Congress about one impact of climate change: “Forest fires are expected to increase in frequency, severity, distribution and duration.”

    But those words were never spoken. They were part of six pages of testimony deleted by White House officials before Dr. Julie Gerberding, director of the Centers for Disease Control, spoke Tuesday to the Senate Environment and Public Works Committee on the health impacts of climate change.

    See also:


    Links to the texts, drafted and edited versions. Note that the final version was 7 pages long, the draft 15 pages.

    Draft text of Dr. Gerberding, with excised material highlighted:

    Senator Boxer’s letter requesting drafts:


    It’s ironic that as we’re seeing an upswing of fires and an outbreak of so many of them in such a relatively small geographic area, the White House is determined to tell us things are not what they so obviously appear to be.

  16. 116
    Cobblyworlds says:

    #103 Figen Mekik,

    Thanks for getting back. Yes I have read Shindell et al and have read a fair amount about
    the AO/NAO (Wallace/Baldwin etc etc).

    My problem is that in their “Perspectives” paper (Science March 2007) Serreze et al note
    inconclusive results of studies looking at the impact of Arctic ice loss, but conclude that
    the seasonal loss of Arctic ice will have climatic impacts. That seems to stand at odds with
    the findings of a tropo/strato source for the mode of the AO. Although a summer-free ice loss
    could affect summer northern hemisphere weather, whereas the winter AO would still have a
    predominant impact on winter NH weather.

    My main problem right now however is that I’m an amateur reader of the science, and my day-job
    has been stupidly hectic for the last few weeks. As such I’m not really able to seriously
    consider this right now. So I’ll bow out of the issue.

    #112 Nick Burnes.

    Thinner ice allows more heat flux through it, and being less mechanically strong is more easily
    disturbed by storms, creating leads that allow more latent heat flux (water vapour). I too expect
    the extent to increase this winter as always. But the loss of perennial ice itself has implications
    for the region (possibly the whole N.H. – although I can’t pin that down). Nobody should be talking
    about ice-free year round this side of ~1000ppm CO2, but a seasonally ice free state would still
    involve a yearly grown seasonal sheet in the winter.

  17. 117
    Adam says:

    Re #103, that helps. Many thanks.

  18. 118
    David Garen says:

    Could you give the full citation for the paper by Cullen and deMenocal (2000) that you reference? When I click on the link, I just get an error message from Wiley that it can’t send a cookie to my browser, and I thought I had enabled this. If I had the citation, I could get to the paper via another means. Thanks …

    [Response: I shortened the link, so it might be ok now. – gavin]

  19. 119
    Figen Mekik says:

    #118. This is the full citation:
    Cullen H., and P. B. deMenocal, 2000: North Atlantic influence on Tigris–Euphrates streamflow. Int. J. Climatol, 20, 853–863.

  20. 120
    Nick Barnes says:

    I note from CT that the anomaly seems to have bottomed out at about -2.9 million square kilometres, and is now heading back towards more familiar territory.

  21. 121
  22. 122
    Ray Ladbury says:

    Re 121.

    “Congress does science:”

    Oh, be still my beating heart. Anyone care to make a wager that the information content will be greater than 1%?

  23. 123
    J. L. K. says:

    “Forest-fire warning cut from testimony
    Somehow it doesn’t seem sinister that an MD’s talk about forest fires was excised! She would have to simply parrot someone else’s opinion, no?
    The fires in SOCAL are simply tribute to the stupidity of the current inhabitants thereof. Everyone wants to live in those neat canyons and on ridge tops that are simply indefensible against fires that are powered by the Santa Ana winds. CDF calls these areas WUI for Wild Urban Interfaces and has a whole list of caveats, best practices and cautions. In the 2003 fires, the losses were mostly to properties that had not complied with weed/brush control, had not removed combustible shrubs and had not built in a manner that discouraged fire spread. [Notable exception is the San Diego County areas that had been infested with pine bark beetles, leaving huge fuel loads of dead pines!] When we get to see the After Action Reports on this season, it will probably be more of the same. We can’t buy and man fire trucks for every house and we can’t lay enough pipe to provide fireflows to a hydrant in front of every house. Some of the toughest duty in the world is fighting fires driven by 50-70mph winds. It can’t be done safely without triage. In short, we have to decide what we can defend and forget about the rest.
    Yet, when we try to do controlled burns during the wet season and work small areas to reduce the fuel load, we are met on every front with obstructions. It takes from 7-13 months to get all of the approvals and the restrictions are awesome. I know many task force leaders who don’t even try. They know that if they screw up, they are history. God forgives sinners, but the CDF sure as hell doesn’t! No one wants to endanger the environment by allowing perfectly natural burns when we can control them, but they scream like hell when we won’t defend their homes against towering flames traveling 50 mph! Go figure! This is truly a situation where you pay now or really pay later! One firefighter’ opinion.

  24. 124
    MoritzKoenig says:

    Interesting article, but wouldn’t the ozone layer depletion and the increase in CO2 emission cause temperatures in the Arctic to increase more than they would close to the equator? This would then lead to a smaller temperature difference between high latitudes and subtropics, causing smaller pressure contrasts, essentially leading to less east-west winds and a smaller chance of positive NAO phases!?!

    Am I completely wrong?