I wished that I had a dime for every time I heard people complain about how big the AGU meeting has grown! It is an enormous meeting: the number of participants this year was rumored to be around 22,000 people. Moscone South, the venue for the poster sessions, stretches one entire city block so that if you forget something at one end and remember it only at the other end, you are in for a hike back to retrieve your stuff! And shuttling back and forth between Moscone South and Moscone West, where the sessions are held, became a bit like a salmon run with hundreds of researchers streaming from one venue to the other. It was hard to advance if you are going against the flow.

 But it is the very size of the AGU that makes it such a useful meeting. While not all the world's earth science researchers attend, a significant fraction do, including very strong representations from East Asia. And the AGU meeting management provides very good tools for navigating the 12,000 posters and 6,000 invited presentations.

 I found so much relevant material at the AGU meeting that I have decided to break this blog entry into two parts. This first part will explain the informational resources that I have brought back from the meeting and highlight some of the talks that caught my attention. The second will explore some of themes that I saw emerging from the conference as relevant to MRI.

Informational Resources

The AGU Scientific Program webpage appears to be public so that you can search the program and create an itinerary (complete with abstracts) as I did (see below) even if you were not registered to attend. 

In addition, poster presenters were encouraged to submit their posters to an "ePosters" site. Thus if you know the session code for the poster (and if the presenters uploaded their poster to the site, which alas many did not), one can download the poster! For instance the very first poster in my itinerary, C11B-0674. Glacier Surge-like signals detected by SAR-based technique in West Kunlun Shan, NW Tibet T. Yasuda; M. Furuya, has been uploaded to the site. By inserting the code "C11B-0674", I can download the poster itself.

Finally, the AGU has finally begun to record some of the presentations and has made them available via a Sessions on Demand button.  The AGU recorded numerous Union sessions and plenary lectures over the five days of the meeting. I suggest that you look at all five days to see what might be of interest to you. I particularly suggest the sessions GC43G Stephen Schneider session and the GC43H session on the history of global warming, both on 8 December 2011.

What exactly did I do? I searched the program using keywords such as "mountain", "Sierra", "Himalaya", "Tibet", and so on. After several iterations, I arrived at a nearly feasible itinerary focused on mountain-related topics, which runs to 254 pages and 18.3 MB. This itinerary kept me shuttling back and forth between sessions and posters for the whole time that I was there. I was extremely pleased with this modus operandi as I felt like I was able to focus right in on the most relevant work presented at the meeting. While I was not present on Friday,  I included Friday in my itinerary document so that the full five days were searched.(If the preceding hot link doesn't work, copy the following URL into your browser: http://194.150.248.152/~swissmat/index.php?option=com_docman&task=doc_download&gid=1268&Itemid=43 )

How can you use this mountain-focused itinerary of posters and papers? If you browse the first pages of the itinerary you will find the titles of papers and posters by session. Yes, there are 27 pages to this section of the document but as the papers and posters are organized by session title, you should be able to see quickly those that are of interest to you.

 To see the abstract of any of the posters or papers listed there, simply search the rest of the document for the name of the first author and you will eventually find the abstract.

Papers and posters that made an impression on me

B12B-01. Mechanisms of Alpine Treeline Stability (Invited) W.K. Smith; G. Wieser; F. Holtmeier

 I am always taken by papers that explain something at one level by mechanisms at another. Bill Smith assessed treeline dynamics by looking first at seeding survival, then at growth of seedling to tree stature and finally the facilitation of survival and growth by the presence of existing vegetation. Seedling survival was a question not just of carbon acquisition under low temperatures, high radiation loads and water stress but also of the processing of that carbon into plant structures. Snow played an important role is protecting seedlings and young trees but also presented physical challenges, especially abrasion, to growth above the snow layer into tree form. Finally, the presence of other plants greatly influenced the success of plants moving from one life stage to the next, a point also made by Lara Kueppers in one of her several papers or posters on the Alpine Treeline Warming Experiment on Niwot Ridge (though I can't seem to find mention of it in the abstracts). Smith emphasized that from an ecophysiological point of view, high elevations differ considerably from low-elevation polar environments in at least two ways - very high incoming radiation which often leads to photoinhibition of carbon processing, and very low partial pressures of CO2 with large impacts of fluxes. These two points seem directly relevant to the "why are mountains different" theme that was proposed by Mark Williams and Dave Schimel in Perth last year.

