Confronting weather extremes

The basic concept of risk is rather straightforward.

Likelihood: First, we need to compute the probability of a weather extreme. Thus, President Donald Trump tweeted "Wow – Now experts are calling #Harvey a once in 500-year flood!" A 500-year event implies the chance of such events occurring in any given year is 0.2%. A 500-year event is not inevitable in the 500th year simply because it did not happen in the preceding 499 years, and once a 500-year event occurs we are not automatically immune for the next 499 years. Each year, the probability remains 0.2 per cent. However, fundamental changes, such as urbanisation or climate change, could turn 500-year events of the past into more frequent (such as 100-year) events in the future. This is what scientists often claim climate change could do to weather extremes globally, and deforestation or urbanisation, regionally or locally. Besides frequencies, climate change may impact intensities or durations of weather extremes. Thus, warmer atmospheres on the average are likely to cause more evaporation and possibly more intense storms and hence lead to heavier precipitation, while warmer oceans may intensify tropical storms, and urbanisation is expected to increase flash floods since paved unvegetated surfaces absorb less water. Human signatures can, therefore, alter the probability of weather extremes, although such alterations, if any, may only be discernible at decadal to even century scales.

Impact: Second, we need to assess the vulnerability of the impacted systems. This can also be thought of as a probability, except this is the probability of damage when the weather extreme occurs. Thus, when hurricanes Katrina hit New Orleans in 2005 or Harvey hit Houston in 2017, the leading causes of increased vulnerability were, respectively, a levee that was near breaking point and houses that were below sea level in New Orleans, and urban lifelines or hydraulic infrastructures that were not prepared for a Harvey in Houston.

Third, we need to consider exposure, which is the total potential loss, whether in human lives or property, or to private sector enterprises, or to government and public sector assets. Vulnerability, as defined here, may occasionally be merged into the definition of hazards likelihoods or into exposure. In any case, relative to the mitigation of the hazards themselves, reduction of vulnerability and exposure are more immediately and directly under human control through appropriate investments into infrastructures and urban planning or land use. Aging of critical infrastructures and their growing cross-vulnerabilities, the breakdown of social community networks, and growth or movement of people into cities and coastlines, are all causing a significant increase in vulnerability.

In addition to changes in the three elements of risks, mathematical complications may arise in the computations because these elements may not be independent of each other. The expected losses from weather extremes will also depend on investments in readiness of first responders in early warning or dissemination systems, such as the use of sensor systems or social media, and in plans made in advance for the effective and efficient recovery of critical systems. Finally, cost-benefit tradeoffs, such as regulatory and policy considerations for balancing jobs in energy sectors and maintaining growth and business competitiveness, versus limiting weather related economic disruptions or loss of lives and property damage, as well as considerations such as immediate social justice or longer term sustainability, will need to be considered. The tradeoff space for decisions may be complex and large, hence real-world risk computations may quickly get complicated. These computations are perhaps best left to the experts. Suffice it to say that there is significant room for improvement both in developing actionable metrics for risk-informed policy and communicating them meaningfully to non-scientists. Our investment priorities and planning decisions can still rely on sound scientific methods and engineering principles, rather than turning into ideological battlegrounds which begin to resemble religious belief systems - which brings me, albeit briefly, to religion.

Let me hasten to add that I have nothing against religion per se, even though religion, in general, has been getting bad press lately. I venture to say that you probably would not immediately associate belief in God with insurance companies or law firms; yet, it is these companies and firms that invoke "Act of God", or "Force Majeure" clauses for weather extremes to minimise or eliminate liability. Despite my best efforts to stay agnostic, here I feel compelled to rise in defence of our Lord God, for I believe She (or He, if that better suits your belief system) is being rather unfairly blamed. First, as we discussed previously, weather extremes in the longer term may depend probabilistically on human-induced changes such as urbanisation, irrigation, or deforestation, and on the burning of fossil fuels leading to warming of our planet. Second, hazards do not turn into disasters, let alone catastrophic events, without the inaction of man. It is he (man) and not She (God) who has made Holland resilient to floods, yet manages to keep America vulnerable.

