Today is our last day at Forskningsparken (Oslo Science Park); CICERO is headed to a new -even shinier- building next week, where we will be able to hang out with a whole posse of other Norwegian environmental institutes.

This feels like the end of an era, but in more ways than one. The CLM-FATES demographic vegetation model configuration for the Norwegian Earth System Model is now frozen, finally concluding a quest to better represent plant ecology in climate models that started (in my case) half a lifetime ago…

For a long time (this is my fourth attempt!), finding an Earth System Modelling centre brave/foolish enough to let us reinvent its land surface model so comprehensively felt out of reach (to represent ecological function properly, almost everything has to be rearranged). But in 2022, Kjetil and I decided that we would try and chart a tentative path to making a viable demographic vegetation configuration for NorESM3 using CLM-FATES.

Kjetil, Jessie, Adele and Betty at our last meeting Kyoto

At first we had (almost) no funding and no people, but gradually, our luck changed, and we had the -immense- good fortune to somehow assemble a talented and tireless ‘NorESM-land’ team of hardcore compulsive problem solvers from across CICERO, MetNo and the University of Oslo. Three years ago, we began meeting every Friday morning in ‘Kyoto’ (at CICERO our meeting rooms are nerdily named after climate treaties) and since then we have all been working crazy hard; developing, running, perturbing, emulating, tuning, code merging, debugging (and more debugging), benchmarking, conflict resolving, regridding, list-making, panicking (a lot), celebrating (sometimes) and eating (cake, often), to push the effort over its extremely complicated finish line, supported by our many similarly long-suffering and amazing colleagues icluding the core FATES developers at Berkeley Lab and the CLM team at NCAR.

Now we have a -lot- of CLM-FATES simulations in the oven (GCB, WIEMIP, CMIP7) and they don’t look too bad… To some (mildly alarming) extent, we don’t yet know what the future experiments will tell us (or indeed if they will crash in some exotic land use scenario, so we aren’t done until we are done!), and while this celebratory feeling might be mildly premature, it does feel like finally, the next stage of science (making it better, and finding out what it can tell us) can finally begin!

The model is open source, and will be released later this year as part of [NorESM3](<https://github.com/NorESMhub/noresm) and CLM6. More info is here :

On the way out, CICERO’s director Christine Tørklep and I said goodbye to the lovely tree on the way to the Blindern Metro. I’ve gotten really attached to that tree. Life has many chapters. Next stop Campus Ullevål!

Chrstine Tørklep and my favourite tree at Blindern Metro station

Land surface components of Earth System Models (land surface models, or LSM’s) underpin a huge array of decision making and climate policy, with uses in projecting the future of the Earth system, informing the Global Carbon Budget, generating climate impacts assessments, and many other applications. Because land surface models are used in a huge array of different disciplines, we land modelers tend to find ourselves at the end of the bill at allied meetings, but rarely are we the main event. Until recently, despite their critical role, there was no single event for land model teams to meet and discuss our scientific process as a community.

Four years ago, the First Land Surface Modeling Summit was held in Oxford, as an experiment in having a conference such a meeting. There are around 10-12 major land model development groups globally, and conference participants were invited to ensure representation across all these groups, in particular from active Earth System Model land development teams.

At the first ‘summit’, we all had such an excellent time discussing the trials and tribulations (and joys!) of our parallel model development and testing processes that we set up a new organization, the International Land Modeling Forum ILMF, (of which I am a member of the steering committee) to continue the work of creating a place to exchange ideas and share opportunities. Through ILMF, an exciting opportunity arose to host the second Land Surface Modeling Summit LSMS2 at the European Space Agency (thanks to the participation of Stephen Plummer from ESA on our committee).

In the run up to the meeting I had some anxiety that we might not manage to capture the same convivial energy and liveliness as the first ‘summit’ (this was a high bar!), but my fears turned out to be ill-founded. I don’t know exactly why, but land surface modeling is a really special community, with so many enduring friendships and mutual support between teams who are, on paper, competitors of one-another. Part of this is that this problem, of how to model and understand the functioning of the entire terrestrial biosphere, decades into a huge array of possible futures, is both mind-blowingly difficult and, on some level, existentially imperative. I think it is clear to most of us that we all need to work together to find the best path fowards. That, and lots of them are just really funny… I think the meeting was a great success, but it has taken me several days for my brain to readjust back to normal life!

