Mesoscale effects on carbon export paper just accepted to GBC

See link to accepted manuscript here:

Most Earth System models (used to study the global climate including the carbon cycle) do not resolve the most energetic scales in the the ocean, the mesoscale (10-100 km), encompassing all of the eddies and jets in the ocean. One key question is does this ocean mesoscale circulation affect the global carbon cycle? In particular, does the export of carbon from the surface to the deep ocean, a process that is responsible for removing a significant fraction of CO2 from the atmosphere, change if the ocean circulation is better represented? We found that the while the global integral of carbon export is similar between a mesoscale-resolving (0.1º) and a standard resolution (1º) Earth System model (<2% difference), regional differences are up to +/-50%, and occur through a number of processes. One key result is that in high-latitude regions where biological production is driven by natural iron fertilization from coastal sediment sources, the resolution of both coastal jets and mesoscale turbulence reduce iron delivery from the coasts signifigantly, limiting biological production and the resulting carbon export. The figure below (left) shows the iron concentration leaking through at the confluence of the Brazil and Malvinas currents off the Patagonian shelf, into the iron-limited waters of the South Atlantic sector of the Southern Ocean. This results in one of the largest, shortest phytoplankton blooms in the Earth’s ocean, shown here in December, late Austral Spring, where the phytoplankton abundance is shown in iron (Fe) units and in log scale. Note that white areas in the left plot are where the phytoplankton concentration in the right plot is high; here the plankton have rapidly used all of the iron that is available, as the Southern Ocean is rich in nitrate and other nutrients. The large bloom to the south, off Antarctica, is due to deep winter vertical mixing, while the blooms off South America and South Africa are driven by horizontal transport and local upwelling. Additionally, one can see the iron fertilization effects of islands and seamounts throughout the region. You can see an animation of this bloom here:

Screen Shot 2018-02-23 at 1.49.57 PM.png

How does mesoscale transport of iron affect carbon export? Read the paper for the full story!  In the below plot, the top panel shows carbon export production (EP) at high resolution (top) and low resolution (bottom), with an inset of the comparative statistics for the two boxed regions. The short answer is that restriction of the iron transport by the mesoscale flow reduces the production and export in this region signifigantly, while also increasing the spatial and temporal variability. This has implications for observational campaigns, highlighting the difficulty in validating modeled carbon export for short, seasonal and turbulent blooms, which represent a large portion of the global carbon export.





Video on Biases in Academic Settings

On ESWN someone posted this great short video from Ohio State U on how biases affect assessment. This includes when we write letters of recommendation. are hiring, and really anytime we do assessment. We all do this unconsciously, and studies show just being aware can help us overcome our biases’ effect on our behavior. A great resource.

Historical impacts of volcanic forcing: famines throughout the Little Ice Age

Tonight and tomorrow we are having a symposium at CU Boulder on the Little Ice Age entitled “The Coldest Centuries in 8000 years: The Little Ice Age Causes and Human Consequence.” The keynote speech is given by one of my collaborators in the nuclear war modeling project, Alan Robock. The Little Ice Age (LIA) was a period of global cooling lasting from roughly 1250 to 1850 CE (Fig. 1). In preparation for this event I have been doing some reading on what we know about the LIA in terms of the climate proxy and historical records, as well as our understanding of what started it and kept it going. In particular, I am interested in the idea that the cause was a series of large volcanic eruptions, and how those might compare with nuclear war in terms of climate and human impacts.

Fig 1. Temperature reconstructions over the last 2000 years, showing the cold anomaly termed the “Little Ice Age”. Source: wikipedia , see link for description of records plotted. Note that the temperature anomaly has a different magnitude in different reconstructions.

The beginning of the Little Ice Age coincided with a number of large volcanic eruptions, recorded in ice cores from Antarctica and Greenland (Fig. 2). One idea is that the cluster of eruptions around 1250 CE initiated a global cooling event, which lead to changes in the global ocean overturning circulation (i.e. AMOC), leading to a climate feedback. Honestly I have yet to wrap my head around how this feedback works.

Fig 2. Records of volcanic eruptions in Antarctic ice cores. From Sigl et al. 2014

One of the interesting things about this era is that we have historical records of the human impacts of this global cooling, and of the individual volcanic eruptions. One map I found particularly fascinating links the large eruptions throughout the LIA to various famines (Fig 3). The stories from these famines are quite disturbing, from legalization of slavery in Japan during the Kangi Famine, where families would sell off one of their members to get food for the rest, to cannibalism and infanticide. Often famines would make populations more susceptible to disease, and it is thought that the “Dark Ages” in Europe were largely an effect of the LIA cooling.


