The syllabus says
- Examine the methods used to make estimates (predictions) of the probability (in time and space) of hazard events occurring, and of their potential impact on lives and property.
- Discuss these methods by examining case studies relating to two different hazard types.
Why predict hazards?
Prediction allows people to reduce their vulnerability through taking preventative action. This includes evacuation, alterations to property, changes in daily routines, stockpiling of resources and, for governments, the long, medium and short term planning of effective strategies to reduce loss of life, injury and destruction to property.
prediction: a general introduction
Elements of prediction
The type of prediction that is done for hazard events has three elements - spatial, temporal and magnitude. All three are needed for an effective response.
Spatial ('where')
Temporal ('when')
Magnitude ('how big')
Prediction methods in general
Prediction methods fall into two main categories:
Primary event - this is a historical approach, looking at the past events, and attempting to deduce patterns. These patterns may be both spatial and temporal, such as looking at the frequency of earthquakes along tectonic plate boundaries
Secondary phenomena - this means looking at other features of the hazard, such as the foreshocks that often occur before a major earthquake. By establishing common features of the hazard event that occur before the event itself, it is hoped that prediction may be in time to reduce impacts.
The type of prediction that is done for hazard events has three elements - spatial, temporal and magnitude. All three are needed for an effective response.
Spatial ('where')
- the location of the hazard event
- the spatial extent of the impacts
- the intensity at which different areas will be affected (this has strong links to magnitude) but is also a result of the vulnerability of the population
Temporal ('when')
- the time of the onset of the hazard event
- the 'peak' time of the event (e.g. the moment of highest seismic activity in an earthquake, or the time of the fastest wind speed in a typhoon)
- the total duration of the event (from start to finish)
- the frequency of future hazard events
- the regularity of future hazard events
Magnitude ('how big')
- the size of the hazard event e.g. the level of shaking in an earthquake, or the level to which water is lacking during a drought
- the intensity of the event (will it be short and strong, or weak and long)
Prediction methods in general
Prediction methods fall into two main categories:
Primary event - this is a historical approach, looking at the past events, and attempting to deduce patterns. These patterns may be both spatial and temporal, such as looking at the frequency of earthquakes along tectonic plate boundaries
Secondary phenomena - this means looking at other features of the hazard, such as the foreshocks that often occur before a major earthquake. By establishing common features of the hazard event that occur before the event itself, it is hoped that prediction may be in time to reduce impacts.
predicting earthquakes
There is a long history of attempts to predict earthquakes. However, it is still not possible to accurately predict any of the three necessary aspects (temporal, spatial or magnitude). Earthquakes can (and do) occur anywhere on the planet, not just in seismically active areas along plate boundaries.
Past attempts to predict earthquakes
These have often been promising but have not accurately predicted future earthquakes:
Modern attempts to predict earthquakes
The video below outlines some of the methods used to predict earthquakes, and reasons why earthquakes are difficult to predict. In summary:
Past attempts to predict earthquakes
These have often been promising but have not accurately predicted future earthquakes:
- Radon gas counter - radon gas is produced during the decay in uranium, and can build up in faults under the surface. The gas is potentially released when the faults move, suggesting that high levels of radon at the surface might indicate movement at depth. However, this line of enquiry was largely abandoned after the 1990s because surface changes don't often indicate what the faults are doing under the ground, and also because radon doesn't precede anywhere near every earthquake. Indeed, it was described in an interview with the New York Times as potentially leading to an 'armchair prediction' because the gas is so easy to detect - creating even more danger as non-experts begin to 'predict' earthquakes.
- Changes in groundwater - historically this has looked at the water table (the point of saturated rock underground) and monitored sudden changes, but again, what is happening at the surface doesn't necessarily indicate what's going on underground. In addition, the lack of sudden movements before many earthquakes means it isn't reliable. However, recent research suggests that chemical composition may change before an earthquake, and this is a future line of enquiry.
- Changes in animal behaviour - famously used in Haicheng in 1975, the Chinese authorities evacuated the 1m population city days before a 7.3 magnitude earthquake, preventing around 150,000 deaths and injuries. However, there's no undisputed evidence that the recollection of strange animal behaviour is not part of a 'psychological focusing effect' whereby people look back to find evidence that would not have been remarkable at the time.
Modern attempts to predict earthquakes
The video below outlines some of the methods used to predict earthquakes, and reasons why earthquakes are difficult to predict. In summary:
- Mapping fault zones - the most simplistic idea is to look at the plate boundaries. But, most major earthquakes originate 15km below the surface, where pressure, temperature and water content vary signficantly. Surveys of the underlying geology are therefore an important area of research.
- Applied mathematics - earthquakes have been shown to follow a scale law, whereby for every 10 magnitude-5 earthquakes, there will be 1 magnitude-6 earthquake - and for every 10 magnitude-6 earthquakes, there will be 1 magnitude-7. This makes it possible to predict earthquakes. Unfortunately the 'foreshocks' are therefore only applicable 10% of the time.
- Monitoring of the Earth's crust from space - GPS devices have made laser, levelling and strainmeter measurements obsolete. Global Positioning Satellite devices make it possible to observe tiny variations in the crust over a very large area.
Further information can be found at these links:
- http://www.earthquakes.bgs.ac.uk/education/faqs/faq19.html
- http://earthquake.usgs.gov/learn/topics/topics.php?topicID=53
- http://www.thegeographeronline.net/hazards-and-disasters---risk-assessment-and-response.html#3
predicting typhoons
Tropical cyclones (also known as typhoons and hurricanes) are more straightforward, because they are developmental - meaning they develop gradually through a fairly sequential process. The problem with better predictions is that not all small storms become tropical cyclones, and the path of the cyclone (i.e. the direction and speed at which it moves) is unpredictable as it is significantly influenced by global scale atmospheric circulation, which varies hugely with both latitude and altitude.
The actual method of prediction is largely computer modelling based on remote sensing. Remote sensing includes satellite monitoring, buoys (floating weather stations) in the sea, and radar. The humidity, temperature, wind speed and wind direction is then fed into large computers which work out the probability of the future of the cyclone based on past event experience. Some countries - notably the US - also use weather planes which fly directly into the storm to gather further data.
More information can be found at the following links:
The video below gives further information about Hurricane Ivan - a precursor to Hurricane Katrina. It's a pretty basic video!
- http://www.wmo.int/pages/mediacentre/factsheet/tropicalcyclones.html
- http://www.tropicalstormrisk.com/
- http://www.hko.gov.hk/informtc/tracking.htm
The video below gives further information about Hurricane Ivan - a precursor to Hurricane Katrina. It's a pretty basic video!