Climate variability is meaningful only when there is a standard of comparison for assessing the perceived deviation. Establishing the climate standard is complicated by the relatively short history of human existence on Earth. The instrument record is still shorter. An appreciation of the climate record is needed to define the context for hydroclimate variability in the past, present, and future. A comprehensive analysis is beyond the scope of this book, but an overview of selected concepts is presented to illustrate the nature of the time scale differences.
The Earth's age is about 4.5 billion years, and the vast majority of this history lies outside of the human perspective (Saltzman, 2002). Little or no information is known about the earliest parts of the Earth's history because the climate system has been a major agent in sweeping away evidence. The Earth's surface changed through time as its plates moved into different alignments and erosion and tectonic activity altered continental configurations. Changes in atmospheric composition and the radiative properties of the atmosphere accompanied the progressive changes in terrestrial and oceanic environments. The changes in the world's continents, oceans, and atmosphere produced a global climate that varied over geologic time.
A general understanding of ancient climatic conditions, known as paleo-climates, has been assembled for specific parts of the geologic past. Since paleoclimates existed before humans began collecting weather data, they are based on reconstructed climates inferred from natural environmental records rather than the observational data that form the framework for expressing present hydroclimates. Because paleoclimate data are highly specific for a given site, they may represent local anomalies rather than large-scale variations. Nevertheless, these reconstructed climates provide a wide range of useful knowledge about the climate system. By extrapolation, we can use this knowledge to expand our understanding of hydroclimate. Paleoclimate insights that are especially transferable to hydroclimate are knowledge of how the climate system works, the range over which climate can vary and has actually varied, the relative stability and/or instability of the climate system, and how fast and to what degree climate can vary.
Paleoclimate evidence indicates that global temperatures varied widely during the last 3.8 billion years, and the global climate was warmer and wetter than today for most of Earth's history. Although periods with widespread glaciation were common early in the Earth's history, the Earth's basic climate pattern has existed for the past 1 billion years even though the position of continents has changed. An exception is the Mesozoic era, between 245 million years ago and 65 million years ago, when global climate was warmer and drier and continents were clustered around the South Pole. The Cretaceous period during the last 80 million years of the Mesozoic era had a globally averaged surface temperature 6-12 °C warmer than the present temperature. Modern tropical to subtropical conditions extended to about 45° N, and higher latitudes were warmer than today. Tropical surface water temperatures were similar to or slightly warmer than temperatures today, but oceanic deep water was warmer. Warmer polar regions supported abundant high-latitude floras and faunas, and there is no direct evidence of polar ice caps (Burroughs, 2001; Saltzman, 2002).
Warm and wet conditions continued into the Tertiary period, but a cold and arid climate resembling today's was dominant by the end of the Oligocene epoch about 25 million years ago. A slight climatic amelioration occurred during the Miocene epoch, but temperatures and precipitation continued downward during the Pliocene epoch. By the end of the Pliocene about 2 million years ago, the climate was colder and drier than today. This set the stage for the most recent and best known of the glacial epochs spanning the Pleistocene. Temperature changes during ice ages were greatest at the high latitudes and small near the equator. Tropical land areas were drier on average during glacial maxima than during warm periods. The global mean temperature was about 5 °C colder than now 20 000 years ago during a full-glacial episode. Temperature increases during the present Holocene interglacial have returned global temperatures to about the level of those occurring 125 000 years ago, and altered precipitation, evaporation, and runoff during this time have affected the expansion and desiccation of lakes worldwide (Melack, 1992). Since the current climate is very near the warmest that has been observed in the last million years, we have little information about what constitutes a really warm climate and how the environment will respond.
What emerges from this paleoclimate background is that our understanding of past climates degrades with time. We have a very good understanding of climate over the past several thousand years, but a poor understanding of climate beyond several hundred million years. Nevertheless, it is evident that during the past 10 000 years, as civilizations developed, our climate has been remarkably stable. In the context of the paleoclimate record, this stability appears to be highly unusual.
Paleoclimates are reconstructed using geologic data and proxy data that serve as a substitute for actual climatic records. Proxy data are physical, biological, and chemical information present in natural features of the environment that are indicators of a particular climate or climate variable. These natural recording systems provide indirect evidence of climatic trends over geologic time. Proxies include tree rings, pollen, ice cores, lake sediments and shorelines, relic soils, marine shorelines, corals, and deep-sea sediments. These natural climate archives substitute for thermometers, rain gauges, and other modern instruments used to record climate, but they have a wide range of differences in terms of the resolution of the climate element they portray. The utility of proxy climate data is achieved by calibrating the proxy with the available instrument record using established dating methods as described by Bradley (1999). The key to meaningful paleoclimatic reconstruction is the interaction between proxy phenomena and present climate.
For hydroclimatic purposes, the most extensive and highest resolution natural record is provided by annual tree rings from temperate and boreal forests (Saltzman, 2002). However, this record is relatively short compared to the time line available using other proxy data. Combining information from tree rings, ice cores, lake sediments, corals, and deep-sea sediments provides improved understanding of climate and the climate system for the past several thousand years. Reconstructing precise atmospheric conditions is hindered by the limited spatial data on which to base definition of a pattern that can be related to reasonable atmospheric circulation features, but some success has been achieved. The stable hydrogen isotope composition of bristlecone pine tree rings has yielded information on temperature and air mass trajectories for North America during the past 8000 years (Feng and Epstein, 1994). Ocean sediment cores have revealed evidence linking shifts of the ITCZ to climate change in Central and South America during the last 14 000 years (Haug et al. 2001). Ice cores from Greenland and West Antarctica have contributed to developing temperature patterns for the past 90 000 years through analysis of gases trapped in the annual ice layers at different depths and inference of a relationship between temperature and the changing composition of the atmosphere through time (Blunier and Brook, 2001). Similarly, pollen in lake sediments has been employed to estimate regional temperature and precipitation variations for the last 130000 years (Adam and West, 1983). For periods covering the past several hundred million years, a highresolution record is being sought from deep-sea sediments. However, this task is difficult because most of the sea floor older than 200 million years has been subducted as part of tectonic recycling, and the hydroclimatic information contained in these sediments is lost (Saltzman, 2002).
