In 2007, U.S. Geological Service (USGS) scientists looked at the future of the polar bear throughout its range: Forecasting the Range-wide Status of Polar Bears at Selected Times in the 21st Century. The USGS report points to high probabilities of extinction for the bear in most of its current range within the next 100 years. "High" gets defined a little more later in this post. The bears are most likely to survive in Canada's northern islands.
Sea ice and ringed seals are essential to the bear's survival. A large part of the USGS results are driven by expected reductions in the availability of both. Here's a readable paper by two Canadian bear scientists that explains why the sea ice and ringed seals are so important: Melting under pressure.
I posted about the USGS report a few weeks ago: The polar bear's last stand. In this post I'd like to explore a little more what the report has to say about the efficacy of possible policy responses. I'm not going to talk about greenhouse gas policy. However important the polar bear is, greenhouse gas policy isn't being driven by it alone. I would like to look at the potential for local policy actions: normal game management applied to polar bears, oil and gas and shipping mitigation measures, tourism management, and so on.
The USGS study used a technique called Bayesian network modeling to incorporate expert opinion and probabilistic information. The authors caution that "...the model presented here should be viewed as a first-generation prototype..." The model output was a set of statements about the relative probabilities of various population size outcomes.
The study aggregated the world's 19 polar bear populations into four larger eco-system regions. The four regions are shown in this figure and described some more below:
The model projected percent probabilities for each of the following five outcomes in each region:
- a population larger than we see now
- a population similar in size to today's
- a smaller population than today's
- polar bears continue to exist in the region, but are rare
- polar bears are extinct within the region
In each region the probabilities for these five outcomes sum to 100. The pie charts below show the probabilities for a set of model runs made using the average results for a set of 10 plausible sea ice models and for other assumptions, including assumptions about changes in human activity in the Arctic (the charts are based on the report's Table 8).
The top two pie charts show the extreme results. The Seasonal eco-region includes Hudson Bay, the waters off of Eastern Canada, and most of the waters between Canada and Greenland. In these areas the ice disappears during the summer. The Archipelago eco-region shows the waters in and around Canada's northern islands.
The figures show the results after 100 years. In the Seasonal region there is an extra-ordinarily high likelihood that the bears will become extinct. In the Archipelago eco-region, there is a very high possibility of extinction, but there also appear to be much more significant chances that the bears will at least be rare, and that the worst impact may be a smaller population. Bear in mind that these are the estimated results after 100 years; if the IPCC scientists are right, global warming is going to continue beyond the 100 year time frame.
The next two regions provide coverage for the Arctic Ocean itself. In the divergent region, off of Alaska and Russia, the ice is formed and moves away; the ice piles up against Canada's Arctic coast in the convergent region. Here's a link to a dynamic map that shows why one region is called divergent and the other convergent: Old ice and the Northwest Passage. Extinction probabilities are high in both regions:
Are the polar bear's doomed, no matter what we do?
Still the Ghost pointed downward to the grave by which it stood.
'Men's courses will foreshadow certain ends, to which, if persevered in, they must lead,' said Scrooge. 'But if the courses be departed from, the ends will change. Say it is thus with what you show me!'
The Spirit never did give Scrooge a clear answer.
The results, summarized in the pie charts, were based on assumptions about likely future human actions and the impacts of these actions in the different areas at different times. For example, future shipping in the Archipelago region was set at "same as now" 45 and 75 years out, and then rises to "increased" 100 years out. I haven't reported the outputs for the pessimistic and optimistic melt runs, but the authors use different shipping patterns for these, depending on what is supposed to happen to the ice (Table 3 in the report outlines the assumptions).
The USGS team did look at how changes in policies with respect to Arctic development and use of Arctic resources might affect the polar bears. The bottom line was not encouraging and I emphasized the pessimistic elements in my earlier post. As I've reread the report I can see it suggests that there is a little more scope for using policy (other than greenhouse policies) to increase the probability of survival than I had thought.