C23F-03. Mountain front precipitation accumulation over a 3300m elevation gradient from scanning LiDAR snow depth and in-situ instrumental measurements, southern Sierra Nevada, California P.B. Kirchner; R.C. Bales; J. Flanagan; K.N. Musselman; N.P. Molotch

 I am always on the lookout for papers that examine the relationship between elevation and precipitation. Pete Kirchner created a very high resolution dataset on spring snow depths from the difference between snow-on and snow-off LIDAR measures, and found three distinction relationships between snow depth and elevation: a steep relationship below 2100m where precipitation was frequently mixed rain and snow, a shallower but highly significant relationship above 2100m (in non-forested areas), and then a flat relationship but with great variance at high elevation where snow redistribution was dominant.

 Related talks included Danny Marks who reported on trends and projections of the rain-snow transition line and its impacts on runoff, a poster by Anders looking at elevation- precipitation relationships from different mountain regions around the world, and another poster by Minders looking at the mechanisms for ascent-forced convective precipitation over a mountainous island in the Caribbean. Minder also gave a good paper on the mechanisms for a lowering of snow line as the rain-snow transition zone approached mountain fronts.

A23E-06. Representation of the Sierra Barrier Jet in 11 years of a high-resolution dynamical reanalysis downscaling M.R. Hughes; P.J. Neiman; E. Sukovich; F.M. Ralph

 Just prior to Minder's paper on declining snow line, I listened to a fascinating talk about a low level jet that runs parallel to the Sierra Nevada in California. For years I lived in the foothills of the Sierra, and never did I suspect that a thousand feet above me was a river of air moving parallel to the crest that forced upward incoming Pacific moisture and thus initiated precipitation well before the mountains. Several other talks in this session examined the nature and the impacts of the jet. I wondered if such jets were features in other mountain ranges as well, such as the Himalaya, that block prevailing winds.

H24D-05. Widespread hillslope gullying on the southeastern Tibetan Plateau: Human or climate-change induced? (Invited) J.D. Pelletier; J. Quade; R. Goble; M. Aldenderfer

 Having seen these gullies myself, I had to go to this talk, which concluded quite simply that the advent of pastoralism with a reduction in vegetation cover was insufficient in itself to initiate gullying, as gullying had not begun in earlier, drier periods when the vegetation cover had been previously reduced. The gullying required as well on-going and increasing precipitation of the late Holocene to push the landscape over a threshold. I like this line of inquiry as I think climate change impacts on geomorphology in mountains is not well researched.

GC32B-03. The Modification of Orographic Snow Growth Processes by Cloud Nucleating Aerosols (Invited) W.R. Cotton; S. Saleeby

 I listened to Bill Cotton four years ago at a meeting that MRI recorded in Boulder CO . I found his talk quite interesting then and so I expected a good talk this time. I was not disappointed. Here he talked about how aerosols, which can come from pollution, changed the rate at which snow flakes form and fall, so much so that horizontal winds push the precipitation further downwind often for 10's of kilometers. In the area of the study (on the Continental Divide in Colorado) this horizontal shift in precipitation moved significant amounts of precipitation from the Colorado River Basin on the Pacific side of the Divide to the North Platte River on the Atlantic side. Thus aerosol pollution can change not only the amount but also the location of precipitation.

GC32B-04. Hydrometeorological Scales Interactions in the Andes and their Impact on the Amazon Climate (Invited) R. Avissar; D. Medvigy; R.L. Walko

 This talk showed the importance of mountains to the climate of surrounding areas. These researchers use the Ocean-Land-Atmosphere Model, which, besides featuring these coupled elements, models the entire planet using a variable sized grid. It thus produces regional climate predictions from a single model with a very fine grid spacing in regions of interest and larger grid spacing over the rest of the globe. Here they demonstrated that if one did not model the topography of the Andes at a sufficiently fine scale, then predictions of the inter-annual variability of precipitation over the Amazon were incorrect. I would have liked to have heard exactly what about the topography was important in getting better predictions of precipitation over the Amazon, but a little more digging in ISI will probably lead me to the answer.

