Tuesday 26 November 2013

Tropical Cyclones: Past, Present and Future

After looking into extra tropical cyclones in some detail, I would like to move on to another important climate hazard, tropical cyclones, in terms of the science, response to a changing climate and their impact on people. The Met Office describes Tropical cyclones as as amongst the most powerful and destructive meteorological systems on earth. I have produced two simple graphics using data from Munich Re (2013). The first shows the costliest natural disasters in term of monetary loss whilst the second depicts the same story for human fatalities. Costs consequent of tropical cyclone appear high up on these lists.  




The awesome might of these tropical systems can be seen in the video below which shows scenes from the aftermath of hurricane sandy, the second costliest natural disaster of all time behind the recent Japanese Tsunami. Tropical storms have the potential to wreck cities, livelihoods and cause long-term demographic, social and political change. 


The mechanics of tropical cyclones are relatively well understood. It is generally accepted that there are 6 critical antecedent conditions which must prevail in order for such systems to develop, intensify and in turn propagate. Of critical importance is the presence of warm ocean waters, at least 27 degrees centigrade. Related to warm waters is the depth of the thermocline, that is, the depth at which there is a sharp transition from warmer waters to deeper cold water. A deep thermocline means that if a tropical cyclone were to develop the associated turbulence only returns more warm water to the surface of the ocean and hence continues to fuel the genesis process In contrast, a shallow thermocline might mean that the strong winds of a hurricane bring cold waters up from below and self prohibit serious intensification. In terms of Atlantic systems, the position of the loop current may play an important role in allowing cyclones to intensify en route to Louisiana or Florida. (Will be considered in more detail in a later post.) The second key precursory condition is an atmosphere which is unstable. That is, an atmosphere which cools significantly wit height. Other requirements are moist layers in the mid-troposphere, a minimum distance of 500km from the equator (so that there is sufficient Coriolis Force to spin the cyclone), pre-existing near-surface disturbances, and low levels of vertical wind shear. The figure below illustrates the tracks of all historical Atlantic tropical storms. The tropical Atlantic source is evident as a favourable location for genesis, as is the northwards trajectory where storms pick up the Gulf Stream's warm ocean waters. Furthermore, It is clear that coastal regions bare the brunt of the tropical storm's force and that inland regions are relatively secure. 


The next image documents the anatomy of a hurricane. The eye which is roughly 20-50 km in diameter is found at the centre of the system. The eye wall immediately adjacent presents very strong winds, intense rain and thunderstorms. It is important to note that the direction of winds at the bottom of the system are in the opposite direction of rotation than at the top. Rising air at the centre of the storm is opposed by subsidence on the outer edges. Energy for intensification of the hurricane comes from sensible heat and latent heat as vapour changes state which creates high pressure, divergence aloft and lower pressure at the surface. The increased pressure gradient generates strong surface level winds. Just as there are conditions that favour the development of tropical storms, there are conditions that inhibit their formation too. Strong trade wind inversion, cooler sea surface temperatures and strong upper level winds play a role. 


Storms can leave distinctive sedimentary deposits in shallow marine deposits and coastal lagoons. Davis et al (1989) inferred that 'hurricanes produced graded or homogeneous facies of sand, shell gravel, and mud found in predominantly clastic sediments of late Holocene age in coastal lagoonal bays of Florida'. The hurricane signature is largely a consequence of their immense energy. As such, they can act as geomorphic agents wit the potential to cause coastal landform changes, particularly when hurricanes reach categories 4 or 5 when the winds get particularly strong. Liu and Fearn (1993) suggest that sediment cores from Lake Shelby on the Alabama coast show more tranquil climatic regimes prior to 3.2 ka BP as a result of a complete absence of sand layers in the cores. Furthermore, the authors propose that a more mesic climate post 3.2 ka BP was responsible for an unusual incursion of hurricanes into southwest Texas. Donnelly and Woodruff (2007) examined the millenial scale variability of Caribbean hurricane activity by reconstructing hurricane-induced overwash events from Laguna Playa Grande, Puerto Rico. The data reveals large scale fluctuation in the frequency of intense hurricanes. Relatively frequent occurrences of intense hurricanes are shown to have occured between 5.4 ka BP and 3.6 ka BP with the exception of a short quiet period from 4.9 ka BP to 5.1 ka BP. Following this period is a spell of quiet activity from 3.6 ka BP to 2.5 ka BP. Another active spell occurs between 2.5 ka BP and 1 ka BP. This is depicted in the figure below. A key future research goal is to relate these activities more reliably to past SSTs based on coral or sediment records. Such issues are at the centre of current debate on the impact of changing SSTs in the face of anthropogenic climate change will have on future hurricane activity. One this that is more certain, however, is that there is a link between La Nina years and intensified NAHU activity and supression during El Nino. This is chiefly due to increased vertical wind shear (noted above as a controlling factor) in strong El Nino seasons. 


