Wednesday, October 30, 2013

Hurricanes and Climate Change


Following the previous two posts that touched on trends in climate and natural disasters in general, we will now finally discuss the links between specific types of natural disasters and the changing climate. So let’s start it off with hurricanes! Just a quick note, hurricanes, cyclones and typhoons all refer to the same weather phenomena, just that they occur in different places. Hence the terms might be used interchangeably in this post. This post is pretty heavy in content, but I thought that it would be better if I addressed all these issues together in a single post for a more coherent discussion, rather than splitting them up.

Hurricane Sandy. Source: NASA Earth Observatory 
Theory: Link between Hurricanes and Climate Change
Hurricane activity generally occurs over the oceans where the sea surface temperatures (SSTs) exceed 26°C. The warm oceans provide the energy source to kickstart the formation of tropical storms and help warm the air within the hurricane and reduce the pressure and density of the air so as to make the winds spin faster.

Fig. 1 Formation of a hurricane. Source: Geogonline.org.uk

As such, in theory, the rise in SSTs due to anthropogenic climate change should provide more heat energy for tropical storm formation and lead to an increase in intensity and frequency of hurricanes. However, it would be an oversimplification to just look at the effects of SST on hurricane activity. SSTs are not the only ‘ingredients’ required in the hurricane ‘recipe’. Other factors that need to be considered include the strength of the vertical shear, availability of moisture in the air and the presence of a triggering disturbance such as the African easterly lower atmospheric winds (Trenberth 2005). Hence we need to take into account how a changing climate would have an impact on all the other factors as well before we can conclude on how hurricane activity change. Take for an example, although a warming climate would lead to higher SSTs, the vertical wind shear is also predicted to increase (Vecchi and Soden 2007) which would tear the incipient tropical cyclones apart, deterring them from intensifying. 

Moreover, we have to take into account the multidecadal variability in SSTs that exist in the ocean basin, especially the Atlantic Ocean. Known as the ‘Atlantic multidecadal mode’, there have been distinct warm and cool phases that have dominated the North Atlantic which are independent of the long-term global warming signal (Goldenberg et al. 2001). As seen in Fig.2, from 1944 to around 1970, the Atlantic multidecadal mode is predominantly warm and hurricane activity in the Atlantic was high. Thereafter, from the early 1970s to early 1990s, the Atlantic multidecadal mode switched to the cool phase and hurricane activity was below average. From 1995 onwards, the Atlantic decadal mode became warm and the hurricane activity increased again. The Atlantic multidecadal mode has been hypothesised to be linked to fluctuations in the intensity of the thermohaline circulation in the North Atlantic (Goldenberg et al. 2001).
 
Fig. 2 SST anomalies of North Atlantic shows multidecadal variability coined as the Atlantic multidecadal mode. The green dotted line shows the five-year running mean. Source: Goldernberg et al. 2001
Besides the mutlidecadal variability in SSTs, ENSO has also led to interannual variability in tropical cyclone activity as well. Atlantic hurricanes seemed to be suppressed when an El Nino is occurring in the Pacific due to an associated increase in vertical wind shear in the Atlantic (Goldenberg et al. 2001). Hence while Pacific activity might increase during an El Nino, there tends to be less activity in the Atlantic (Trenberth 2005).

Observed trends so far
Frequency
A study done by Webster et al. (2005) has shown that the there is no statistically significant change in the global frequency of tropical cyclones from 1970 to 2004 (Fig.3). Likewise, the same study shows that in each ocean basin, there is also no significant change in tropical cyclone numbers over the 35-year period, except for the North Atlantic Ocean (Fig.4). There seems to be an increasing trend in number of hurricanes in the North Atlantic that is significant at the 99% confidence level. 

Fig.3 Annual global frequency of hurricanes and storms between 1970 and 2004. Source: Webster et al. 2005

