Hi everybody, I recently finished writing a short essay about extreme weather events due to global warming, so I thought I'd post it here. It's not too long or science-heavy and it took me a little over a day to write. Let me know what you think!
Much of current climate science pertaining to anthropogenic global warming focuses on averaged changes in temperature, precipitation, atmospheric circulation, and other factors. The 2007 IPCC conference, though comprehensive, was dominated by the discussion of the broad, long-term repercussions of global warming such as the decrease of summer arctic sea ice extent and increases in sea level due to ice sheets melting worldwide. Among the most impactful effects of global warming on society, however, are not these long-term averages, but the extreme events that come with them. Previous research by many climate scientists has shown that global warming will, on average, not only cause an increase in the number of record-breaking events but in the severity of them as well. I will focus on the changes that will be expected in the number and extremity of three specific types of extreme weather: heat waves, hurricanes, and tornadoes.
A plethora of previous research has been done on these extreme events; Tyler Ruff and David Neelin have done extensive research using Gaussian curves to prove that even a small increase in mean temperature can have drastic consequences for the frequency and strength of major heat waves. Kevin Trenberth’s research shows a mixed picture for hurricanes; hurricanes are predicted to increase in intensity due to warmer SST (sea surface temperatures) and a moistening of the lower troposphere, but their individual tracks are uncertain, meaning that it is hard to know at this point if highly populated places like the U.S. coastline will see an increased number of hurricanes or if hurricanes will trend away from the U.S. coast. Although 2011 was a record-breaking year for tornadoes and 2012 is off to a fast start, Diffenbaugh, Trapp, and Brooks attest that it is difficult to forecast any future trends at this time due to seasonal variability, subjective tornado data, and unclear model predictions. However, some conclusions about what the future may hold for tornadoes can still be made, and these authors stress that much more research needs to be done on the subject. Indeed, many of the mechanisms that cause extreme weather events are poorly understood, but we still have a wealth of information to draw reasonable and scientifically sound predictions on future extreme meteorological events. It is essential that we understand the predicted effects the future greenhouse climate so that we, as a society, can prevent catastrophic losses of life and property due to extreme events in the future.
Nearly all of the scientific community believes that anthropogenic global warming is real, but there are differences in opinion regarding the exact effects of anthropogenic global warming on the frequency and severity of extreme events, particularly for specific regions. Climate models are the main tools used to predict future extreme events, and although climate science is far from intuitive, many of the results shown by the models are similar to what would be expected based on our current knowledge of how weather works. For example, climate models pick up on the fact that anthropogenic greenhouse forcing causes enhanced tropospheric temperature, and they go even further to forecast that increases in temperature lead to increases in the amount of water vapor in the air and enhanced precipitation rates as a result. Additionally, confidence in models is increased through their agreement with observed changes in Earth’s climate. Many models have forecast an increase in the intensity of rainfall events and an increase in the severity of heat waves, and these forecasts have been validated by recent observations in many parts of the world (Easterling, 2000). Unfortunately, climate models are continually hampered by the relative lack of data in certain regions of the world for optimal model initialization and the lack of computing power to increase model resolution. These two problems make it difficult for scientists to predict the tendencies of extreme events in the future greenhouse climate for specific regions.
Short-term climate records of extreme events are also very important for scientists aiming to forecast the tendency of extreme events. These records only extend 150-200 years back in many parts of the U.S., but they provide the detailed information needed for the historical analysis and future forecasting of short-duration extreme events. Many of these records show trends in themselves; northern California and southern Oregon have shown a decrease in extreme flooding and wind events from extratropical cyclones over the past 50 years, while northern Oregon and Washington have seen an increase in these events. Coupling these observed climate trends with model output helps scientists gain insight into the particular climate trends in certain regions, especially if the climate models and observations show strong similarities in terms of extreme events.
Ruff and Neelin propose that although many localities have Gaussian distributions of temperature extremes, there are many places that have non-Gaussian “tails” of temperature distribution, and that these tails play a huge role in determining the future for extreme temperature events in a particular locale. The figure to the left shows the distribution of the probabilities of three temperature variables: Tmax (the daily maximum) Tavg (the daily average), and Tmin (the daily minimum) over a Gaussian curve. Graph (a) shows the distribution over June, July, and August, and graph (b) shows the distribution over December, January, and February. This input data was derived from the synoptic/hourly observations contained in U.S. Air Force DATSAV3 Surface data and Federal Climate Complex Integrated Surface Data (Lott et al., 2008).
