Skip to main content
Hong Kong Observatory Brand Hong Kong - Asia's world city
GovHK Traditional Chinese Simplified Chinese Search
Search Site Map Contact Us
red dot
Print Version
Print Version

Monday, 22nd March 2010

Risk management

'Risk management' is not just a popular term these days. It is actually practiced in many areas --- in investment, insurance and finance, just to name a few.

In the provision of weather service, probability forecast is being issued in some parts of the world, notably North America. For instance, the probability of rain is routinely given. Such a forecast is useful to decision makers or those who operate a decision model in their operations, e.g. ports and terminals. It can also be used in deciding on an outdoor activity, e.g. a barbecue or sports event. The risk of an event being ruined by rain can be managed through a cost-benefit analysis. For instance, people can make up their mind after knowing the chance of rain and evaluating the relative cost between a dry and a wet programme for the event.

Not many people are aware that some kind of risk management is now being increasingly adopted in typhoon forecasting. Since the early 2000s meteorologists found that with advances in the numerical prediction of tropical cyclone, it is possible to minimize the forecast error in a typhoon track by a 'multi-model consensus' approach. The approach takes account of the predicted storm positions generated by global NWP (numerical weather prediction) models run by major weather centres in Europe, Japan and the United States. As a matter of fact, research work by Observatory colleagues indicates that even a simple average of these predicted storm positions would produce a significant improvement in typhoon tracking, in general better than the tracking by any individual model.

Here is an example of tracking of Typhoon Koppu when the storm threatened Hong Kong in September 2009, necessitating the issuing of the No. 8 Gale or Storm Force Wind Signal:

Figure 1     Forecast tracks of the various major weather centres of Typhoon Koppu for 1200 UTC (8 p.m. Hong Kong Time) on 14 September 2009
Figure 1      Forecast tracks of the various major weather centres of  Typhoon Koppu
                  for 1200 UTC (8 p.m. Hong Kong Time) on 14 September 2009 (the label
                  1412 represents fhe forecast position for 1200 UTC on 15 September, etc.)

For the approach to be useful, the performance of the global models individually has to be reasonably good in the first place. There is little point in producing a 'consensus' if, as an idealized example, one model takes the storm to the east while another takes it to the west.

Several years of experience with the multi-model approach suggests that it is more than risk management --- it brings about significant improvement in typhoon tracking as a result of scientific advances in computer modeling of the weather.

There is no lack of risk management in history. I can name a couple of examples in the history of religion. In the incipient days of Buddhism in China about 2000 years ago, it was not uncommon to see Buddhist deities in Taoist temples, Tao being an indigenous religion and philosophy in China. In those days, one could also see figures of Confucius featuring in Buddhist and Taoist temples. To get a firm foothold in China, Buddhist monks were also asked to study Confucian and Taoist teachings.

One reason for this was that Chinese people are not strongly religious. In their mind, there was perhaps no harm in including more deities in the temple lest they upset those being left out. That Buddhist deities could be found in Taoist temples reflected the monks' effort at the time to get a share in the religious 'market'. What better strategy it was than to 'infiltrate' temples that believers normally went.

In the western world, King Redwald (died 624 or 625 AD), of East Anglia in England, built two altars, one for indigenous gods and one for Christianity. It was said that he did this as a 'double insurance', a way to safeguard against oracles not going well. England fully adopted Christianity by around 670 AD.

The concept of risk management equally applies to the issue of climate change. Recently, there has been increasing media coverage on scepticism about the work of the IPCC (Intergovernmental Panel on Climate Change), the Nobel prizewinner for peace, and on the science of climate change in general. While there has been some carelessness in a couple of places in the IPCC report, the science behind climate change is sound and the rigour in the studies and arguments leading to the key results are strong.

The IPCC report points out, all along, that it is 90% certain that global warming is brought about by human activities. Such certainty, or uncertainty if you like, is quite normal in the field of science. For the decision maker, he or she would be wise to bear this in mind when planning their way forward. It would not be wise, and certainly not in their best interest, not to take account of the key results in their decisions simply because of some minor mistakes in the report.

B.Y. Lee

Last revision date: <17 Jan 2013>