Fig.4 Annual frequency of hurricanes in each basin between 1970 and 2005. WPAC = Western North Pacific, NATL = North Atlantic, SH = Southern Hemisphere, NIO = North Indian, EPAC = Eastern North Pacific. Source: Webster et al. 2005 
Although some papers like Webster and Holland (2007) have suggested that greenhouse warming have led to a rise in SST that resulted in the increase in hurricane frequency in the North Atlantic, others have begged to differ. Landsea et al. (2010) pointed out that the increasing frequency of tropical cyclones in the Atlantic could actually be due to the large trend in the reported frequency of very short-lived (<2 days) Atlantic tropical cyclones as seen in Fig.5. The detection of these short-lived tropical cyclones have significantly improved due to the advancement of satellite technology and analysis techniques over the last few decades. Removal of the short-lived tropical cyclones from the full frequency record shows that the frequency of medium to long-lived storms has actually remained almost flat over the past century (Fig.6).
Fig.5 Frequency of short-lived (<2.0 days) North Atlantic tropical cyclones from 1878 to 2008. The black curve is the 5 year running mean and the blue line is the 1878-2008 trend. Source: Landsea et al. 2010.
Fig.6 Frequency of medium to long-lived North Atlantic tropical cyclones from 1878 to 2008. The black curve is the 5 year running mean and the blue line is the 1878-2008 trend. Source: Landsea et al. 2010. 
Intensity
On the one hand, some studies have shown that there has been an increase in mean tropical cyclone intensities over the years. Emanuel (2005) suggested the total Atlantic hurricane power dissipation has more than doubled in the past 30 years. Likewise, Elsner et al. (2008) has reported that over the period 1981 to 2006, the intensities of the strongest tropical cyclones has increased globally, especially in the Atlantic basin. On the other hand, other studies show that there has been a shift towards higher frequency of cyclones of higher intensity levels (Webster et al. 2005). Nonetheless, similar to the case of hurricane frequency, the increase in hurricane frequency might have been an artifact of the advancement of satellite technology over the years. Even if that factor is removed from the discussion, it is still hard to determine if anthropogenic climate change has played any role in the intensifying hurricane activity, given that the strongest signals come from the Atlantic Ocean where multidecadal variability (as mentioned above) is dominant (Knutson et al. 2010).

Projections
Despite all the different factors (other than just SST) that we have to consider while analysing hurricane activity, it has been predicted that global warming will have an impact on hurricane activity. Nonetheless, the predictions do vary across studies.

The downscaling of CMIP5 climate models by Emanuel (2013) has suggested that under the RCP8.5 emissions pathway, the power dissipation increases by 45% over the 21st century (Fig.7), with a 40% increase globally in hurricanes of Saffir-Simpson category 3 and higher. Likewise, Bender et al.’s (2010) model projections show that there will nearly be a doubling of frequency of category 4 and 5 storms in the Atlantic by the end of the 21st century.
Fig. 7 Power dissipation index of tropical cyclones averaged in 10-year blocks, using historical simulations for the period 1950 to 2005 and RCP8.5 scenario for the period 2006-2100. In each box, the red line marks the median among the six general circulation models (GCMs) used and the bottom and top of the boxes represent the 25th and 75th percentiles respectively. Source: Emanuel 2013.

Meanwhile, Emanuel (2013) suggests that under the RCP8.5 emissions pathway, the global frequency of tropical cyclones will increase in the range of 10-40% in the first three quarters of the 21st century, with most of the increase in frequency in the North Pacific, but substantial increases in the North Atlantic and South Indian Oceans as well. However, Knutson et al. (2010) concludes that it the global mean frequency will experience decreases ranging from -6 to -34% with greater decreases experienced in the Southern Hemisphere than Northern Hemisphere. Likewise, Bender et al. (2010) also suggest that there will be an overall decrease in frequency of tropical cyclones.

At the end of the day however, what matters the most might not be the intensity or frequency of tropical cyclones but how vulnerable we are to these tropical cyclones that make landfall. Should populations continue to increase along the coastal areas along with a lack of comprehensive disaster management schemes place then the damage incurred (economic losses and deaths) could also be very devastating even without a category 5 hurricane (e.g. Sandy was a Cat 3 hurricane but brought massive destruction and devastation nonetheless).

Phew..massive information overload. Let's just finish this post with two videos from the Weather Channel on the top 5 most powerful and deadliest hurricanes to ever hit the United States! 


Top 5 most powerful hurricanes (US)



Top 5 deadliest hurricanes (US)

Sunday, October 27, 2013

Storm Alert: St Jude heading towards the south of UK

Apparently we are going to experience strong winds and heavy rainfall tonight and tomorrow morning. Take care and suit up everybody!

Video from BBC 

Monday, October 21, 2013

Natural Disaster Trends


Since I have broadly touched on the topic of whether our climate is getting more extreme, I will now turn to look at the trends in natural disasters over the years. However, before we begin to look at the stats, let’s begin this post with the second episode of Disaster Bite on the devastating earthquake that struck Bohol, Philippines just last Tuesday 15th October 2013.