The asymmetry of these curves shows us that a uniform mean increase in temperature throughout several different regions does not necessarily correspond to a uniform distribution of record-breaking temperature events. An increase of 2 degrees Celsius could have profound effects on the strength and occurrence of heat waves in Phoenix while having little impact on the heat waves in Los Angeles. If the mean temperature is increased and the Gaussian curve is shifted to the right, the occurrence of future extreme temperature events increases at different rates for different regions. Locations with high-side tails that are roughly Gaussian, such as Phoenix and Houston in the summer, will see a large increase in the number of extreme heat events if the curve is shifted to the right because the slope of the curve at the ends is very steep and therefore very sensitive to a slight change in the mean temperature. Long Beach and Los Angeles may see the same mean warming effect, but because the tails at these locations for maximum temperature are long and have a more gradual slope, the increase in instances where a given high threshold value is exceeded will be proportionately lower than those experienced in Phoenix and Houston. Additionally, because places like Los Angeles and Long Beach already have strong heat waves in the summer that depart from the average temperature, they likely have the infrastructure in place to deal with an increased number of these events in the future. Places with curves like Phoenix and Houston do not experience such a range of extreme heat waves during the summer, and therefore likely lack the infrastructure to deal with a sharp increase in the severity and number of particularly hot periods in the summer. The curves for winter temperatures suggest that there will be a relatively gradual decline in the occurrences of extreme cold events but a sharp increase in wintertime warm events. In order to predict the risks from increasing average temperature in a given location, applying known Gaussian curves and the exponential tails that often come with them is necessary.
It is safe to say that there will be an increase in extreme temperature events under global warming, but it is harder to draw conclusions to the number and strength of hurricanes forecast in a future greenhouse climate. It would seem intuitive that the warmer sea-surface-temperatures (SST) forecast due to global warming would fuel a greater number of hurricanes, but this isn’t necessarily the case. Greg Holland and Peter Webster found in their “Heightened Tropical Cyclone Activity in the North Atlantic: Natural Variability or Climate Trend?” paper that there was an increase in the number of observed storms in the North Atlantic basin over the past century, but this was quickly refuted by Chris Landsea’s “Counting Atlantic Tropical Cyclones Back to 1900,” which states that the observed increase in hurricanes is due to improved monitoring via satellites and other instruments over the past century. Even if the Atlantic has seen an anomalous period of enhanced hurricane activity recently, it is hard to know if this activity is due to global warming or is simply the result of decadal oscillations and ENSO (El Nino Southern Oscillation).
Kevin Trenberth makes an important distinction between the differences in the number of tropical cyclones and the number of intense hurricanes. Tropical cyclones form when a group of tropical thunderstorms develops an axis of rotation, and the main obstacle they face to forming is wind shear, which can tear these thunderstorms apart before they can organize and become a stronger system. Once a tropical system has formed, though, it derives its energy from warm, moist air over a warm ocean that is at least 80 degrees Fahrenheit for up to 150 feet in depth. The effects of global warming on wind shear are not yet known, but climate models increase SST over nearly every world region, and these predictions have been verified in observed trends throughout the world. It is unclear if hurricanes will increase in number, but they are forecast to increase in intensity.
These forecast increases in intensity have been verified by Kerry Emanuel in his 2005 paper “Increasing Destructiveness of Tropical Cyclones over the Past 30 Years.” Emanuel plotted the PDI, or “Power Dissipation Index,” which measures the total energy that a hurricane releases over a year, and the increases in SST over the past 90 years and found a closely correlated increase in both PDI and SST over the past 30 years, even though he admits that the number of hurricanes in most regions around the world has stayed relatively constant.
The figure to the right shows the PDI and averaged SST in the North Atlantic. The increase in the power dissipation index is either because the storms have become more intense or lasted longer (or both). The duration of storms in both the North Atlantic and the western North Pacific has increased by around 60 percent since 1949, and the average peak wind speed for storms has also increased by 50 percent. Stronger storms usually live longer, because it takes them more time to obtain a certain high wind speed and takes them longer to dissipate from this peak. This figure additionally shows that SST and PDI are strongly correlated, and that SST is much more important for the strengthening of hurricanes than wind shear in the atmosphere. In conclusion, it is hard to know if there will be an increase in the number of hurricanes, but if models are correct and trends continue, we will likely see an increase in the number of major, destructive hurricanes in the future.
2011 was the most was one of the most prolific years on record for tornados and was capped off by one of the greatest natural disasters in the history of the U.S. – the April 25-28th 2011 Super Outbreak. While it is even more difficult to forecast how tornadoes will change in severity in a warmer climate than hurricanes, it is very important to attempt to do so, as major tornado outbreaks have major consequences in the U.S. and abroad.The diagram on the next page shows an increasing trend in the number of tornado reports and a flat or even slightly downward trend in the number of F2 to F5 tornadoes. These observations should be taken with a grain of salt, however, because they are likely not indicative of major changes in tornado activity. As the U.S. observational network has expanded, more and more weak tornadoes are being reported now than ever before. Most significant (F2-F5) tornadoes were observed and recorded in earlier decades, but the decrease in them is most likely due to subtle changes in strength ratings as time goes forward (Diffenbaugh, Trapp, and Brooks, 2008). Because tornadoes are rated via a human-based assessment of damage, it is hard to set a common definition for tornado strength, especially when the damage that a given strength of tornado can cause varies widely based on topography, terrain, and infrastructure of an affected region.