C33F-08. Past peak water in Peru’s Cordillera Blanca: diagnosing the demise of glacier influence on stream discharge M. Baraer; B.G. Mark; J.M. McKenzie

Michel Baraer participated as well in the MRI KCW on 4 Dec in Berkeley. His talk here reviewed the expected time course for stream discharge from basins in which glaciers are receding, that is, an initial increase in discharge due to the melting ice volume and then when the glaciers are considerably small a reduction in discharge. He then reviewed discharge data to show that "peak water" had already occurred in the Cordillera Blanca of Peru, and that stream discharge was now on the descending limb of the expected curve. Thus the Cordillera Blanca is now where other major "glacierized" basins, such as those in the Alps, Central Asia and the Himalayas, may be in the future.

GC34A-08. Non-linear feedbacks between climate change, hydrologic partitioning, plant available water, and carbon cycling in montane forests (Invited) P.D. Brooks; M.E. Litvak; A.A. Harpold; N.P. Molotch; J.C. McIntosh; P.A. Troch; X. Zapata

Paul Brook's talk was one of eight talks in the CIRMOUNT-organized session: GC34A. Climate Change and Drought: Climatic Water Deficit and Water Balance in Mountain Systems: Biophysics, Ecohydrology, and Impacts.  All the talks here revolved around the ecological importance of potential evapotranspiration, actual evapotranspiration, and the difference between the two, climatic water deficit. All of these measures integrate precipitation and energy over the annual cycle and so are more pertinent than total annual precip or mean annual temperature at explaining the distribution and abundance of organism. I recommend that you look at all the abstracts in this session.

 Brook's talk focused on some of the mechanisms related to snow accumulation and melt that strongly influenced water availability to vegetation. These include sublimation of snow, which is strongly influenced by tree canopy cover, and the depth of early snow pack as it influences the depth of frost in the soil and the soil's subsequent porosity at the time of spring snow melt. Once again, factors that one might not suspect at first (e.g. the degree of tree canopy or the timing and depth of first snowfall) turn out to be very important in the ecohydrology of mountains. Only when we know these will we be able to project better the impact of future climate and land cover change.

C44A-04. Past and future glacier changes the western Nyainqentanglha Range on the Tibetan Plateau (Invited) T. Bolch; F. Chen; W. Yang; M. Buchroithner; E. Huintjes; S. Kang; A. Linsbauer; F. Maussion; F. Paul; C. Schneider; D. Scherer; T. Yao

Tobias Bolch of the University of Zürich gave several talks at the AGU, all of them excellent. One (C43A-02) focused on a summary of glacier trends over the entire Himalaya, a definitive statement of which is essential in the face of the erroneous statements in the IPCC AR4. This talk (C44A-04) highlighted a method to assess glacial area and volume over time that is applicable at the scale of an entire mountain range. The relevance and clarity of the question was matched by the scope and completeness of the answer, representing to me the best of what research can do.

GC13C-06. Hurricanes in a Warming Climate (Invited) K. Emanuel

One of the reasons to go to the AGU is to hear talks from people who have made important contributions to the climate change debate (or to other scientific questions, such as the origin of life), even if their contribution has nothing to do with mountains. Kerry Emanuel updated his findings on the growing intensity of Atlantic hurricanes. He presented an amazing map of historic tropical cyclone paths, which clearly showed that the Pacific Ocean has many more tropical cyclones than the Atlantic ( a similar image is here) and leads one to wonder why the media focuses so strongly on Atlantic hurricanes.  He noted that there is no climate change signal apparent in the intensity of Pacific cyclones, which seems to vary with ENSO. However the intensity of Atlantic hurricanes does show a climate change signal, a combination of warming sea surface temperatures coupled with a cooling tropical tropopause. He emphasized that the damage from cyclones is overwhelming associated with a few high-intensity storms so even a minor increase in intensity would have major impacts on damage.

So I am in Urumqi at the International Symposium on Changing Cryosphere, Water Availability and Sustainable Development in Central Asia. Nice title, you might say but first, where is Urumqi? Urumqi is the capital of Xinjiang, an enormous province that constitutes northwestern China. It is in the center of Asia. Look at a map of Asia and find what looks like the center and you will not be far from Urumqi. I was told that it is farther from the ocean than any other city on the planet. It is a region of high mountains and enormous desert basins, arid and cold. The local language is written using Arabic script. Somewhere to the west of here is the Dzungarian Gate, a gap in the mountains with a name that seems to come straight from myth.