More specifically, the team varied assumptions about the numbers of bears killed or removed from the population through direct human interaction, including hunting, takes for scientific purposes, takes for zoos, and, I assume, defensive kills of problem bears. The team also looked at the
...effects of a variety of contaminants, including: petroleum hydrocarbons, persistent organic pollutants, and metals. Although we don't know much quantitatively about effects of these contaminants at the population level, we know qualitatively that effects on immune systems and steroid levels etc. will ultimately have such effects. We also know that oil spills will have immediate and dire effects. It also includes effects of human activities and developments which may directly affect habitat quality, including: shipping and transportation activities, habitat change, noise, spills, ballast discharge, and ecotourism.
The bar chart below (based on the report's Tables 8 and 14) summarizes the probabilities of everything but extinction after 100 years for the central tendency of the ice simulations and for alternative assumptions about human activity.
The bar for "More" represents the results summarized in the pie charts above. As noted in the description of the shipping assumptions under the basic scenario, "more" does not necessarily mean more intense every activity at all time periods; the word is used to capture a general tendency. The "Same" bar assumes human activity continues at current levels. The "Fewer" bar assumes improved conditions compared to now.
I'd guess the "Same" and "Fewer" possibilities are unlikely. Here's a figure I used in the last post. I've focused on the Archipelago region and superimposed blue circles for two nearby oil and gas basins, and red lines to mark two key Northwest Passage routes. The northern oil and gas basin may already be economically viable, and there is intense interest in the Northwest Passage. Other types of development are in prospect as well.
I'm not sure too much emphasis should be put on the precise probabilities of the different outcomes, and the "same" and "fewer" development scenarios seem unlikely. However, the results suggest to me that management of human activities within a region can have an impact on the probability of population survival in a region.
One of the factors considered was the presence of persistent organic pollutants. It is likely that these are being produced outside the Arctic and carried in by air currents, ocean currents, or rivers. Efforts to address this issue would have to extend far outside the Arctic.
If policy actions can affect the likelihood of population survival, we have a tradeoff. Increasing levels of polar bear specific policy intervention can have a positive marginal benefit. I assume this will be a decreasing marginal benefit - as we reduce the likelihood of extinction the benefits from additional incremental reductions gradually decline. The benefit of reducing the probability of extinction from 80% to 70% will be greater than the benefit of reducing the probability of extinction from 20% to 10%.
There are two steps in the determination of this benefit - (1) the determination of the relationship between policy interventions and the impact on the probability of survival, and (2) the valuation of that change in probability. The USGS work addresses the first element.
The best level of policy intervention will depend on the marginal benefit of increasing levels of protection, and on the marginal cost of increasing those protection levels. Increasingly restrictive measures will come at increasing costs. It may be relatively inexpensive for oil companies to create bear management guidelines and conduct training for their employees, and that may have a good sized payoff. It may be much more expensive to prohibit oil and gas exploration and development in large swaths of the Archipelago region.
At some point, the additional benefits of an increment in protection will become equal to the additional costs of that increment.
The optimal response is not necessarily to drive the probability of extinction to zero (although the optimal probability is not necessarily about 50% as the diagram might suggest, the curves are only drawn to make a general point). It would be nice to have a better feel for the marginal cost curve; it would be nice to find ways to reduce the marginal cost of achieving every level of survival probability. That would shift the marginal cost curve down and a higher probability of survival would become optimal.
Polar bears have increased in population from about 8,000 in the 70’s to 30,000 to 40,000 at present. Your suggestion is alarmist and unfounded. And regarding warming, the last 10 years have been cooling, arctic ice is now thicker by about 20cm than it has been in decades. World temps have been so cold the last 12 months that the 1 degree c increase of the last 100 years has been erased. So relax it will be OK.
Posted by: buy r4 dsi | January 25, 2010 at 08:18 PM