During the conference I gave a talk on “Demographic Vegetation Models and Earth Observatioons: A new era of carbon cycle science”, much of which was inspired and informed by our NextGenCarbon EU research project from my previous trip. I also led a breakout group) on the same topic, thinking about how we can best describe our new crop of models so that the upcoming raft of CMIP/TRENDY/WIEMIP simulations is better informed on the important nuances of their structure. Adrianna, Kjetil, Jennifer, Charlie and Stephen de Hertog (U. Delft) also gave poster presentations on different aspects of CLM-FATES work.

A specially huge thank you is due to Eleanor Blyth, who organized and hosted the first LSMS meeting in Oxford, and for whom this meeting marked the very last day of her career at the Centre for Ecology and Hydrology, near Oxford in the UK. I have known Eleanor since my first post-doc in 2005, and for all that time she has been a uniquely unifying force in the UK land modeling community, and in the ILMF we will miss her enormously.

Another great part of the trip was the team-CLM/FATES extracurricular warm-up weekend that Adrianna organized in Bolsena, and staying in a big staircase of flats in Rome all together! It’s been many years since I last saw most of the CLM-FATES crew (whom I talk to online every few days) in real life, and it was amazing to share such fun times (and approx. ~49 espressos, 16 pizzas and 56 gelato) with them!

Left to right: Jennifer Holm, Kjetil Aas, Daniel Kennedy, Will Wieder, Danica Lombardozzi, me, Charlie Koven, Adrianna Foster, Dave Lawrence, Linnea Hawkins and Katie Dagon. (Though we failed to get Marcos Longo in this picture. Sorry, Marcos!)

It has been quite a month of work travel. The busiest for a decade! My heart and my inbox are full in equal measure. Now time to settle down into writing annual project reports and documenting our new and shiny implementation of CLM-FATES in NorESM3 :)

The annual meeting of our NextGenCarbon project was held in early September in Umeå at the Swedish University of Agricultural Sciences SLU, which is also the home of the project coordinator, Rubén Valbuena.

NextGenCarbon is focused on better use and integration of Earth observations (via ESA) into models of the Carbon Cycle. Jessie and I travelled by train from Oslo; the astonishingly extensive forests of central and Northern Sweden providing a psychological warm-up for our consideration of human legacies on terrestrial biomass storage. This was our second annual event, but I was sick during the last event in Paris, so this year I enjoyed learning a great deal more about the work of our project partners!

NGC is working across a wide spectrum of carbon cycle understanding, from ICOS towers, through National Forest Inventories, novel Earth observations of forest structure and growth/loss, demographic land surface and book-keeping models, land use mapping and future scenario generation. One critical element of this is consideration of how different emerging demographic vegetation models represent vegetation. ‘Standard’ land surface models are typically quite similar in their charecterisation of ecosystems as average properties of plant types in a particular location.

The new generation of higher-fidelity models, of which CLM-FATES is an example, diverge in how they represent ecosystems in terms of their age and size and history. This makes it challenging to know how satellite and ground measurements of these things can be used to make comparisons between models and data that make sense.

The nice thing about NGC is that many of the modeling groups developing this new generation are represented (CLM-FATES, ORCHIDEEE, JULES-RED, JSBACHv4) so we can have the conversations we need to have to figure out what the next generation of land model benchmarking looks like.

A big highlight of the trip was the genuine Swedish sauna and swimming in the gorgeous glassy river Ume. Simon Besnard and I even saw the Northern lights on the way home!

I also caught up with my old friend Dan Metcalfe who has lived in Sweden for many years since we both worked at Caxiuanã for our PhD’s. Dan is the mastermind of several methods now widely used to measure the details of carbon budgets in tropical forests, and he is expanding these to Scandiavian forests, opening the potential for collaboration with our projects on high latitude systems.

In mid August i was lucky enough to be invited to teach at the Advanced Climate Dynamics Course ACDC at Abisko in Northern Sweden.