That's it for now, later this week I hope to discuss a case study of the recent cyclone to have devastated the Philippines as well as prompting a discussion on the future for hurricanes in the face of anthropogenic climate change. The latest scientific hurricane research will be debated and participation is encouraged! 


Wednesday 13 November 2013

YOLO

YOLO. True. But as we move further into the Anthropocene, we are living much longer than our ancestors did a few thousand years ago. This definitely warrants a few minor diversions:

- 'How Long Will I Live?' a website where you can calculate how many years you have got left
- an insight into the use of the acronym on twitter

Let's get back to life expectancy. My post will talk about the general upward trend of life expectancy, what that means for population and explore reasons why we are living longer.

What's trending?
A trend in the life expectancy of humans during the past thousand years has been characterised by a slow and steady increase. Epidemics, famines and warfare were to blame for frequently upsetting this upward trend with volatile death rates however this curtailed in the mid-19th century. Why? Due to improved living conditions, advances in public health and medical interventions. 

The figures below suggest the changing picture of mortality. It is crystal clear that life expectancy is climbing and projections suggest that this trend will continue to increase. Only 50 years ago, life expectancy was just 68 years of age, now we can expect to live past 79. That's an increase of about 16%. A key indicator of the dramatic change in life expectancy is the growth of people aged over 100 in our society. In terms of centenarians per million, by 2030 estimates predict anything between 515 and 3,500. Now that's a LOT of letters from the queen. 



Figures 1 and 2 use data taken from RMS

We must be careful however when generalising life expectancy across the globe, as it can be different within the same city. Places just a few miles apart geographically have life expectancy spans varying by years. The diagram gives an example of a small area in London along the Jubilee Line. It suggests that men living in Westminster can expect to live at least 4 years longer than those in Canning Town!


Why the upward trend?
The greatest advance of medical science in history has helped to push life expectancy through the roof. Let's break this down and look at the major social changes year by year: 


Lifestyle
Medical
Health Economics
1950
Two thirds of adults smoke; smoking linked to cancer proved in 1954
Discovery of structure of DNA. Penicillin in mass production
Healthcare insurance offered by US employers. Universal health services set up in Europe
1990
Tobacco companies settle law suit for $206 bn; tobacco advertising banned. Food calorie labelling
Clinical trials prove statins cut heart deaths, 5% of adults take statins by 1998; heart attack mortality rate 80%
Healthcare expenditure averages $2000 per person. Consolidation of drug industry into giant pharmaceutical companies
2010
A quarter of adult population smokes; smoking bans in many countries
Lipitor becomes world’s best-selling drug; stem cell transplants; heart attack mortality rate 40%
Health expenditure 9% of GDP (average $3000 per person)
2030 – What If?
Nobody smokes; obesity wave halted; fitness levels higher than today
Cancer becomes managed disease
20% of GDP spent on medicine

This obsession with longevity feeds into our every day lives. Headlines such as 'sex adds years to your life' this century, have triggered a mad outburst in people searching for activities, foods, diets and wild and wondrous ways to live longer. wikiHow gives you an opportunity to live a very long life just by having a read of this article.  