Fig. 3. Links between various volcanic eruptions in the LIA and human famines. From PAGES magazine

To put this in perspective to what we might expect from nuclear war, the temperature reconstructions suggest that cooling was at most -1ºC during the LIA (Fig 1), where the climate simulations of Mills et al. 2014 suggest that the global cooling effect of regional nuclear war could more than this (Fig 3). Global nuclear war would have an even larger cooling effect, as it is expected to deliver more light-blocking aerosols to the stratosphere.


Fig 4. Atmospheric aerosol mass burden (soot in atmosphere), anomalies of global mean solar flux, temperature, and precipitation in a number of nuclear war model simulations. From Mills et al. 2014. Note that the maximum temperature anomaly is greater than 1ºC/K, larger that what is recorded for the LIA, and lasts for a number of years.

Would nuclear war have the same climate-related impacts on humans as volcanic eruptions in the past? Our farming methods have much improved, with much higher yields per acre, as has our ability to store and distribute food surplus. However, there would likely be very large disruptions in weather and food supply, resulting from crop failures, which would be expected to have unpredictable regional and global political impacts. The losers would likely be poorer regions and nations, especially those relying on subsistence farming, as is the case for any sort of climate disturbance. The regional and large-scale impacts of recent refugee crises in Syria, Central America and Myanmar come to mind. Though these were driven by political unrest, we might expect similar levels of migration during climate disturbances.


First update on Nuclear War project

In short, the project is applying a state-of-the-art earth system modeling framework to study the climate implications of nuclear war (aka nuclear winter), funded by the Open Philanthropy Project.

The science summary is this: Like volcanoes and asteroid impacts, regional nuclear war and the resulting firestorms are expected to inject aerosols like soot into the stratosphere, where the lack of precipitation leads to residence times from years to decades. The soot (aka black carbon) absorbs radiation, blocking light to the surface, causing global cooling and an associated decrease in precipitation. At the same time, heating of the stratosphere destroys ozone, potentially creating a global ozone hole, leading to high levels of UV exposure for plants and animals world-wide. In the historical record, volcanic eruptions, causing less cooling than we expect for even a regional nuclear war, have resulted in climate perturbations leading to massive famines, riots and disease outbreak, e.g. “The Year Without a Summer” in 1816. Regional nuclear war is expected to lead to much more extensive global cooling, and could result in a global nuclear famine. Global nuclear war would likely lead to global temperatures far below the last ice age, and the cooling is expected to last for decades.

The details: In this study we are using expert knowledge of likely nuclear conflict scenarios to drive regional urban fire models, the output of which will be plugged into state-of-the art global earth system models as a climate forcing perturbation. The resulting impacts on temperature, precipitation, crops, ocean biogeochemistry and fisheries will be studied. The urban fire modeling is building on recent advances in urban wind modeling, and a state of the art fire model, which will be combined for the first time. The earth systems modeling effort is building on recent studies of asteroid impact effects on global climate (Toon et al. 2016Bardeen et al. 2017), previous nuclear war modeling work by Mills et al. 2014, the Last Millennium Ensemble, which includes the effects of volcanic eruptions on climate, and concurrent efforts by myself and Clay Tabor to look at ocean biogeochemical effects of the K-Pg asteroid impact that killed the dinosaurs, as well as work I am doing with Samantha Stevens to look at climate variability (such as ENSO) effects on fisheries. The effects of volcanic, asteroid, and nuclear forcing on ocean biogeochemistry have not yet been explored.

In July we had our first meeting with nuclear policy experts. One of the likely scenarios our experts proposed was that these weapons could be easily hacked. Bruce Blair said something to the effect of: “We could go tonight, and I could show you where we could pitch a tent, here in Colorado, and hack into the data lines leading to NORAD.” He also reported that it is confirmed that Russia has an unmanned sub with a mega, dirty nuclear bomb, so that even with a “first strike” effort by another power (i.e. the US) to take out all nuclear capabilities of an opposing country could still result in a severe retaliation. This is beside the point that a first strike could result in a global climate crisis, in effect self-assured destruction. One of our issues as a group is do we try to figure out at what level nuclear bombing is not a global catastrophe, or do we focus more on the potentially harmful effects. This is a bit of a moral conundrum I feel ill-prepared for.

Nuclear war project in the news

This Fall I am moving down to CU Boulder as research faculty where I will be working on the impact of extreme forcing events on ocean circulation, fisheries, and biogeochemistry. This includes forcing events such as volcanoes, asteroids, and in particular, nuclear bombs.

The main funding for this work comes from a large private foundation grant to study the effects of nuclear war on the earth system.

