As Climate Has Changed, More Homes Near Wildland Vegetation Have Burned

The title might seem self-evident, but with science, nothing ever is.

I’ve been excited to write about this paper since it came out a couple months ago, and here’s a pdf for those who can’t get past the paywall on Science. This is new research coming out of Volker Radeloff’s lab – one of the best in the business.

The authors present compelling data showing that the amount of Wildland-Urban Interface (WUI) has increased over the past decades, and that consequently the number of homes in the WUI has increased, so it’s not a surprise that many more homes have burned in the WUI. About 150,000 buildings were exposed to wildland fire from 2000-2013. Of these, about 10% were destroyed. In some forest types, however, 20-23% of buildings were destroyed.

One surprising conclusion from their data is that fires in grasslands and shrublands destroyed far more homes than fires in forests. Intuitively, because forest fires have the potential for greater intensity and severity, we’d expect them to destroy more structures than fires in shrublands. But because so much of the WUI is in the west, and so much of that landscape is either grassland or shrubland, there have been many structural losses in those vegetation types. Interestingly, shrublands can actually accumulate fuels faster than many forest ecosystems, meaning that reburning over time has the potential to keep these vegetation types higher-risk places for housing development. This might be of special concern in Texas, where housing in the WUI has increased by about 20% since 2010.

With many homes already in the WUI, and more development being planned, it becomes increasingly important that we have intelligent policies in place to manage the risk from wildland fire, along with all the other pressures on ecosystems in the WUI, as well as the needs of people who live there. Anyone who is interested in wildland fire, climate change, and public policy should take a careful look at the data in this paper.

FDA Approves CRISPR/Cas9 Treatment for Sickle Cell Anemia

The science here is so cool that I have to make a note of this advance in biomedical technology and treatment. The Food and Drug Administration approved a gene editing treatment for sickle cell anemia last week.

Sickle cell anemia is a genetic disease caused by a single recessive allele. Having two copies of the gene is highly deleterious – that genotype causes sickle cell anemia, in which malformed red blood cells interfere with circulation and cause debilitating pain, organ failure, and cardiovascular disease.

There’s an interesting evolutionary biology story here, too. Because the gene is so harmful, we might expect it to be eliminated from our population over time. But a single copy of the gene confers resistance to malaria. The plasmodium parasite that causes malaria has a hard time affecting red blood cells that carry just one copy of the gene. In places where malaria is endemic, heterozygotes have a fitness advantage. The genes for sickle cell therefore remain relatively common in populations. This is true for much of Africa, including the regions from which people were captured and forced into slavery in the United States. That history of evolutionary biology and slavery has combined to make sickle cell disease much more common for Black Americans than for other Americans.

The CRISPR/Cas9 system uses a combination of bacterial and viral mechanisms to precisely edit the genes of all cells in a living organism’s body. Cas9 is a naturally occurring enzyme in bacteria. It’s a nuclease – an enzyme that cuts up DNA. When some bacteria are infected by a virus, they can produce matching RNA sequences to the viral DNA, which Cas9 can then target, chop up, and deactivate. Cas-9 uses a sequence of short, repeated palindromes to figure out what to target. That’s the CRISPR part of the system (clustered regularly interspaced short palindromic repeats). Cas-9 is also an abbreviation, for “CRISPR-associated protein 9”. I’m sorry for all the jargon and abbreviations, but that’s just how molecular biology is – it’s so complex that we’re constantly inventing new language to describe what we’re studying.

Recently, scientists realized that we could use Cas9 to precisely cut up DNA in people, using a customized RNA sequence to target the spot in the genome that we want to make a cut. We can cut around a mutated gene that isn’t functioning properly and then insert a functional copy of the gene, which will fit into the cut ends of the DNA and result in an intact, functional genome to allow a patient to live a healthy life, free from that genetic disease.

Of course, there are some downsides and caveats to this treatment. First, it’s incredibly expensive. The treatment from Vertex Pharmaceuticals will cost about $2.2 million. An alternative Cas9 sickle cell treatment from Bluebird Bio will cost about $3.1 million. That’s a ton of money. At the same time, living with sickle cell disease is also expensive, costing about $1.7 million over a patient’s shortened lifetime. And that’s a life plagued by constant illness. We can expect more treatment options to become available, and costs should come down over time.