Why am I here? In large part, because of the time I spent in China last year. When I visited Lanzhou, Liu Shiyin, one of the organizers of this conference, asked if I would participate on the steering committee. I was pleased to do so for many reasons, one of which was that I would be helping shape an IGBP-sponsored activity, another part of which is a synthesis workshop that MRI and IGBP will stage in Switzerland next year.

 Qin Dahe, co-chair of IPCC's WG I with Thomas Stocker on University of Bern, gave a great talk on the current status of WG I's work.  Unfortunately he can't give out copies of his presentation, otherwise I would put it on the MRI website. The world has clearly warmed but not uniformly, as his maps showed. The hockey stick continues to be developed, now with previous unused Russian data. Ocean acidification has joined climate warming as a major impact of CO2 enrichment. Modeling of the Earth System continues to progress, narrowing the bands of uncertainty, even as more and more elements of the system are included, while accumulating observation data vindicate previous generations of models. He showed a particularly good slide summarizing what is known regarding radiative forcing, and an even better slide to show that while geo-engineering with sulphates could indeed halt temperature rises for a few decades,  it would not eliminate the warming caused by CO2 unless of course sulphate pollution were maintained  indefinitely. As a result warming that might otherwise have occurred over 30-50 years would occur in 10-20 years after the cessation of sulphate emissions, a scary idea.  Qin's talk was almost rousing: the kind of Big Picture statements the science can deliver.

 Ray Bradley, one of MRI's most active SAB members, followed Qin and emphasized that the importance of high mountain climates, particularly those with large cryospheres, is exceeded only by our lack of observations within them and our consequent ignorance of exactly what is happening there! One of the perpetual questions is whether the rate of climate warming increases with altitude, with data in the literature on both sides of the question, and various mechanisms proposed for something the reality of which we are not yet sure. And yet in places such as Central Asia and the tropical Andes the fate of cities and agriculture are tightly linked to very high elevation glaciers, snow cover and permafrost. This situation calls for a "high elevation climate observation campaign" that both compiles existing dispersed and often unhomogenous climate data à la HISTALP and promotes the installation of new observation stations at very high altitudes. I saw several examples of such dispersed data sets at the conference. For instance, He Xiaobo of CAREEI has 5 months of high-quality data (e.g. with ventilated thermal sensors) for each of 7 consecutive years at 5600m. Ray gave me his presentation and so you can download it here). And while this is not the first time that high-elevation observations have been proposed (Gianni Tartari and Vladimir Aizen launched CEOP-HE with such a focus), the focus here shifted away from operational observations, always a hard sell to cash strapped hydromet agencies, and toward a research network motivated by a few key questions. 

 Dong Wenjie of Beijing Normal University did something that I have never seen before. He ran the emissions histories of all countries, of Annex 1 countries alone, and of non-Annex I countries alone through a community model to quantify the climate change impacts attributable to these different groups of nations. Not surprisingly, his preliminary results show that the non-Annex 1 countries' emission alone would lead to very modest changes in the earth system with the obvious conclusion that the really serious impacts are due to Annex 1 country emissions. This is not a terribly surprising result, particularly for an author from a non-Annex 1 country, but what I found so interesting were the questions related to path dependency in such a model. Are the impact of Annex 1 countries, as measured by Dong as the difference between those created all countries and those created by non-Annex 1 countries the same as one would find if one ran just the Annex 1 countries' emissions alone? I expect not. Beyond that, are the marginal environmental costs of emissions constant, decreasing or increasing, and what does that meant for the obligations for various countries? I could easily imagine someone from an Annex 1 country asserting that the marginal costs of their emissions in the last century were much lower than the marginal costs of current emissions by non-Annex 1 countries and BRICS. If true this raises dizzying questions of what Annex 1 countries owe, if anything, for their much earlier consumption of the global atmospheric commons.

 Xiao Conde added to the growing body of literature of the real fate of Himalayan glaciers with a study of climate records along the Himalaya and a review of what glacier inventories from remotely sensed data say about glacier extent. A key take-home message for me was his use of just 13 long-term climate records for mountain range that extends of 3000 km. Clearly we need to get more climate data on the Himalaya, yet another reason for a high elevation climate observing network.

