Steve Bowen is an Associate Director and Meteorologist for Aon Benfield’s Impact Forecasting division in Chicago.
The El Niño-Southern Oscillation (ENSO) is an oceanic pattern that tracks anomalous warming or cooling of the waters in the central Pacific Ocean. A warm phase is known as El Niño, and a cool phase is known as La Niña. These phases typically occur every three to seven years, and often trigger significant weather disruptions that are felt around the globe. When these anomalously warm or cool sea surface temperatures do not exist, ENSO is considered to be in a neutral phase. In the 66 years since official ENSO records began in 1950, there has been a nearly uniform distribution of ENSO phase years: La Niña (21), El Niño (22), and Neutral (23). The question then becomes: Do the different phases of ENSO lead to weather catastrophe loss variability?
Overall Global Weather Catastrophe Losses
When breaking down historical weather catastrophe loss data from Aon Benfield’s database, a distinct trend is recognized. La Niña years have clearly shown greater average annual losses on both an economic and insured basis in comparison to El Niño and Neutral phases. The results indicate that there is a 42 percent difference in annual economic losses alone between La Niña ($77 billion) and El Niño ($45 billion) years when using inflation-adjusted dollar values. Much of the increase in losses during a La Niña year surrounds the historical increased frequency of costly landfalling tropical cyclone events in the Atlantic Ocean basin and heightened flooding events across Asia Pacific. Similar percentage differences are found when analyzing from an insured loss perspective as well.
Global Weather Catastrophe Losses by Region
Significant distinctions between economic and insured losses by ENSO phases are found amongst the different regions of the globe. Asia Pacific shows a 121 percent increase during La Niña ($47 billion) on average as opposed to El Niño ($21 billion). The difference is less pronounced between Neutral ($27 billion) and El Niño. The United States also shows greater rates of economic loss during La Niña ($32 billion) and Neutral conditions ($33 billion) in comparison to an average El Niño year ($20 billion). Interestingly, the Americas have recorded more annual losses in La Niña years ($11 billion) than El Niño ($4.7 billion). This is surprising given warmer water temperatures off the west coast of South America and the southern half of North America typically leads to more tropical and non-tropical storm events that induce major flooding. EMEA, particularly Europe, does not typically sustain major weather variability during ENSO events. This is confirmed in the loss analysis. It is worth noting that data collection remains challenging in Latin America, which may be impacting this analysis.
Insured losses largely mirror the economic loss trend. The insured losses are primarily impacted by the location of the occurring event and the level of insurance penetration.
Global Weather Catastrophe Losses by Peril
When breaking down the economic loss data even further by perils, there are some expected yet slightly unexpected results. For this analysis, five major perils that are typically directly impacted by ENSO phases were selected: Tropical Cyclone, Flooding, Severe Thunderstorm, Winter Weather, and Drought.
The most striking find indicated that economic losses from the flood peril during La Niña ($28 billion) were 94 percent higher than during El Niño ($14 billion) and 51 percent higher than during ENSO-neutral conditions ($19 billion). The two costliest flood events in modern history – Thailand in 2011 ($47 billion) and China in 1998 ($44 billion) – each occurred during La Niña phases. Of the nine non-tropical cyclone induced flood events that have caused more than $20 billion in economic damage on an inflation-adjusted basis since 1950, five occurred during La Niña years. Three others were during El Niño and one was during Neutral conditions.
Similar ratios were attributed to the tropical cyclone peril, though not quite as dramatic. La Niña years have averaged $22 billion annually, which is 79 percent higher than El Niño years ($12 billion) and 15 percent higher than neutral years ($19 billion). The breakdown of the top 10 costliest hurricane events with inflation-adjusted economic losses beyond $15 billion by ENSO phase includes: Neutral (4), La Niña (3), and El Niño (3).
The rest of the major perils did not show distinctly greater differences. In each case, El Niño years were lower than La Niña and ENSO-neutral years. With the severe convective storm peril, thunderstorm activity produces slightly more tornadoes in the United States during La Niña or neutral years and the loss data on an economic and insured loss basis suggests this to be a contributing factor. Perhaps surprisingly it was determined that ENSO-neutral years are costlier for the winter weather and drought perils as opposed to warm or cool phases of ENSO.
Similar patterns were found when analyzing the insured loss data. An increased spike of insurance claims during La Niña was detected for the tropical cyclone and flooding perils around the world, though the majority of these claims were driven by “mega” events in the United States. If insurance penetration was greater in parts of Asia and Latin America, it is entirely likely that the percentage of insured losses for the flood peril would be much closer to – or perhaps even greater – than tropical cyclone.
Regardless of the current phase of ENSO, extremely costly and catastrophic weather events can occur anywhere around the world. However, as this analysis concludes, the financial implications tend to be even greater during La Niña years on an economic and insured loss basis.
This analysis, and more, can be found in Aon Benfield’s 2015 Annual Global Climate and Catastrophe Report. It is available for download here.