This treatment is difficult, and doing it safely requires a tremendous amount of specialized work. Companies expect that they might be able to treat up to 10 patients per year given their current capabilities. If the treatment proves effective and safe, this capacity will grow, but it will still take a long time to be able to provide treatment to everyone who needs it.

And long-term safety is also a concern. We have no theoretical reason to suppose why there should be long-term health risks. But biology and medicine are complicated; there is still a lot that we don’t know, and there may be unknown risks. I’ll keep paying attention to the new data coming out about these treatments, and I’m excited and optimistic that this is the start of a new era in human health.

Carbon Accounting Through Climate Extremes

I want to draw attention to recent research by a large group of scientists showing the impacts of a recent drought on forest carbon dynamics. The team reported their findings recently in Nature Communications, in a paper led by Auke van der Woude. The work is fascinating, and is already generating substantial commentary among scientists. I want to highlight just a couple of take-home messages from this work.

First, forests are dynamic. As temperature and available moisture change, the trees in a forest change what they’re doing in response to their environment. If a leaf is too hot or the soil is too dry, for example, photosynthesis slows, and the tree’s overall ability to store carbon declines. At the scale of a forest, this reduction can be substantial. And if we are relying on forests to sequester carbon for us, their ability to do so will be dynamic over time. In this case, the authors showed that extreme temperatures and drought in 2022 reduced the ability of European forests to pull about 60 Tg of carbon out of the atmosphere in the peak growing season. As the climate anomaly continued into the fall, forests grew more quickly, resulting in an annual reduction in carbon sequestration of about 40 Tg. For context, that’s a little less carbon than Spain emits in a year, and a little more than the Netherlands. This is crucial information to keep in mind for anyone working with carbon credits or offsets that have anything to do with forests.

The second major point of this research is the massive network of scientists necessary to make the observations behind this paper. Scientists relied on precise local measurements at sites distributed across a continent, all to be able to determine the impacts of one drought just on forest carbon dynamics. The world desperately needs the kind of information that only a global network of scientists can provide. Again, anyone working with managing, validating, or selling carbon credits should keep this in mind.

Making Carbon Offsets Genuinely Effective

A recent paper in Science (West et al. 2023, pdf available here) reviewed the effectiveness of projects intended to reduce greenhouse gas emissions by avoiding deforestation, and the results were discouraging. One common way that carbon offsets work is by paying land managers to leave their forests in place, with their trees unharvested. This is called an “avoided deforestation” project and is a common way that carbon offsets are sold. In 2021, for example, the trading market for such projects was about $1.3 billion.

The authors found that most of the projects they assessed failed to reduce deforestation, meaning that even though they were sold as a means of offsetting carbon emissions elsewhere in the world, the carbon associated with logging these forests was still released into the atmosphere. And even for the projects that reduced deforestation, the reductions were much lower than initially claimed when the offsets were sold. The projects that they looked at were all in tropical forests in Peru, Colombia, the Democratic Republic of Congo, Tanzania, Zambia, and Cambodia. Of these countries, Peru had the best record of successfully avoiding deforestation – half of their projects had a discernible effect. There was no evidence that the African projects were successful, and success was poor in Colombia and Cambodia.

Additionally, the effectiveness of the projects, even in Peru, was far less than was hoped before the projects, and far less than the project managers claimed. In the end, 71% of the offsets sold from these projects had no significant effect on deforestation. Of the remaining offsets (29%), the impact on reducing deforestation was so marginal that they only had the emissions-reducing equivalency of about 6% of the overall projects. Essentially, for clients who bought those emissions offsets, they only got about 6% of the intended effect. This is like paying for a gallon of milk and getting a little less than a cup. Except that in this case, the point isn’t to have milk to drink; the point here is to prevent worsening an ongoing crises. For anyone trying to use these offsets, unfortunately, the crisis was not averted.