Shocks to the system?
The last 20 years have seen unprecedented mortality improvement. Medical advances have been fundamental in this undeniable upward trend in life expectancy, however there have been numerous 'mortality shocks' which work against this trend. 
  • Asian Flu pandemic 1957/8 kills up to 4 million people worldwide 
  • HIV/AIDs emerges in 1980s, reaching 8 million cases by 1990
  • SARS emerges as a new infectious disease in 2002
  • Terrorist attack 2001 kills 3000
  • Avian flu outbreaks from 2006 puts world on pandemic alert

Pandemics are a threat to life expectancy now, but will they always be? Will technological fixes make us immune to pandemics in the future? What type of mortality shocks will gain momentum in years to come? What impact will climate hazards have on life expectancy? The recent Philippines typhoon alone is expected to have claimed approximately 2000 lives. 

The 5 'catastrophe perils' in the diagram below are shocks which have the potential to cause excess mortality in the future. Infectious diseases have featured throughout history, however natural hazards and mass-casualty terrorism are factors which have merely been considered as a threat to life expectancy in the past.


RMS include 'natural hazards' in their life-risk models, a factor which could start to gain momentum. Unprecedented urbanisation to coastal areas (explored in a previous post) means thousands of people are vulnerable to the increasing frequency of climate hazards. Do climate hazards have the capacity to reverse the upward trend of life expectancy? That is the question.

Sunday 3 November 2013

NAO in the News

Related to much of my previous discussion, I have found this great article in the Daily Mail.

'Why is it so cold? Simple... it's the North Atlantic Oscillation - and it's got a bit stuck'

Well worth a read! 

Question Time

My recent posts have raised some interesting questions. I thought I'd tackle them as a post in itself such was their relevance.

Ok here goes, question number 1:

"Cool post Isabela. It seems that the NAO is really important in determining the formation of extra-tropical cyclones. You mentioned that anthropogenic climate change might be influencing the NAO. Despite the lack of unequivocal evidence, has a causal mechanism by which climate change influences the NAO been identified?"

Thanks for your question. A causal mechanism for NAO variability is a complex and controversial topic. These have been a number of interpretations and many have used GCM simulations under different parameterisation. Robertson et al (2000) suggest SST distribution plays a role in modulating both interannual variability and regional modes. It is certainly clear that SSTs are changing in the midst of a changing climate, too.


Others have suggested anthropogenic climate change may influence the strength of the stratospheric vortex and in turn the NAO. In both cases, though, predictive models lack skill and robustness. Modelling improvements will be key in attributing a causal mechanism.

Question 2:

"Looking at the NAO index diagram you have there also appears to be a prolonged positive phase between 1900 and 1920, were there an increased storm frequency then as well?"

Here's a reminder of exactly what we mean by a positive phase...


Firstly, an important point to note is that the term North Atlantic Oscillation was only coined in the 1920s by Sir Gilbert Walker. Walker's 1924 writings eluded to the traditional description of the NAO through correlation between pressure anomalies between the sub tropics and Iceland. Consequently uncertainty is induced through an incomplete knowledge of the atmospheric system itself, and, secondly, through limited recording of weather events. Nevertheless the period from ~1900 to 1930 did represent a period where a positive phase persisted over multiple winters. I recommend this paper by Andrade et al (2008). The authors look at historic records of storm frequency from the Azores region. This extract summarises some important conclusions from the paper:

"A number of major periods of contrasting storminess were identified. The first period (AD 1836–1870) is characterized by a distribution with two maxima and low yearly storm frequency (2–3 storms/yr). A second period extends between AD 1870 and 1920 and corresponds to an irregular distribution of storm frequency that decreases in time, maximum values reported in 1879 and 1886; the first years of this period record an abrupt peak of storminess (> 8 storms/yr) that drops to 1 storm/yr by the end of the 1911–1920 decade. A third period corresponds to the decades of AD 1920–1940 with uniform distribution and low storm frequency. After 1940 (period 4) a general trend of increasing storminess emerges (2–3 to 4–5 storms/yr on average) with maxima occurring in the middle of the 1980s."