 I was pleased to see that human dimensions appeared in at least three talks, two on vulnerability assessments and a third on land use change. What I found lacking in most of these talks was a focus on the exact mechanisms involved in vulnerability and a more analytical approach to defining exactly who is vulnerable. When we did similar work for the California Department of Forestry and Fire Protection (in my previous life) we could trace vulnerability to fire through equations involving probability of  a fire at a given location, the probability that the fire would destroy the structure and the value of the structure so that we could calculate an expected value for loss. At the same time, other researchers in California were modeling land use change, not using general linear models, but models of how the underlying processes of land use change, principally market economics, actually worked. Such an attention on mechanism rather than a fishing expedition for correlations would move this theme along more quickly.

 The use of  CMT, circulatory mode technique, in statistical downscaling illustrates nicely the utility of thinking mechanistically. Some forms of statistical downscaling of temperature and precipitation assume that the variation in T and P will persist under future climates but that their central tendency will change by the amount of change in T and P from GCM results. The assumption that the variation will remain similar in the future seems hard to defend as that variation is the result of a number of different mechanisms. CMT focuses more closely on those mechanisms by looking at different circulation modes, i.e. the proportion of the year when the winds aloft come from different directions. Different amounts of T and P are associated with those circulatory modes: think rains in California from SW flows, drying from north or easterly flows. As GCMs produce above all else potential future circulation, one characterizes a future climate in terms of different proportions of circulatory modes. Assuming that the relations between T and P remain unchanged within a circulatory mode, a more intuitive proposition than assuming that the overall annual variation of T and P will remain constant, one can then predict a future pattern of T and P. It is quite likely that this new pattern of T and P has quite different statistical properties than the current climate, yet it is firmly based on a mechanistic understand of climate as the net cumulative effects of circulatory modes.

Anyhow, I thought it was cool!

As I was not able to participate in the Sonnblick Conference I asked Chris Ritter, MRI's Communications and Events Manager, to write up his impressions of the meeting:

Each day at the Climate Change in High Mountain Regions conference we saw some beautiful images of remote sites and slides of remarkable data analyses, often from those same locations. Reinhard Böhm opened his historical tour of Sonnblick with a lovely photograph of the Hohe Tauern summit from which a stunning diversity of observations have been collected continuously for, now, 125 years. I clearly recall as well the photograph of the remarkable group gathered at the observatory for the international meteorological conference in 1922. I will return to this second image.

Among slides of results from analyses and models that spanned the full spectrum from straightforward to subtle and complex, I also recall (as a lapsed radiochemist would) a slide from Ingeborg Levin’s talk on atmospheric carbon cycle research using radiocarbon data. Data from tree rings plainly illustrate the Suess effect, the slow dilution of atmospheric ¹⁴C by CO₂ from fossil fuels (in which the t_1/2 = 5700 y ¹⁴C has long since decayed away). It’s such a simple graph that originates in such a consequential thing: a planetary change that humans have accomplished in a geological instant by moving carbon from underground into the atmosphere.

Following the slow reduction of the fraction of ¹⁴C in the atmosphere to the moment at which on another mountaintop half-way round the Earth from Sonnblick Charles David Keeling initiated data collection for what has become his famous, eponymous curve, Levin’s ¹⁴C curve illustrates another human intervention on a planetary scale: the virtually instantaneous spike in atmospheric ¹⁴C due to nuclear weapons testing. At 1963, when the limited test ban treaty relocated weapons tests into the domain of fossil fuels, the atmospheric ¹⁴C curve begins to decline. Although the decline looks like a radioactive decay curve, it’s not (¹⁴C’s half-life is too long). What this unintended but revelatory global tracer experiment now reflects is carbon partitioning in the biosphere, and the surface ocean and below.

Levin’s slide is only one of several hundred that extract indicators of global change from what I thought was a stunningly diverse range of analyses and proxies emerging from a rare and wonderfully organized gathering of researchers from so many different scientific disciplines. I wonder whether those scientists who gathered at Sonnblick in 1922 imagined either the climate changes we study or the great scope of the science we bring to bear nine decades later on climate change. But no matter. We have been doing science for a long time (or at least a long time in human terms) on this and other summits that has taught us about meteorology and climate, environment and pollution, technology and its human and planetary consequences, the heavens and the Earth. And so much more. And there is much more than that to come.

I wonder, what will our scientific descendants be talking about at Sonnblick 250?