The authors don’t point to many potential solutions, noting only that with greater scrutiny and monitoring, perhaps only validated projects should be traded in voluntary carbon markets. I agree that more rigorous assessment of project performance is crucial. And this study also indicates that other approaches than avoided deforestation carbon offset projects will be necessary to address climate change.

Predicting Major Changes in Ocean Circulation

A recent paper in Nature Communications about a potential state-change in the Atlantic meridional overturning circulation (AMOC) has been getting some attention, and I wanted to dig into what the authors did.

First of all, what is the AMOC? It’s not the Gulf Stream current, which is a flow of water on the surface of the Atlantic Ocean. The Gulf Stream brings warm water north along the east coast of North America, all the way to Europe, and is the main reason why the climate is London is more temperate than the climate in, say, Calgary (which is at the same latitude). The AMOC, on the other hand, moves warm water and nutrients throughout the Atlantic Ocean in a huge loop, from north to south and back again. A complete journey takes about 1,000 years. And that time has been increasing over the last hundred years as the AMOC has slowed.

As with many major earth system processes, it would be pretty bad for a lot of people if the AMOC were to change substantially, let alone stop completely. Increasing meltwater from polar regions could shift the AMOC, causing major changes in where, when, and how much precipitation happens. Places that are great for agriculture now could become to cold or dry if the AMOC were to shift or stop. This has been a focus of research for a long time, so none of this is really news within the scientific community.

With that background, a new paper uses statistical modeling to determine when the AMOC might collapse. Their basic answer is that, using their statistics, the average time at which the AMOC will collapse is 2050, and the 95% confidence interval runs from 2025 to 2095. It’s hard to think intuitively about things like 95% confidence intervals. Accepting it at face value, it’s possible that the AMOC could collapse outside that interval, say, in 2100… or 2023!

But we shouldn’t accept that finding at face value. A statistical model like this makes sense when we have little information about the processes at play. We’ve never seen a major oceanic overturning circulation collapse, so it might make sense to claim ignorance and turn to a statistical model. But we actually know a lot about climate and planetary processes, and have a ton of measurements that aren’t a part of this statistical modeling approach. As the authors noted, when other scientists have modeled this phenomenon, either for the Intergovernmental Panel on Climate Change or the Climate Model Intercomparison Project, they have determined with moderate confidence that the AMOC will slow through the 21st century, but persist. Those projections are based on modeling projects that attempt to simulate how the oceans and atmosphere move around the planet. Provided that the people doing that work have some expertise (which they do), they will almost certainly come up with a better idea of what’s going to happen that scientists who are only relying on statistical correlations.

It’s overblown to move from statistical correlations to project a state change outside of anything represented in the data. This is an example of the statistical sin of projecting beyond the range of your regression. Nevertheless, the authors do show that there is a high degree of variation in key metrics associated with AMOC. As other authors have found, this variability itself will contribute to extreme weather events as climate changes. I’m glad that the authors did this work, if for no other reason than the increased attention it will bring to climate challenges we are all facing.

Reality-Based Carbon Markets

I wanted to share an interesting article I read recently, and some thoughts that I’ve had about its implications for carbon markets. Damien Beillouin, from the University of Montpellier, along with many colleagues, recently completed an expansive meta-analysis of over 25,000 primary studies, all looking at soil organic carbon. I know that it’s tempting to look at something like a forest and think that the story about carbon is in the trees. But for many forests, and many other terrestrial ecosystems, most of the carbon in the forest is belowground, in the soil. So considering the dynamics of soil carbon is crucial to understanding climate change, and evaluating the potential effectiveness of many of the commercial applications that seek to curb climate change and its catastrophic impacts.

The authors had three main findings. First, that converting land to agriculture leads to a lot of carbon moving from the soil to the atmosphere. That has been a longstanding finding, so I’m glad to see that it has held up over time. Second, land management application in forests generally lead to carbon moving from the soil to the atmosphere, but this is highly variable. More on this in a moment. And third, wildland fire can lead to huge and sudden movement of carbon from the soil to the atmosphere. Again, this isn’t surprising to me. Sure, warmer temperatures may lead to increased decomposition, but if a fire burns through an area, a lot of carbon can get burned away suddenly, and without as much vegetation, a great deal of carbon can leave the soil over the next year or two.

