Carbon dioxide (CO2) is the product obtained when humans burn fossil fuels for energy. CO2 is a greenhouse gas, meaning that its presence in the atmosphere traps heat produced by incident sunlight, preventing its release back into space. Although CO2has been present in the atmosphere for millions of years, the amount added since the industrial revolution began has been steadily increasing, and continues to do so even at the present time. This is because the worldwide use offossil fuels has been growing at an ever-increasing rate since the industrial revolution began.
Climate scientists over the past several decades have developed climate models to reproduce past global temperature trends and predict future trends. Models correctly reflect the temperature increases of recent decades when the additional CO2 from burning fossil fuels is included, but they fail to show the increases when this CO2is omitted. With this success at reproducing past behavior, scientists use climate models to project future developments. These predict increasing extremes of weather and climate, including generally higher temperatures, and higher likelihoods of heat waves and droughts, or of intense rainfall and floods, depending on geographic location on Earth.
Extreme Weather and Climate Events Have Been Increasing in Recent Years.
Two publications reviewed here, detailing extreme weather events, avoid reliance on climate models. Rather, these publications analyze actual weather and climate data, and use rigorous statistical methods to develop their results and conclusions.
Hansen and Coworkers Find That Extreme Events are Due to Increased Global Temperatures. James Hansen is a veteran climate scientist at the National Aeronautics and Space Administration’s Goddard Institute for Space Studies (GISS). He and his colleagues published “Perceptionof climate change” online in the Proceedings of the [U.S. August 6, 2012 . (See Details at the end of this post.)
They divided the Earth’s surface into a grid and used recorded temperature data for each grid location from 1951 to 2011. Their results show, for example, that many grid locations for years in the period 2006-2011 had average summer temperatures that were 2-3ºC (3.6-5.4ºF) greater than the temperature that the same grid location had during 1951-1980, which the authors assigned as their base period. (See Details) Several locations had annual average temperatures that were even higher.
Using rigorous statistical analysis the authors further showed that, compared to the same base period, the temperature variation for each of the decades 1981-1990, 1991-2000, and 2001-2011, shifted successively to higher temperatures. More and more grid locations had decadal average temperatures that were much higher than would be expected from the normal variability observed during the base period (see Details).
Hansen and coworkers call this unprecedented finding “probably the most important change” from the base period, one characterizing a “new category of ‘extremely hot summers’”. For example, the authors state there is a “high degree of confidence that events such as the extreme summer heat in [Russia Texas
The authors point out that specific weather events that some invoke as leading to global warming, such as blocking patterns and La Niña events, cannot be considered responsible for the observed warming trend of the last 3 decades. These patterns have been occurring for a long time, not just in the recent past. In contrast, the extreme warming, such as documented in their paper, has arisen only in conjunction with global warming.
Coumou and Rahmstorf Find That Extreme Weather Events Are Linked to Human Activity. They published the Perspective, “A decade of weather extremes”, in Nature Climate Change, July 2012 (published online: 25 March 2012 | doi: 10.1038/nclimate1452). They reviewed the work of others detailing extreme weather events occurring between 2000 and 2011 (see Details). Several record-breaking events were singled out, detailing economic impacts and human costs. While physical principles can help explain their occurrence, the authors rely on statistical analysis to identify true outliers among weather events. For past events, this drives home that in the decade under consideration, extreme events have occurred with unprecedented frequency. The number of new record hot days, analyzed on a monthly basis, is now more than three times that expected if the climate were not undergoing a long-term warming trend. Climate models help substantiate this trend; they predict higher record temperatures because of “human influence on the climate”.
Extreme rainfall events are also increasing in recent years. This too is understandable on a physical basis; warm air has the capacity to hold more water vapor than cool air (see this post). This extra moisture is then available for precipitation should it condense out of the air. The authors also discuss the important role that climate models play in understanding these events, and in helping attribute their origins to particular causes.
Importantly, the authors point out that “attribution is not a ‘yes or no’ issue as the media might prefer, it is an issue of probability.” Based on their analyses they conclude “it is very likely that several of the unprecedented extremes of the past decade would not have occurred without [man-made] global warming….now…the evidence is strong that [man-made], unprecedented heat and rainfall extremes are here — and are causing intense human suffering.”
The U. S. Is Experiencing Its Hottest Year-To-Date On Record. The National Climatic Data Center of the U. S. National Oceanic and Atmospheric Administration reports (accessed Aug. 10, 2012) that July 2012 was the hottest month on record, since recordkeeping began in 1895. The July average temperature was 3.3ºF (1.8ºC) above the average for that month. In addition, the period January-July 2012 was the hottest year-to-date, and the previous 12 month period was the hottest for that window on record.
The January-July 2012 period was also the twelfth driest recorded. As of early August 2012 “moderate to exceptional drought” conditions prevailed over 62.9% of the 48 contiguous states of the U. S. Current news accounts relate that the drought conditions are already reducing U. S.
While these phenomena are consistent with expected trends arising from a long-term increase in the U. S.
