Kristine N is a graduate student in the Earth and Atmospheric sciences department at Purdue University . She recieved a B.S. in Geology from Caltech and an M.S. in Geology from Penn State. Her master’s research, conducted with Kate Freeman and Chuck Fisher, focused on the sulfur stable isotope geochemistry and lipid geochemistry of sediments associated with vestimentiferan tubeworms in the Gulf of Mexico (which sounds far more pretentious than it really is). She is currently growing Sea monkeys for her PhD in an effort to reconstruct drought freqency and intensity in the Great Basin.
Ben Santer works on identifying human influences on global climate. He and Tom Wigley co-authored the eighth chapter of the second IPCC assessment, which included the controversial statement, “The balance of evidence suggests a discernible human influence on global climate.” Having survived the maligning of climate skeptics, the third IPCC assessment, and the most recent CCSP report, which was released in May of this year, he spoke to a group of us at Purdue University. Today I am presenting, for your edification and entertainment, my notes from his talk.
Climate varies naturally, and would change over time whether humans were present or not. Solar input is the most important climate forcing (a forcing is a mechanism that influences climate). The energy balance, or radiative forcing of the climate system depend on solar output and on the amount of energy absorbed or reflected by the earth’s atmosphere and surface. Orbital cycles, or Milankovitch cycles produce the pattern of long ice ages punctuated by short interglacial periods that characterizes the Quaternary period by changing the radiative forcing of the earth. On a shorter time scale, solar output varies due to the sunspot cycle, which also impacts global temperatures. Volcanic eruptions are another significant climate forcing. Increases in volcanic dust and in the concentration of sulfates in the stratosphere increase the albedo (reflectivity) of the earth. If the earth reflects more solar radiation, temperatures decrease, as was observed in the year without a summer after the explosion of Krakatoa. Solar output and volcanic dust are considered natural external forcings. Variability inherent to the system is also produced by internal forcings, such as the interactions between the atmosphere and ocean that produce El Nino and La Nina cycles.
Human activities produce climate forcings. We have demonstrably altered the concentrations of carbon dioxide, a gas that absorbs infrared radiation, in the atmosphere. The Keeling curve is considered the most famous figure in all of geophysics and shows the impact of burning fossil fuels on atmospheric CO2 concentrations. Biomass burning, ore smelting, and other combustion-type activities produce aerosols that increase the albedo of the atmosphere. As we alter ground cover by cutting down forests, growing crops. and building cities and roads we change the albedo of the ground, again impacting the energy balance of the planet.
We know we produce forcings, but do not know the magnitude of these forcings. Unfortunately, we don’t have another parallel earth sans people acting as a control that would allow us to quantify the magnitude of human-induced forcings. To get around this limitation, we use models. Models are a tool we have to predict climate change signals. They are simplifications of a complex and chaotic system, and every model posesses sytematic errors that must be taken into account. Even if it were possible to create a model with perfect physics, because climate is a chaotic system and we don’t know exactly the initial state of the earth’s climate we would see a mismatch between the modeled and observed climate. Modelers thus perform multiple realizations, starting with slightly different initial conditions, and then average the results, producing a model with surprisingly high skill and an “envelope” of the variability around the model. Models are tested against data over many time scales–from recent, daily or even hourly instrumental records, to longer paleoclimate records produced by proxy reconstruction.
Agreement between modeled and observed climate is considered evidence the model is “good;” however, mismatch between a model and observations can be just as informative. For example, in 1982 the eruption of Mt. Chichon in Mexico produced copious amounts of sulfur dioxide, which should cool climate. At the same time, the strongest recorded El Nino was occurring. The conflucence of the two events resulted in less cooling than expected based on known climate sensitivity to volcanic eruptions. The mismatch between the modeled response and observation provides information on the relative importance of each of these forcings on the climate system. Any forcing can be left out of a model intentionally and compared to observations or to other models to test the relative contribution of the forcing on the climate system. Estimates from such comparative studies suggest the contribution of human forcings is responsible for much of the warming observed over the past 50 years.
Different forcing produce unique warming and cooling patterns, or “fingerprints.” For example, increases in solar radiation warm the entire atmosphere, while increased greenhouse gases warm the troposphere and cool the stratosphere. These fingerprints provide testable hypotheses–we can test predicted patterns based on physics against observations and say which of the forcings gives the most likely explanation. According to Santer, increased height of the troposphere predicted by models matches observations, supporting the theory that increased greenhouse gas concentrations are responsible for the observed 0.6C to 0.8C increase in average global temperature.
Climate skeptics freqently point out that not all observations are consistent with a warmer troposphere. Atmospheric temperature profiles collected by radiosonde and satellite do not show the same warming trend observed in surface temperature records. In the last year, papers by Mears and Wentz, Sherwood, Lanzante, and Meyer, and by Santer and colleagues suggest corrections applied to the original data, daytime solar heating of the instrument, and uncertainties in the data explain the observed lack of warming.
We know human activities have changed the composition of the atmosphere. Concentrations of carbon dioxide have increased about 30% since 1850 and are today higher than they have been in the past 650,000 years (sorry, graph is only 400,000 years) , and we know the 20th century is the warmest century of the past 2,000 years. We have seen surface temperatures increase 0.6C to 0.8C since the mid 1800′s. We see the fingerprints of human activities in multiple aspects of the climate system. Together, we are presented with a consistent story–human activities are influencing global climate.
What we don’t know is what our future climate will look like. We don’t know how warm we will be in 2100, in part because we don’t understand every possible climate feedback, and in part because we as a people haven’t decided what to do about CO2 emissions. We aren’t sure how the carbon cycle will respond to increased CO2, or what temperature changes will do to thermohaline circulation in the ocean. Climate does not always respond linearly to forcings; there are “tipping points” where the climate will shift from one equilibrium state to another over a very short time period. We don’t know how the frequency and intensity of extreme climate events, like hurricanes and tornadoes, will change in a warmer world, and we are only now starting to investigate regional features of climate change (see here for a recent Purdue publication on a region near and dear to my own heart).
Thanks to Dr. Santer for letting me post these notes.