Even so, this data is useful in that tornado reports for a certain year can be applied to the climatological record for that year, and, through statistical models, be used to estimate what the tornado activity was like in a given year, but it will take a longer period of modern, detailed tornado observations to be able to outline clear trends in tornado activity.
Tornadoes spin out of supercell thunderstorms, which get their energy from CAPE (convective available potential energy) and get their rotation and longevity from wind shear. Global warming is expected to increase CAPE by increasing temperature and humidity within the atmospheric boundary layer while simultaneously decreasing vertical wind shear by weakening the meridional temperature gradient (Diffenbaugh, Trapp, and Brooks, 2008). These two changes may offset each other with regard to tornado frequency and intensity in many locations, but they are likely to significantly change tornado seasons and occurrences. The meridional temperature gradient is greatest during cold seasons, and increased CAPE year-round may lead to an increased number of tornadoes in the winter. Many models also shift the peak meridional temperature gradient and the associated mid-latitude jet stream northward under global warming, meaning that there may be a northward trend in peak tornado activity in the future. The limitations in current knowledge and model resolution at this time pose challenges to accurately predicting tornadoes, much more so than heat waves or hurricanes. To accurately predict trends in future tornado activity requires that models have enough resolution to account for local, small-scale features and the physics of thunderstorm growth within these regions, and this resolution will require massive amounts of computing resources. Forecasting changes in the prevalence and intensity of tornadoes for different regions is in its infant stage at this point, but as computing power increases and climate model resolution becomes finer, we should be able to see clearer predictions of what may happen.
High-impact extreme weather events from global warming are projected to increase in severity in many regions, but it is misleading to simply say that global warming will cause more extreme weather. Global warming is a worldwide problem, but the repercussions of it, particularly extreme weather events, are very region specific. Mesoscale climate modeling is necessary both to accurately predict extreme events for certain regions and the regional implications of these events. Los Angeles should experience a slight increase in record-breaking heat waves, but Phoenix, 350 miles away, could experience many more in the future. Even though predicting hurricanes in the future climate requires more synoptic scale modeling, the effects of extreme hurricanes for certain regions, such as New Orleans, need to be modeled with high resolution. The University of Washington has taken some steps with its WRF (Weather Research and Forecasting) climate models, but for the rest of the country, an increase in resolution is necessary to determine the regional implications of extreme weather due to global warming.
Diffenbaugh, N. S., Pal, J., Trapp, R., & Giorgi, F. (2005). Fine-scale processes regulate the response of extreme events to climate change. PNAS, 102(44), 15774-15778. Retrieved from http://www.pnas.org/content/102/44/15774.full.pdf
Diffenbaugh, N. S., Trapp, R. J., & Brooks, H. (2008). Does global warming influence tornado activity?.American Geophysical Union, 89(53), 553-560. Retrieved from http://www.stanford.edu/~omramom/Diffenbaugh_Eos_08.pdf
Easterling, D. (2000). Climate extremes: Observations, modeling, and impacts. Science, 289(5487), 2068-2074. Retrieved from http://www.sciencemag.org/content/289/5487/2068.full.pdf
Holland, G. J., & Webster, P. J. (2007). Heightened tropical cyclone activity in the north atlantic: Natural variability or climate trend?. Philosophical Transactions of the Royal Society A, 365(1860), 2695-2716. Retrieved from http://rsta.royalsocietypublishing.org/content/365/1860/2695.full.pdf html
Landsea, C. W. (2007). Counting atlantic tropical cyclones back to 1900. American Geophysical Union,88(18), 197-208. Retrieved from http://www.aoml.noaa.gov/hrd/Landsea/landsea-eos-may012007.pdf
Lubchenco, J., & Karl, T. R. (2012). Predicting and managing extreme weather events. Physics Today,65, 31-37. Retrieved from http://scitation.aip.org/getpdf/servlet/GetPDFServlet?filetype=pdf&id=PHTOAD000065000003000031000001&idtype=cvips&doi=10.1063/PT.3.1475&prog=normal
Ruff, T. W. and J. D. Neelin (2012), Long tails in regional surface temperature probability distributions with implications for extremes under global warming, Geophys. Res. Lett., 39, L04704, doi:10.1029/2011GL050610.
Trenberth, K. (2005). Uncertainty in hurricanes and global warming. Science, 308(1753), 1753-1754. Retrieved from http://www.sciencemag.org/content/308/5729/1753.full.pdf
Thanks for reading!
I've already turned my paper in,but if you have any suggestions for my writing, let me know!