-----
Disaster Bite: Bohol Earthquake
Epicenter of Bohol earthquake. Source: Inquirer News
  •  7.1 (previously reported as 7.2) magnitude on the Richter Scale
  • Epicenter was at Bohol (about 620m south-southeast of Manila), at a depth of 33km
  • It struck on 08:12 (local time) on a Tuesday 15th October 2013, which also happened to coincide with a national holiday
  • The earthquake was caused by a vertical movement of the East Bohol Fault
  • Earthquake did not result in a tsunami
  • The death toll has surpassed 100 on Wednesday, with the greatest fatalities in Bohol and Cebu provinces
  • Many damaged buildings and stampedes were reported Bohol and Cebu
  • Tremor triggered power cuts in both provinces
  • Strongest tremor felt in the area in the last 23 years

More information:

-----
With reports saying that this was one of the deadliest quakes in Philippine history, which released energy equivalent to 32 Hiroshima bombs, it seems like it has joined the ranks of deadly natural disasters that have brought great tragedy to mankind in recent years (think Hurricane Sandy 2012, Tohoku earthquake 2011, Haiti earthquake 2010 etc.). A question that pops into my head right now is: are natural disasters becoming more frequent, intense and devastating?

Frequency

Fig.1 Frequency of natural disasters from 1900 to 2010. Source: EM-DAT

Based on the graph showing the number of natural disasters reported from 1900 to 2011 (Fig.1) produced by The International Emergency Disaster Database (EM-DAT), it does seem at first glance that the frequency of natural disasters have increased tremendously over the last century. However, this graph might actually be misleading as the increase in number of natural disasters over the years could in fact be due to better media reporting and advances in communications (Guha-Sapir et al. 2004). Moreover, with the launch of agencies like Office of US Foreign Disaster Assistance (OFDA) and Centre for Research on the Epidemiology of Disasters (CRED) in the 1960s and 70s, data collection on natural disasters have markedly improved. Hence, what the graph is showing might in fact be the evolution of the registration of natural disasters over time. Guha-Sapir et al. (2004) thus suggested that it might in fact be more appropriate to review the statistics of natural disasters over a shorter time span to spot trends (say between 1980 and 2000). Even so, the frequency of natural disasters during these 3 decades is still increasing. This might partly be due to the increases in hydro-meteorological disasters as seen in Fig.2. Some reports attribute such increases to climate change, which I will cover in the weeks to come. 

Fig.2 Worldwide polynomial trends for the four major types of natural disasters from 1900 to 2003. Source: Guha-Sapir et al. 2005

Intensity

Some studies have proposed that the intensity of some types of natural disasters has increased over the years. This include work done by hurricane expert Kerry Emanuel, who has shown that the total North Atlantic and western North Pacific hurricane power dissipation have more than doubled over the past 30 years (Fig.3) (Emanuel 2005). We’ll discuss hurricanes in greater depth in future posts. However, for other types of natural disasters such as earthquakes, it is much more difficult to spot any trends in intensity over the years. 

Fig.3 Annually accumulated power dissipation index (PDI) for the western North Pacific and North Atlantic, compared to annually averaged sea surface temperatures (SST). PDI has nearly doubled over the past 3 decades. Source: Emanuel 2005

Vulnerability and devastation

Nonetheless, it does seem that the impacts (e.g. loss of life, economic damages) of natural disasters are increasing over the years. It has been proposed that this is due to the increasing vulnerability due to large populations living in high-risk areas as a result of population growth and urbanisation (Huppert and Sparks 2006). Jackson (2006) cites the example of Tehran, which has grown from being a small village town to a megacity with 12 million inhabitants. This has greatly increased the vulnerability of Tehran to earthquakes given that it is built on an active fault system. In fact, in 2012, more than 300 were killed in twin earthquakes that rocked the region. Moreover, as our assets increase, there is more to be lost when a disaster strikes. The UN Office for Disaster Risk Reduction (UNISDR) reported that for the first time in history, the annual economic losses caused by natural disasters have exceeded $100billion for three consecutive years (2010-2012). This is due to the major increases in exposure of our industrial assets and private property to these disasters. 


Based on the discussion above, it seems that it is hard to conclude whether natural disasters have increased in frequency and intensity over the past few years. However, we are indeed becoming more vulnerable to natural disasters due to growth in population and assets. Nonetheless, with the recent IPCC AR5 report (2013) attributing some natural disasters to climate change as well as a recent NOAA report (Peterson et al. 2013) that attributed single natural disaster events to climate change, it is interesting to see how the changing climate as (discussed in the previous post) would have an impact on the intensity and frequency of future natural disasters. Stay tuned over the next few weeks as I continue on my quest to unravel the links between the changing climate and natural disasters.