How about those land management applications? Interestingly, these application may include deliberately putting biochar into the soil, which is becoming more common as a carbon mitigation techniques, and more popular in carbon markets. The study finds that biochar application can work very well. On average, it can lead to a gain in soil carbon of 67%, which is remarkable. It can be most effective when there is abundant biomass available nearby for conversion to biochar, and where that biomass isn’t destined for other long-lived sequestration.

On the other hand, one common carbon mitigation tactic is agroforestry, where trees are planted, or planted in croplands. And the authors find that this is much less effective at storing carbon in the soil for the long-term. Likewise, crop rotation and diversification were not especially effective at storing carbon in the soil. But diversifying tree species in forestry made a small and desirable difference in soil carbon.

Some treatments proved to be absolutely misguided, such as removing dead leaves, stems, or branches from forests or grasslands. Their Figure 3 should be mandatory reading for any policy-maker, anyone buying or validating carbon credits, and anyone who works on greenhouse gas emissions in ESG. But the short version is this: in forestry, plant a mixture of species; in grasslands, apply biochar.

Hyperbolic Scientific Claims

I want to share an interesting article I read, and dig through its evidence and reasoning.

Writing in The American Journal of Medicine, Armin Alaedini, Mary Heinricher, and Gary Wormser provided a couple recent examples of scientists making huge claims in their publications that turned out not to be justified. It’s not that those claims were later shown to be fraudulent – there isn’t malfeasance of that sort at play here. (If you’re interested in schadenfreude about that sort of thing, may I offer you this article about the researcher who studies public dishonesty… making up data?) Instead, the data in those papers themselves did not support the conclusions the authors drew. Hoisted by their own petards!

First up was a paper from a group at Yale that claimed to show sex differences in immune responses to COVID-19. A year later, in the same journal, a group of public health experts reexamined the data from that paper and found that, no, in fact the reported sex differences were not present in the dataset.

Next up: a claim that a particular SARS-CoV2 spike protein persists in the blood of people with long COVID. And, counterpoint: no it does not. (This was is a little less clear to me, in no small part because the work is in pre-prints and comments. Also this editorial comes from the first author of the AJM article that I’m considering here.)

Alaedini and colleagues explain that scientists are likely engaged in a process of marketing and selling their results. There are incentives in academic publishing to get one’s work into the most prestigious journals. Doing so usually requires novel, surprising results. The authors also place blame on journalists, who may take bold claims by scientific articles, and then uncritically amplify them into even more grandiose headlines and articles in pursuit of their own perverse incentives. This tendency toward sensationalism and hyperbole is bad for all of us, because so much of what our society does rests on a foundation of scientific information.

This whole topic reminded me of something I heard Arthur Miller say in an interview:

“It seems to me the whole United States is engaged in selling. Most of the world recognizes what’s at the bottom of that idea; namely trying to become great without being great, which is what most of us are doing.”

Alaedini and colleagues then make a further claim, that these tendencies toward inflating scientific claims “contribute to the weakening of public trust in science and scientists, which is a serious and growing problem.” I’ve written about this in the past, and it’s worth repeating: Americans trust scientists a lot, more than pretty much anyone else. What evidence do the authors provide to show a decline in trust in science? This article from AAMC, which starts from a headline about trust in science, proceeds to a subheading about trust in medical science, and then, when it comes time to cite evidence, refers to the same Pew polls that I wrote about. In discussing that evidence, the article reports that trust in science remains high. The claim of bloated headlines that the authors are complaining about is present in the evidence that those authors bring to bear in support of their assertions! That doesn’t mean that the authors are wrong about their overall claim, even if they are wrong about this component of their argument. And I take delight in finding how deep this problem goes.

Nevertheless, the authors’ proposed solutions are sound: remove the perverse incentives, incentivize peer review, “improve science journalism” (sure, and, that’s a big task!), and change our culture to better understand and share science’s values. That last point is one that I agree with. It’s also a huge task, and a topic for another day.