Analysis
The papers by Hansen and coworkers, and Coumou and Rahmstorf, examine recent extreme temperature and precipitation events. Their results are highly significant, for each separately shows by statistical analysis of past temperature records that the differences of the observed temperatures from expected averages are so large that they would be extremely improbable in a climate that was not warming. Rather, the occurrence of these extreme events is a hallmark of the warming of the planet. The authors conclude that it is very highly probable that these events arise from that warming, perhaps added to effects of other, more short-term climate patterns (La Niña/El Niño, jet stream blocking patterns). (These other effects have always been occurring, including during times that predate the rise in global average temperature. In those times extreme events such as are now occurring did not arise.)
These papers complement two reports over a year ago that for the first time directly linked extremes of rainfall and flooding to the warming of the average global temperature as a result of greenhouse gas emissions (see this post). Statistical analyses of rainfall and flooding were used to correlate observed rainfall patterns or flooding resulting from heavy rains to the predictions of several climate models.
Most CO2, and certain other greenhouse gases, persist in the atmosphere for hundreds, or even a thousand years. For this reason, within the 1-3 decade time frame under discussion for abating long-term warming of the planet, their concentration in the atmosphere can never be significantly reduced, but at best can only be held constant at today’s level or the higher level reached at some time in the future. Even if all emissions were to cease right away, the CO2concentration would remain constant. That is why the European Union’s Roadmap, to lower emissions by 80-95% by 2050, is so significant and praiseworthy. Even so, however, new emissions from those nations will continue to accumulate, at lower and lower annual rates, so that concentrations by 2050 will still be higher than now because of continued emission during the intervening years.
The use of fossil fuels for energy around the world is projected to increase, not to level off or decrease, in the foreseeable future. The U. S. Energy Information Agency projects worldwide energy usage from 2008 through 2035 will increase, reaching 53% higher in 2035 than in 2008. Much of that increase will arise in China India
This means that the occurrence of extreme weather events such as analyzed here will be more frequent and/or intense, and damages inflicted on humanity as a result (see the table in Details and this post, for example) will likewise be more severe. The economic expenses of responding to weather-inflicted tragedies are borne by governments and their taxpayers, as well as by private insurers. Any food shortages from droughts or floods will be felt worldwide as increased food prices.
Dealing with the worsening increases in global average temperatures can be approximated as a zero-sum enterprise. On the one hand, the governments of the world, and private corporations, can undertake investment as soon as possible to develop energy alternatives that contribute to lowering the rate of greenhouse gas emissions. Those investments would have the beneficial effect of contributing to a reduction in extreme weather-related disasters and their resulting damages. On the other hand, we can persist in business-as-usual, with the certainty that relief and restitution expenditures will continue increasing. The choice is ours to make.
Details
Hansen and coworkers analyzed world-wide weather data accumulated at GISS. On a globally based grid they determined average annual temperatures and standard deviations (SD values, measures of variability) of the annual temperatures for each grid position.
(Standard deviation is a measure of variability observed in repeated measurements for a particular data item such as temperature; in this case the variability is over the time interval considered. For a typical “bell shaped” curve expected for random variations of the data item, 1 SD on either side of the average value at the top of the “bell” contains 67% of all the measurements; 2 SDs on either side contains 97% of all the measurements; and 3 SDs contain 99.8% of all the measurements.)
Hansen and coworkers evaluated data sets covering 1951 to 2011. They used the period 1951-1980 to establish baseline values for the averages and the SDs for each grid position against which differences that have arisen in more recent years were determined.
Examples for one year from the base period, 1965, and the recent year 2010 are shown below.
Global grids in the years 1965 and 2010. The maps show temperature differences from average values for each location determined over 1951-1980 in ºC for the months June through August. Each grid location is color coded for the difference from the average for that location according to the color scale at the bottom (gray indicates locations with no data).
Source: Proceedings of the [U.S.
The map display for 1965, within the base period, shows that globally, temperature differences from local averages were both negative (cooler) and positive (warmer) compared to the averages. The magnitudes of the differences, using the color scale for temperature in ºC, were relatively small. On the other hand, many grid locations for years in the period 2006-2011 had average annual temperatures that were 2-3ºC (3.6-5.4ºF) greater than the temperature that the same grid location had in the base period of 1951-1980; and indeed several locations had annual average temperatures that were even more than 3ºC higher than those in the base period. The global display for 2010, above, shows that extreme hot weather prevailed over a large part of Eurasian Russia, and relatively hot areas in portions of China U. S.
The authors continued with a rigorous statistical analysis of the data they collated, shown below.
Frequency distributions for each value of the variability (SD) found for successive ten-year global average temperatures. These curves may be considered as highly compressed histograms showing the fractional occurrence for each value of SD. (All curves sum to 1.000.) The curve for a set of numbers that would be found from fully random valuations about the average (“bell-shaped curve”) is shown in black. Any deviation from the bell-shaped curve shows the existence of a bias in the distribution of the values. Decades in the base period are: crimson, 1951-1961; yellow, 1961-1971; and green, 1971-1981. Decades showing warming are aqua, 1981-1991; dark blue, 1991-2001; and magenta, 2001-2011.