The relatively brief record of instrument data severely hinders efforts to develop a comprehensive understanding of hydroclimate over the past 1000 years. For the immediate pre-instrument period, seasonal and monthly temperature and precipitation can be gleaned from a variety of written and illustrated sources ranging from crop harvest dates to artistic works. These historic data are in documentary form rather than quantitative and are commonly grouped into four categories. First is the observations of specific weather phenomena, such as the first day of frost or the occurrence of snow, which are frequent entries in personal journals and diaries. A second category is the dates of floods, drought, and other weather-dependent natural phenomena, sometimes called parameteorological phenomena, which appear in municipal and governmental reports. The third group is phenological records that deal with the timing of recurrent weather-dependent biological phenomena, such as crop harvest dates or crop yield data (Bradley and Jones, 1995). A fourth group is records of forcing factors, such as sunspot activity, that influence climatic conditions. Utilizing this historical information requires recognition that the information is subjective, sketchy and limited to a few geographical areas. In addition, extreme events that have a high probability of appearing in historical accounts may not be representative of the overall climate (Bradley, 1999).
Reconstruction of past weather and climate using anthropogenic climatic data in the form of written and illustrated records found in archives, libraries, and museums constitutes hydroclimatic documentary data. In one sense, this is another form of proxy data, but documentary data are derived through a human filter rather than being produced by a natural biological or chemical response to climate. The accurate and appropriate use of documentary sources involves a rigorous methodology for isolating reliable materials from those less faithful in conveying climatic information. Content analysis has been useful as a technique for assessing this information in quantitative terms, allowing the documentary data to be calibrated with instrumental measurements to estimate specific meteorological variables (Bradley and Jones, 1995).
Archives of central and local governments have been the source of some of the most useful documentary data. An illustrative case is the dates of tithe auctions in Western and Central Europe. Tithe auction dates are highly correlated with wine harvest dates and with Central England temperature time-series. The tithe dates are earlier than the mean during warm periods and are later than the mean during cold periods (Pfister, 1980). Similarly strong relationships are found between wine harvest dates and wine yields in France, Switzerland, and other parts of Europe. Grapevines are good climatic indicators because the plant remains the same for 25 to 50 years and does not require annual planting. Also, the entire length of the growing season from March or April to October is needed to bring grapes to maturity. This means that harvest date, yield per acre, and wine quality can be used as climatic proxy evidence for three different periods of the growing season, even though the date of harvest is the most popular proxy and the least subject to interpretation controversy (Ladurie and Baulant, 1980).
Although the overwhelming majority of documentary data come from written records, pictorial documents have proven useful. Paintings of Alpine glaciers in Europe record changes in glaciers since the sixteenth century, and cloud cover depicted by Dutch and British landscape painters since 1550 shows summer atmospheric characteristics. However, the use of such materials involves major interpretative difficulties related to subjective biases and the fallibility of the untrained observer. This presents major problems in quantifying the record so we really know what the event is telling us.
The utility of documentary data for developing time-series of hydroclimatic data is constrained because the period of observation is variable and often incomplete and/or discontinuous. Spatial coverage further constrains documentary data because the availability of these data is linked to settlement histories (Bradley, 1999). Documentary data suitable for climatic interpretation are available for about 700 years for Europe and Asia, but the earliest data for Australia begins about the mid to late 1700s. For western North America, a catalog of historical agricultural disasters in Mexico provides a basis for identifying historical droughts from 1450 to 1899 (Mendoza et al., 2006). The spread of Spanish presidios and missions into California between 1776 and 1834 provides an important source of documentary data. Annual reports of the crops planted and harvested at the presidios and missions have been transformed into estimates of rainfall variability in southern California for this period (Rowntree, 1985).
Derived climate histories using all available sources of documentary data can provide very useful numerical expressions of climate and valuable information
Fig 8.1. Cross-section of a hickory stem showing annual growth rings in the heartwood (darker wood in the center) and the sapwood (lighter wood near the bark). The small variation in tree-ring widths indicates the tree experienced relatively little environmental stress each year. (Photo by author.)
on climate trends. Lamb (1982) used documentary data to estimate temperature and precipitation for England and Wales for 50-year blocks beginning in AD 1100. What stands out in these data is the certainty of a warmer period that lasted several centuries in the Middle Ages (1100-1300) that is commonly identified as the Medieval Warm Epoch or Little Optimum. An equally long colder period follows from 1500-1700 corresponding to the Little Ice Age, characterized by wetter summers and more frequent severe winters in Europe. This record emphasizes that the Little Ice Age (c. 1500-1800) occurred at the transition from documentary data to instrument data. Instrument data are reliable for Central England beginning in about 1700.
Documentary evidence is a promising source of data regarding the details of short-term shifts and fluctuations for the millennium or so immediately preceding the era of modern instrumentation. Documentary data may be a particularly important source of information where other proxy data, such as tree rings, are not available. Bradley (1999) discusses the nature of documentary data and emphasizes its use in regional climate studies.
Was this article helpful?