Brian Toon and I were recently interviewed on Colorado Public Radio about this project:

Here is a press release:


Researchers to study environmental, human impacts of nuclear war

Scientists and students led by CU Boulder and Rutgers University are calculating the environmental and human impacts of a potential nuclear war using the most sophisticated scientific tools available.

The lead researchers, CU Boulder Professor Brian Toon and Rutgers Professor Alan Robock, have been studying the threat in-depth for decades. They were among the first scientists to formulate the “nuclear winter” theory, which indicated that a nuclear war between two countries could cool parts of the planet and trigger famine and mass starvation, even in nations not involved in the war.

In 1983, Toon, Robock and others—including Cornell University’s Carl Sagan—followed on the heels of Dutch scientist Paul Crutzen and now-retired CU Boulder professor John Birks, who published a 1982 study concluding that smoke from burning forests, cities and oil reserves caused by nuclear blasts would block sunlight and cool Earth. The 1983 nuclear winter paper, which Toon co-authored, was published in Science and grabbed worldwide attention from scientists, politicians and the public.

“I find it surprising and frustrating that the potential catastrophic effects of a nuclear war have fallen off the radar of many people following the nuclear winter discussions that began in the early 1980s,” said Toon. “One of our goals for this study is inform people how dangerous these weapons are by providing a solid scientific analysis of the issues.”

Ongoing threat of ‘nuclear winter’

The new study will calculate in detail for the first time the impacts of nuclear war on agriculture and the oceanic food chain and on humans, including food availability and migration activity, said Toon of CU Boulder’s Laboratory for Atmospheric and Space Physics. The team is using various scenarios to calculate how much smoke produced by fires in modern cities initiated by nuclear blasts would be produced by urban firestorms and their available fuels, Toon said.

“The most important factor is the amount of smoke which would be generated from fires started by nuclear detonations in cities and industrial areas and lofted into the upper atmosphere,” said Robock, a distinguished professor in the Department of Environmental Sciences at Rutgers University-New Brunswick. “For the first time, we will model the fires and firestorms, using detailed estimates of what would burn, based on new credible scenarios of how a nuclear war might be fought.”

Although the global nuclear arsenal was reduced by about 75 percent following the end of the Cold War in the 1980s, there are still about 15,000 nuclear weapons distributed among nine nations. While the United States and Russia have the bulk of the weapons, the other members of the world’s “Nuclear Club” are Britain, China, France, Israel, Pakistan, India and North Korea.

Toon stressed the threats of a nuclear incident have not diminished and could arise from miscommunications, international panic, computer hacking or malfunction, terrorism or action by a rogue leader of a nuclear nation. North Korea, which has 10 to 20 nuclear weapons, continues to flaunt its military power—most recently with the launch of an intercontinental ballistic missile believed capable of reaching Alaska or Hawaii—that was condemned by many nations, including the U.S., Russia and China.

Supercomputers, climate models tell a nuclear war story

The team is using supercomputers and sophisticated climate models developed by the National Center for Atmospheric Research (NCAR) in Boulder to calculate the amount of fire fuels in major cities and how much smoke might be produced by nuclear blasts. The researchers also are using agricultural and world food trade models to assess the impact on crops from a potential nuclear war and the possibility of widespread famine.

“Calculations show there is enough food on the planet to feed people for about 60 days, and an average city has about enough food to feed residents for just seven days,” said Toon, also a professor in the atmospheric and oceanic sciences department (ATOC). “The functioning of our society is based in large part on our ability to transport food, fuel and other goods—activities that would be severely affected by a nuclear war.”

In 2016 Robock and Toon authored a commentary piece in The New York Times titled “Let’s End the Peril of a Nuclear Winter. In it they point to their 2007 study on the potential impact of a nuclear war between India and Pakistan, with each country detonating 50 Hiroshima-sized bombs.

One result? Smoke from the explosions would make temperatures plunge, causing wheat, rice, corn and soybean production to be reduced globally by 10 to 40 percent for five years. The explosions also would cause severe depletion of the Earth’s ozone layer, damaging human health and the environment, said Toon.

The new project is funded by a three-year, $3 million grant from the Open Philanthropy Project headquartered in San Francisco. Open Philanthropy focuses on funding projects in four broad categories: U.S. policy, global catastrophic risks, scientific research, and global health and development.

Students involved in research

As part of the Open Philanthropy effort, CU Boulder Professor Yunping Xi and his students will assess the amount of flammable building material in modern cities in various parts of the world, as well as the flammable contents in such buildings. CU Boulder Professor Julie Lundquist and her students will use sophisticated weather research and forecasting models developed at NCAR to run computer simulations on how terrain and surface roughness might impact fire behavior after a nuclear detonation.