Source: Proceedings of the [U.S.
Using the base period 1951-1981, they showed that ten-year distributions for the measure of variability, the standard deviation, closely followed the “bell-shaped curve” (black curve in the figure) expected for random, unbiased data sets for the three decades of the base period (as is expected). But for each of the decades 1981-1991, 1991-2001, and 2001-2011, the variability distribution shifts more and more to higher (i.e. positive) standard deviation values as time passes. This indicates that more and more grid locations had decadal average temperatures that were much higher than would be expected from the normal variability of the base period. Whereas the base period had essentially no grid locations with temperatures that were 3 standard deviations above the average (the unbiased bell curve would have only 1 in 769 data points at that position), by the time of the most recent decade, about 10% (1 in 10 data points for land-based grid locations) had temperatures 3 standard deviations above the average, and many data points whose temperatures gave even 4 and 5 standard deviations above the averages of the base period (see the figure).
Coumou and Rahmstorf reviewed extreme weather events during the decade of 2000 to 2011. These included both high temperature events and heavy rainfall events, and their consequences. Examples they chose for mention are shown in the following table.
Year | Region | Meteorological record-breaking event | Impact, costs |
2000 | Wettest autumn on record83 since 1766. | £1.3 billion (ref. 27). | |
2002 | Highest daily rainfall record in Germany42 since at least 1901. | Flooding of | |
2003 | Hottest summer in at least 500 years30. | Death toll exceeding 70,000 (ref. 31). | |
2004 | First hurricane in the | Three deaths, US$425 million damage85. | |
2005 | Record number of tropical storms, hurricanes and category 5 hurricanes52 since 1970. | Costliest | |
2007 | Strongest tropical cyclone in the Arabian Sea53 since 1970. | Biggest natural disaster in the history of Oman53. | |
May–July wettest since records began in 1766 (ref. 43). | Major flooding causing ~£3 billion damage. | ||
Hottest summer on record in Greece33 since 1891. | Devastating wildfires. | ||
2009 | Heatwave breaking many station temperature records (32–154 years of data)34 | Worst bushfires on record, 173 deaths, 3,500 houses destroyed34. | |
2010 | Hottest summer since 1500 (ref. 69). | 500 wildfires around | |
Rainfall records44. | Worst flooding in | ||
Eastern Australia | Highest December rainfall recorded since 1900 (ref. 45). | Brisbane flooding in January 2011, costing 23 lives and an estimated US$2.55 billion86 | |
2011 | Most active tornado month on record (April)3 since 1950. | Tornado hit | |
January–October wettest on record1 since 1880. | Severe floods when Hurricane Irene hit. | ||
Most extreme July heat and drought since 18802. | Wildfires burning 3 million acres (preliminary impact of US$6–8 billion). | ||
Hottest and driest spring on record in France1 since 1880. | French grain harvest down by 12%. | ||
Wettest summer on record (The Netherlands, | Not yet documented. | ||
72-hour rainfall record (Nara Prefecture)1. | 73 deaths, 20 missing, severe damage. | ||
Wettest summer on record1 since 1908. | Flooding of |
© 2012 Macmillan Publishers Limited
The causes of extreme weather events can be difficult to pin down. Climate models are useful in projecting long-term and large scale trends across broad regions of the planet. They are less adept at correlating local or regional weather events such as these with large scale trends. For this reason, the authors note, rigorous statistical analysis is helpful to identify that a particular event actually falls outside the realm of what would be expected if the climate were not changing due, for example, to the man-made greenhouse effect.
Statistical evaluation of extreme heat events shows that, considered globally, the number of monthly average temperature records occurring at present is as high as three times what would be expected if the climate were not changing to higher temperatures. This is shown in the graphic below.
© 2012 Macmillan Publishers Limited
Source: Nature Climate Science doi: 10.1038/nclimate1452
It is seen that the ratio of the observed number of monthly average records to that expected in an unchanging climate begins to grow higher than the value of 1 as early as about 1920. The ratio has grown to values in the range of 3.0 to 3.5 by 1990.
The table above also includes many extreme rainfall events and the floods that ensued from them.
Extremes will result not simply from the long-term trend of increasing average global temperatures. Coumou and Rahmstorf point out, as do Hansen and coworkers, that the Pacific Ocean pattern of La Niña/El Niño/Southern Oscillation, as well as blocked jet stream patterns, affect shorter term climate and weather events. They conclude that statistical analysis of past weather data, modeling of climate events, and reasoning based on physical principles all have a place in assessing the progress of global warming. They find that heatwaves and precipitation extremes have already greatly increased and will continue to do so as the climate warms further.
Coumou and Rahmstorf include a graphic that originally appeared in a paper by Barriopedro, D. et al. (Science 332, 220–224 (2011)) dealing with the temperature extreme in Russia
0 comments:
Post a Comment