Geology 101 teaches that the past is the key to the present.
This premise, however, depends on accurate interpretation of the past.
Accuracy becomes particularly important in paleoclimate analysis, which can be highly challenging because stratigraphy doesn’t always record the complete story.
Geologists historically have visited the outcrops making interpretations based on certain types of paleoclimate indicators, e.g., evaporites. Error is introduced when the climate being interpreted is assumed to apply for longer periods of time or to a broader area than it actually does.
Furthermore, merely keying in on certain paleoclimate markers fails to provide a picture of the complete suite of climate variability actually occurring.
Indeed, relying on the geographic occurrence of a single climate indicator in a certain area as evidence of past climate is rife with spatial and temporal generalizations.
“Geologic interpretations of past climate commonly do not occur at the time scale that climate actually varies,” said Martin Perlmutter, team leader of reservoir and seal prediction at Chevron. “This problem causes our interpretations of paleoclimates to be averaged at best and misleading at worst. Subsequent calibration of quantitative paleoclimate models to these interpretations wouldn’t be very useful.
Annual average does not always tell the present climate story, as shown in fig.1 and fig. 2. Some regions have large differences in seasonal rainfall -- a factor often overlooked.
“It’s like claiming the only thing you need to know about Houston weather is that the average temperature is 68 degrees and the humidity is 77 percent,” Perlmutter said. “All past periods have had highly varying climates that don’t always leave the most easily interpretable patterns.
“Many people think the mid-Cretaceous, because of the high CO2, was supposed to be less variable and more equable,” Perlmutter said, “and, indeed, it was a lot milder.
“But when you look at the range of paleoclimate indicators and run climate models at the right temporal scale, it had the same amount of variability as today,” he noted. “There were regions that varied from rainforests to deserts, and likely polar regions with annual snowfalls at certain points in the climate cycle.”
In fact, Perlmutter has a presentation planned for the upcoming AAPG meeting in Perth that will show:
- How large the variability was for the Cretaceous and the Permian periods.
- What this means for those who are modeling and interpreting climate.
- How this information is useful for predicting reservoirs and source rock potential.
Most paleoclimate analyses are resolved only for mean annual conditions at no less than the scale of eccentricity. Eccentricity causes the orbit of the earth around the sun to periodically vary from elliptical to almost circular with periodicities of about 100,000 years and 400,000 years.
However, the greatest changes in insolation -- the amount of solar radiation reaching the earth -- occur seasonally at the scale of precession, which has approximately 20,000-year cycles. These precession cycles can cause Northern and Southern hemisphere insolation to be about 10,000 years out of phase.
“Hot summers and cold winters in one hemisphere correspond to mild summers and mild winters in the other,” Perlmutter noted. “The pattern reverses itself over a precession cycle so that similar climatic successions in the opposite hemisphere, and their associated sediment yield cycles, will be 10,000 years out of phase as well.”
Until the Plio-Pleistocene, glaciation was unipolar and precession-scale eustasy tended to track the insolation cycle of the glaciated hemisphere, according to Perlmutter. As a result, similar climatic successions in opposite hemispheres would have had sediment yield cycles with markedly different phase relationships to glacioeustasy.
Large polar glaciers have long been recognized as the cause of cyclic, global sea level fluctuations. But the occurrence of glaciers through time, particularly short-lived ones, may well have been underestimated.
Direct evidence, such as moraines and striations, may have been eroded or reworked through time or even amalgamated by later, larger glacial events. Also, stratigraphy doesn’t always clearly yield information on small, rapid sea level changes.
When it comes to paleoclimate interpretation, preservation bias caused by the climate cycle itself plays a key role.
For example, some locales are exceedingly wet and then very dry. The question becomes one of whether it is more wet or more dry -- fig.3 and fig. 4.
“The Sahara desert may be a good example of preservation bias,” Perlmutter said. “Ten thousand years ago, the Northern Hemisphere summertime was warmer than present with a lot more rainfall.
“The active Niger River drainage area was bigger, and there were a lot of lakes in the areas we consider a desert today,” he noted. “Now, because it’s so dry, active dune fields have eroded most of the lake beds.
“In another few thousand years, there may be few or no lake beds preserved at all,” Perlmutter said. “So, if a geologist comes along in three million years and observes the remaining paleoclimate indicators, the interpretation is likely to be that the Sahara region was dry all the time when, in fact, the significant wet period just wasn’t preserved.
“Calibrating a climate model to the geologic record has to be done with the right perspective,” Perlmutter added. “If it’s done based on misinterpretation of the past because of what hasn’t been preserved in that location, then you’ll get a climate model that’s not useful for hindcasts or forecasts.”
In the oil industry, climate information can be useful to predict distribution of sand deposits for potential reservoirs.
“If we destabilize climate in a big way with alternating wet and arid cycles, the dune fields that are created at one point in the cycle get flushed out during the rainier part of the cycle,” Perlmutter said. “Basically, you get the sand coming out in a big pulse, and it’s more concentrated. It’s not distributed through the stratigraphic record but focused in one or two particular units.
“That’s what we’re trying to look for,” he added, “some of these major climate changes that switch back and forth and pump sand out.”
This must then be referenced back to the sea level phase, i.e., whether the sand was pumped out at high stand or low stand. Perlmutter noted there are different potential reservoir locations depending on the phase relationships of the sediment and sea level.