Climate Proxy Datasets: How Do We Know How Sea Levels Have Changed?

Can Rock and Ice Cores be Used as Proxies to Quantify Past Climate Change?

As mentioned on the last page, geologists are exploring the connections between sedimentary rocks and ice cores to help evaluate past climate change and associated sea levels both locally and globally. In order to evaluate this a little further, we will explore briefly some of the connections they are making.

The Vostok ice core from Antartica has been analyzed for its carbon dioxide and methane concentrations. Temperature is calculated from measuring ratios of oxygen isotopes (oxygen 18 and oxygen 16) that were trapped in air bubbles within the layers of ice. These isotopic ratios change according to the global temperatures in place during the interval when each layer of snow was deposited. More info about the ice core record can be found on the CDIAC website if you are interested.

A great site for visualizing the data collected from the Vostok Ice Core and other related climate proxies has been beautifully organized into an informative, interactive website which you are encouraged to explore on your own if you are interested. The data is presented on the Temecula Valley Bright Stars website. When you look at the data you will see that it is organized in such a way that you can toggle individual variables to compare each in turn, or all at the same time. A couple of pointers if you spend some time looking at the site. Remember, temperature is a temperature anomaly relative to modern average temperatures and modern oxygen isotopic compositions.

×

Figure 4.29: Screen shot for the last 20,000 years of paleoclimate data from ice cores and other proxies. This output shows more recent data on the right and prehistoric data on the left. Specific climate factors are shown and labeled on the right along the y‐axis. These include Milankovitch's orbital parameters, variations in incoming solar radiation (insolation), methane (CH4) concentrations, carbon dioxide (CO2), and temperature anomaly data. Note that any of these parameters can be toggled on/off by clicking on the buttons on the left hand side of the screen.

Credit: Screenshot is from the Temecula Valley Bright Stars website.

What you are seeing in Figure 4.29 is that, at 20,000 years (left side of the graph) average temperatures would have been some 5 to 6 degrees Celsius lower than modern levels. Temperatures then began to rise (with some up-down wiggles) so that by around 11,000 years ago temperatures may have reached modern average temperatures. This interval of time roughly correlates with the end of the Pleistocene and the beginning of the Holocene Epoch.

Since the beginning of the Holocene, temperatures have varied a little bit (ranging from approximately -1.5 degrees Celsius to about +1.0 degree Celsius), but, in general, much of the Holocene has shown average temperatures below the modern average. When looking at the methane and carbon dioxide records, the number of data points are too few to evaluate high-resolution changes, so we can only look at the longer term trends. Other websites report some of these higher resolution data and explore some of the controversies surrounding the linkages between temperature and greenhouse gases, for instance see this Joanne Nova webpage and the scientific references therein for more info.

We would venture that temperature variation appears to track changes in carbon dioxide and to a lesser degree changes in methane concentrations. Although cause and effect cannot be established from these data, temperature co-varies with changes in greenhouse gas concentrations such that warming occurred as greenhouse gas concentrations went up.

Given that our last glacial maximum occurred between 20,000 and 13,000 years ago, and that global temperatures began to rise to near modern values after this time, we still need to think about how this warming may have occurred. One variable that we haven't looked at in detail is the insolation factor. In our graph above, you were asked to turn on the insolation estimates for 65 degrees N. So, what does this variable tell us, if anything?I

Insolation and Recent Climate Change?

Today, as in the Pleistocene, the biggest proportion of land is located between 30 to 60 degrees. By implication, as white reflective glaciers melt in response to warming and expose dark soils and rock, it is possible that changes in the amount of incoming solar radiation (and albedo) could be a driving factor in melting glacial ice and causing sea level rise. Check out the USGS Repeat Photography Project that compares photos of different glaciers today relative to historic photos. This will help you understand visually what changes in albedo can look like. There are many dramatic photo comparisons, but see the Chaney Glacier in Glacier National Park as a representative example. Located in the northern Rocky Mountains in Glacier National Park, Chaney Glacier had already begun to retreat by the early 1900s. The striking piece is that by 2005 the glacier is almost non-existent, so much so that many scientists worry that in another decade or two, Glacier National park may no longer have any glaciers.

Exploring our dataset above shows us that Earth saw larger incoming solar radiation values (at all northern latitudes), and likely reduced albedo (reflection of insolation). In our dataset, summertime (June) insolation (measured in watts per square meter) began to rise during the period that temperatures began to rise and preceded the rise in greenhouse gases.

Interestingly, northern latitude insolation began to fall at around 11,000 years ago, about the time we reached near-modern temperatures. At the same time, greenhouse gas concentrations became elevated and remained high.Today we see insolation values at all northern latitudes (where most land is located) are near low points, but are beginning to rise once again, suggesting that more energy will be delivered to these latitudes in the future. These observations should raise some questions on your part! Keep writing them down in your notebook, and stay engaged with your instructor and the rest of the class!

One thing that is clear here is that there are lag and offsets in data set responses that need to be explored further. Are these lags real, or are they related to errors in alignment/calibrations between different datasets? We can't answer this question here, but Joanne Nova does explore this question if you are interested in her website.