Remembering Cormac McCarthy

I was sad reading this remembrance of Cormac McCarthy today. He was a writer who added a lot of meaning to my life over the years. His writing, both rich and terse, is a great antidote to reading the scientific literature. A couple of things from the linked article stood out to me.

McCarthy started writing in the 60s, and published his first novel in 1965. He never sold more than 5,000 copies of a book until he published All the Pretty Horses in 1992. That’s almost 30 years that he spent doing something with basically no financial success. During that time, he refused to give interviews, write blurbs for other authors, or essentially do any of the networking or marketing or what we would now call personal branding work that is usually a necessary component of a successful career. We could say that the exception proves the rule. But I always love seeing the exception in practice.

Once he was successful, he was able to sell some of his accessories for lavish prices. He sold his typewriter that he used to write all of his work. At auction, the typewriter brought in about $250,000. Then he turned around and bought the same model of typewriter, used, for $20.

What can we take from these two details of his life? First, keep practicing your craft. And second, don’t mistake the tool for the artisan.

Rest in peace, Cormac McCarthy.

Academic Science and Mental Health

I noticed a lot of people talking about mental health in academic science recently, probably because of this recent summary from Nature. I can certainly identify with some of the stressors mentioned in the article. For my friends who are still in academia, I highly recommend reading the article, and thinking about positive next steps they could take for their mental health. As the article notes, even just getting beyond the taboo of discussing mental well-being would be a major step forward. Major structural changes are of course the solution, as usual, for systemic problems. But as individuals, taking time to focus on relationships and physical health is always a great place to start. When I started graduate school, I recall a staff psychologist from the university addressing all of us incoming graduate students, and forcing us to parrot back to him: “I get more work done when I make time for work AND play.” It’s still true.

I thought it would be fun to dig into some of the research behind Nature’s news summary. Figure 4 from Jacobi et al. 2022 (pdf) was especially interesting. Those researchers looked at the correspondence between aesthetic experiences and various measures of well-being. They found, unsurprisingly, that the frequency of aesthetic experiences they were interested in – feelings of awe, or noticing the beauty of a research project or result – was associated with stronger relationships and a more meaningful life. At its best, scientific research definitely provided me with those moments. What’s unclear is the direction of causation. So continue to look for the beautiful results in your work. And at the same time, keep up with your personal relationships.

Current Job Search Status

This is less of something that I read, and more about making figures to entertain myself while I look for a new job.

I’ve been seeing a lot of figures on the Internet of how many jobs people have applied for, with the many hoops they have had to jump through to finally find a job. Many people like to show these data in a sankey chart, which I made with sankeymatic and looks like this.

It’s a nice looking figure, but it’s missing what has been the key element for me: time. I have spent SO LONG waiting to hear back from places I have applied. I wanted to make a figure that shows those wait times. The sankey chart also loses the information about the pathway of each individual application, and I wanted to preserve that. Here’s a figure I’m satisfied with.

I made this in R (4.3.0) and R studio (2023.03.1), working as usual with ggplot2 (3.4.2). I posted my code here. I like to write my code with a lot of annotations. In theory this makes it easier for other people to read. Really it’s just so that I can more easily cannibalize my own code for my next project. Annotations aren’t for someone else – they’re for time travelling you!

Visualizing the data this way, it’s easy to see many interesting things. For example, most of the jobs I applied for haven’t gotten back to me. For some, that’s because I applied recently. But there are others (I’m looking at you, Job 7!) where I’ve been waiting a long time without hearing anything. Then on the other end of the spectrum, there’s Job 10, which let me know the day after I applied that they had rejected me. That’s the kind of company I want to work for – they knew I was a bad fit, and told me right away, so I didn’t spend two months in anxious limbo. You can also notice job searches that moved through interviews quickly (Job 5) or slowly (Job 1).

I had a good time tinkering with the details of this figure to make it work well. This included doing things like setting meaningful tick marks on the x axis, inverting the y axis, and choosing custom colors that should still work for people who have colorblindness.

But all good things must come to an end, and now I have to stop working with data and go back to writing cover letters.