Robock is working with several graduate students, including Joshua Coupe, who will be helping on climate modeling. Another of his graduate students, Guangoh Jheong, will work on agricultural modeling with University of Chicago postdoctoral researcher Florian Zabel. Rutgers Associate Professor Gal Hochman and graduate student Hainan Zhang will work on economic modeling for the effort, said Robock.

NCAR scientists Charles Bardeen and Michael Mills, who both received their doctoral degrees at CU Boulder, will use the latest atmospheric and aerosol climate models developed at the center to better understand the response of the climate system to the soot from fires. Based on current scientific knowledge, some could end up in the stratosphere 10 to 30 miles above Earth’s surface and remain aloft for years or even decades, said Toon.

Working with Toon, Bardeen and Mills—who previously collaborated with him on nuclear winter simulations—will track the injections of gases and aerosols from city fires, calculating their transport, removal and the interaction of the particles with clouds, incoming sunlight and climate.

In addition, CU Boulder Assistant Professor Nicole Lovenduski, researcher Cheryl Harrison and students will be studying how the oceanic food chain might change in response to the climatic disruption and enhanced ultraviolet radiation from nuclear explosions.

“Our work will provide a much clearer description of the global humanitarian consequences backed up with state-of-the-art calculations of the fires, the climate change and the impact on food production, prices and restrictions for a number of different possible nuclear wars,” said Robock.

What diversity matters and how to achieve it

At the suggestion of a colleague, I have been adding a few slides about diversity at the end of my scientific talks. This is something I have been interested in since being an undergrad, where I was one of four women in my class of 50 or so in physics. There has been a lot of research on what works to increase diversity within academia and in general.

Here are the slides:

Screen Shot 2017-03-20 at 4.23.06 PM

Michael Faraday, who discovered electromagnetic fields, was born the son of an itinerant blacksmith in South London at a time when class stratification was intense. Apprenticed to a bookmaker in his early teens, he would take texts home at night to read them. After attending public physics lectures he would write up his own notes, and presented them to the lecturers. He only acquired a scientist position due to the imprisonment of Sir Humphry Davy’s assistant after a drunken brawl. We almost missed his contribution due to the lack of opportunity for someone like Faraday: poor and lower class. Historians of science tell us that because of his outsider’s view he was able to see physics in an entirely different way, thus making a discovery that changed the course of human history and enabled the world as we know it. This demonstrates how diversity of thinking breeds innovation, thus we need to nurture it by building an environment that is inclusive of all backgrounds.

Link to the BBC In Our Time podcast on Michael Faraday:

Harvard business review had a great article last year covering a meta-study of the outcome of diversity programs in mid to large size US companies. The results are that the programs everyone has been doing don’t work, and actually decrease diversity (have negative effects)



They also cover what does work:


Since mentoring is a big component of what does work, it is important to consider how to be a good mentor. Dena Samuels has a great book summarizing the research on what works in a classroom setting in effectively teach all students, and this research is easily extended to mentoring practices. Here is my summary of that research in pseudocode:


Link to the Harvard implicit bias test (IBT):

Link to Dena Samuel’s book “The Culturally Inclusive Educator” and website:–diversity-training-.html


Sim Turtle


I am simulating Loggerhead sea turtles in the North Pacific as part of Sim Turtle, a project to look at how ocean currents and food resources impact turtle hatchlings in their first year of life. Sea turtles are endangered, having low reproductive success (they are very tasty to predators), being susceptible to ocean bycatch by major fisheries (like tuna), and due to habitat destruction of their nesting beaches by human activities.

Here I have released simulated turtles over a few months from Yakushima Island, just south of the mainland of Japan at around 31N. This island is an ideal nesting area as it gets the hatchlings into the region north of the Kuroshio, the most productive region in the area. The Kuroshio is the Pacific version of the Gulf Stream, an ocean jet that brings warm water up the western edge of the Pacific. The Kurioshio brings high-iron, nitrate-limited waters from the south to mix with the high-nitrate, iron-limited waters in the north, driving a locally large phytoplankton (ocean algae) population. This phytoplankton concentration drives secondary production, including zooplankton and jellies, which the young turtles feed on.

We are simulating the baby turtles moving around in the global hi-res (0.1º resolution) CESM earth system model, which includes ocean ecosystem components like plankton. We simulate turtles as passive particles, moving with the ocean currents (for now), and eating the modeled plankton. In the figure, I am plotting each particle for five days, with the last day in green. Note the large concentration within eddies (vortices) that have pinched off the Kuroshio, and in the eddy trapped against the coast of Japan. One question is if this distribution of meanders and eddies is the same from year to year, and if the simulation is matching the observed structure of the Kuroshio.