Evaluating Likelihood and Addressing Uncertainty when Applying Solutions This is the third article in a five part series.
In the first two articles in this series we looked at identifying a problem and a functional understanding of that problem and then at determining all the causes of that problem. Now, we are going to consider how to take account of uncertain data regarding future events both in the probability that they may occur and in the degree to which they might occur. People are generally bad at statistics, which makes this part of the risk assessment process counterintuitive at times. However, risk is something we do accept on a daily basis, and so I will try and find some analogies. Again, this series is about taking politics and argumentative posturing out of the decision making process regarding climate change. To be sure we adhere to that premise, I will continue to make my examples totally neutral regarding the data and whether we are talking about a warming or cooling climate; I am discussing the decision making process.
First, there are some terms with specific meanings in risk assessment, that have more general meanings in normal speech, so we should define those. Likelihood is probability that an event might occur at some point in the future or over some time period. Severity (sometimes called criticality depending on the field) is a measure of how bad an event might be. This could be measured in dollars, injuries, lost production, public relations issues, loss of important secrets or anything you desire to prevent. Risk is a measure combining the severity of an event with its likelihood as shown in this earlier article. You can probably imagine why this measure is important as we would seek to avoid severe undesired events with a low probability of occurence and also less severe events with a higher probability of occurence.
So, we've established that our undesired event is a change in temperature of plus or minus more than 1.5 degrees Celsius. How do we know that we should do something about it or not? Well, many of our actions are built around controlling risk, even when we are not certain we will see those events happen. Consider insurance. If I have a family, I may purchase life insurance to protect them from the loss of my income in the event of my death. Even though it is very unlikely that my family will actually need that insurance, we will pass up other goods and services we could have purchased instead to fund a financial solution for my death. But, there is a limit to this, of course. If insurance cost 20 times as much, I may not find the trade off between the risk and the solution to be beneficial for my family. That same reasoning is used (with more mathematics) to evaluate the costs and benefits of potential solutions in a risk assessment program.
If we face uncertainty in our predictive ability or in our set of historical data about climate change, we can use risk assessment to evaluate what are legitimate solution costs to consider in light of that uncertainty. If I knew with 100% certainty that tomorrow I would face a loss of $10,000 to my property, it would be logical pay up to $10,000 today to prevent it, if I could be sure. If I knew there was a 20% chance of facing a $10,000 loss in the next year, I may want to spend up to $2,000 if I had an absolutely effective solution. If my solution was less effective, I would spend less. I have oversimplified the effect of probabilities, statistics, and costs over time, but the essense of the argument remains the same.
So, we will take our list of potential causes of climate change from the last article and provide assumed solutions and their effectiveness. Those are listed in the table below. In the next article, we will discuss more about evaluating our solutions and formulating policy based on risk assessment results.