Developing resilience to weather extremes may look very different in developing versus developed nations, and not just because of the disparities in resources that can be brought to bear for preparedness, emergency management and recovery. This summer a few researchers from across the globe gathered in Jakarta to discuss flood resilience of coastal cities. We discussed, among other things, how lessons learned in Holland, arguably the most resilient nation to weather extremes, may not directly generalise upon Indonesia. One example was the planned project Garuda: a USD 40 Billion investment in coastal protective infrastructures to save Jakarta from floods. Conceived with help from the Dutch, even this project was found wanting. The problem was not with the infrastructure as such, but with how investments needed to be allocated to optimise flood resilience. The problem in Jakarta is not limited to inadequate infrastructure but derives from a combination of climate change, deforestation, watershed management, urban form and stratification, waste management, social justice, politics, water quality, and public health.

Public health makes me think of an analogy with human health. The general health of a person may depend on lifestyle and dietary choices as well as an exercise regimen. A healthy person's body may be able to fight certain diseases better, but ultimately everyone is susceptible to disease. Thus, vaccines need to be administered prior to an outbreak. However, once disease strikes, the primary focus needs to be on treatment and recovery, rather than in attempting to relate exactly how much of the disease may be attributed to, say, longer term lifestyle or dietary choices. However, healthy lifestyle choices must be emphasised before disease strikes or following the immediate recovery phase. Finally, advances in medical science will not translate to better healthcare unless medical insurance, regulations and institutions are well developed. This analogy translates to weather extremes, specifically, for disaster management and post-disaster recovery, investments in reducing vulnerability and exposure, as well as longer-term adaptation or mitigation plans. This brings us to herding cats, swans, and butterflies.

"Herding cats" is an idiom: it is a figurative way to describe attempts to organise uncontrollable or chaotic entities. Our ability to confront weather extremes and their potentially devastating impacts depend on three critical factors that are all somewhat difficult to control: managing financial incentives and insurance strategies, preparing for unprecedented events that are no longer surprising, and taming the inherent complexity in weather and climate systems and the impacted coupled natural-built-human systems.

Cats: The insurance sector determines premiums and portfolios based on Natural Catastrophe ("Nat Cat", or simply, Cat) models (which are basically risk models for hazards developed for insurance clients), and has started to develop and explore new financial instruments such as catastrophe ("Cat") bonds. Cat models were adopted by the insurance sector after vigorous campaigning by industry pioneers such as Karen Clark but only after the devastating impacts of the Category 5 Hurricane Andrew which made landfall in Florida in 1992.

Swans: Black swans have been defined, by Nassim Nicholas Taleb in his book with the same name, as extremely rare and unpredictable events with disproportionate impacts. Grey swans were defined as events that are unprecedented but no longer surprising (e.g., owing to fundamental changes in extremes, vulnerability, and exposure). A recent climate study prophesied grey swan tropical cyclones across the world, including along the US coastlines. As if on cue, Harvey struck Houston.

Butterflies: The butterfly effect issued to describe complexity specifically in the context of chaos theory, which describes extreme sensitivity to initial conditions in what are called nonlinear dynamical systems. Weather and climate, as well as coupled natural-built-human systems impacted by weather extremes, are examples of such complex nonlinear dynamical systems, and may even exhibit chaos. Projections from such systems obey basic rules of physics in terms of aggregate level trends and statistical attributes but tend to generate a wide array of plausible futures at scales and for variables of interest in the context of weather extremes and their impacts. This complexity needs to be tamed to enable risk-informed policy and resilient design.

What we need to make our cities and nations resilient to weather extremes are not precise predictions which may not be feasible owing to the presence of what are called irreducible uncertainties, but a characterisation of the complexity, especially for extreme events, and appropriate financial incentives. When civil engineers build a bridge with a design life, they do not require precise traffic projections over the entire period, but the statistics of extreme values for traffic and other loadings (e.g., wind or earthquakes), which in turn need to be balanced with economic considerations. Fundamental science breakthroughs and practical engineering innovations must be blended with financial incentive structures to solve the challenge of weather extremes. The need is clear and present, and solutions are within reach if only we care enough to build them.