25 years ago (!), my office friends in Edinburgh (Lorna Street and Mark van Wijk), spent much of their time observing, theorising, talking and writing about Abisko, and so this remote and scenic field station is somewhat legendary to me, further because it is surrounded by (without any doubt) the most well studied and experimented-upon bit of arctic tundra in Europe. I’ve been calibrating models with data measured there sporadically ever since (not least Alistair Rogers’ measurements of his all-time-highest photosynthetic capacity of the shrubby vegetation down by the lake, so he tells me…). So naturally, I was very enthusiastic to visit - even more so as it meant I got to circumnavigate Scandinavia by train (Oslo-Stockholm-Abisko-Narvik-Trondheim-Oslo) in the process :)

ACDC has been running since 2009, and gained an enviable reputation along the way. Land modeling is typically out of their remit (I don’t consider myself a ‘climate dynamicist’!), but this year’s theme was ’landscapes and climate’, and so happily I was asked to join to complement Marysa Laguë and Lucas Vargas Zeppetello’s land-atmosphere interaction perspectives. Much of the time I spent trying to grapple with the thorny question of how complex models can learn from more idealized representations of the system (perhaps there is a need to create even-more reduced form configurations of our models?).

One huge highlight was having a little time to briefly think hard about matters not directly relevant to operational land surface modeling! As such, it was an immense delight to discover Øyvind Paasche’s fascinating and unique journey from philosophy to geoscience (and maybe back again!), as well as experience his field tutorial in how to think like a geomorphologist (which I apparently ruined by chatting too much, so I don’t think there’s much hope for me in that arena…). I also had my mind thoroughly blown away by Axel Timmerman’s insanely brilliant and interesting research into prehistoric mammalian food webs, human dispersel and the evolution of culture. Given that the vegetation predictions of CLM underpin the relationship between climate dynamics and ecosystem productivity in his paleoclimate simulation, I left awash with distacting ideas for how to link FATES’ treatment of vegetation dynamics, ’tipping point’ thresholds and fire into that system.

Many thanks in particular to David Battisti who is the maestro of ACDC and has invested a huge part of himself into its legacy and pedagogy. I now (at long last) think I might actually know what Rossby waves are, and moreover, learned a great deal about how to create an excellent and immersive environment for teaching and learning…

I’ve a bit more travel in September to the NextGenCarbon meeting in Ümea, and the Land Model Summit in Frascati, so I couldn’t stay for the whole two weeks of the summer school. But If you all invite me back I’ll make sure to stay the whole time!

CICERO has been funded by the Spark Foundation to run CLM-FATES as part of the Warming Induced Emissions Model Inter-Comparison Project. In past iterations of CMIP, there has been no platform to look specifically at the responses of the terrestrial biosphere to future climate forcing, aside from the limited set of coupled experiments inside the C4MIP protocol (which are coupled climate model experiments and so subject to large variations in baseline climate which can overwhelm differences in behaviour). Future offline experiments were part of ISIMIP, which has a timeline delayed by several years as waits for the output of coupled model future scanerios to drive impact models (none of which is focused on the responses of natural ecosystems or the carbon cycle to future climate forcing). This is major gap, but happily time and dedicated resources have been provided by the SPARK foundation to provide this critical missing information.

Simulations are running and submissions are planned before summer to allow analysis to inform IPCC-AR7 timetables.

For several years we in the NorESM, CLM & FATES network have been very busy coding, calibration, testing debugging, recalibrating, coupling, spinning up, and analysing output from the first global version of CLM-FATES. This has been a monumental and all-consuming effort on the part of a large team of dedicated scientists. To more broadly share our activities with the newly minted global version of the model. There are a few goals.

1. To provide a means of pooling infomation from across several institutions and networks on the many scientists working on CLM-FATES related activities in Norway (see network page)
2. To provide publically accessible updates on these activities (via news and blog posts)
3. In think that land surface models are a unique and undervalued as a tool to help humanity plan for the future, both near and distant. Here I plan to discuss the myriad of interesting ways we could, should, or do leverage them to provide insights into the functioning of our fabulously complex and interlinked world.