The Climate Record
The clearest signal of an enhanced greenhouse effect, or greenhouse warming, that is projected by climate models is a widespread and substantial increase of surface temperature, especially over continental areas. For this reason, special attention has been given to the observed temporal and spatial variations of land surface temperatures (typically measured 1 to 2 m (3.3 to 6.6 feet) above ground level) and sea surface temperatures (typically measured in the layer from the surface to several meters below the ocean's surface). Estimates of global temperature change are credible only for measurements that have been taken since the late nineteenth century. Earlier temperature data are available, but their spatial distribution and their sparseness seriously compromise the attempts to infer global averages. Additional difficulties arise because measurements were not taken with an eye to the small and gradual global average changes that are characteristic of climate change. This problem is exacerbated when regional changes are at issue simply because the statistical amelioration of offsetting bias is not very helpful when the data are sparse.
The most comprehensive assessment of the instrumental record of surface temperature has recently been completed by the Intergovernmental Panel on Climate Change (1990). A significant warming has been observed since the late nineteenth century. Uncertainties underlie the interpretation of the data, but there is consensus that the actual rise probably lies within the range of 0.3° to 0.6°C (0.5° to 1.1°F) (see Figure 14.1). Qualitatively, an inference of global warming is supported by several independent sources of information: (1) land and sea surface temperatures are independently measured, and both depict significant warming (Figure 14.2), (2) the collective
behavior of many glaciers is likely to be sensitive to large-scale temperature variations, and many mountain glaciers have retreated since the end of the nineteenth century, some of them markedly, and (3) estimates of sea level variations over the past century indicate a general rise in sea level that is qualitatively consistent with the global increase of surface temperature.
Uncertainties in the Record
Despite the fact that there is virtually unanimous agreement that global temperatures have increased since the nineteenth century, the exact magnitude is still somewhat uncertain for a variety of reasons. For land-based surface temperatures, these uncertainties arise for the following reasons:
• Station locations have changed and thereby have introduced uncertainties into the interpretation of the temporal record.
• Many stations are located in and around cities that have increased in size and population density, thereby introducing a systematic increase in the local temperature record that is unrelated to global climate change.
Additional uncertainties for both land and sea surface temperatures include the following:
• The spatial coverage of the data is neither globally comprehensive nor optimally sited, and the degree of coverage changes over time.
• Changes have occurred in instruments, exposures of thermometers, observing schedules, observing practices, and other factors including land use, thus introducing artificial changes in the temperature record, sometimes of a systematic nature.
Most of these uncertainties have been assessed in various studies, as reviewed by Karl et al. (1989). Arguably the most serious of these uncertainties
are the changes in observation practices related to sea surface temperature measurements. Changes in these practices have resulted in systematic biases as large as the observed warming depicted in Figure 14.1. The sea surface temperature has been measured by the use of various types of buckets lowered from the ship's deck, and more recently by the water flowing to the ship's engine intake. Various researchers have addressed this problem (Farmer et al., 1989; Bottomley et al., 1990; Folland and Parker, 1990). Differences in the assumptions, in approaches used to correct for these systematic biases, and in the input data lead to an uncertainty associated with the warming since the nineteenth century of about ±0.1°C (±0.18°F).
Urbanization effects in the thermometric record have now been assessed by various researchers for many, but by no means all, regions of the globe. Studies by Hansen and Lebedeff (1987, 1988), Balling and Idso (1989), Karl et al. (1988), Karl and Jones (1989, 1990), and Jones et al. (1989, 1990) suggest that the bias introduced into the observed warming of the land surface varies from a high of 0.1° to 0.2°C (0.18° to 0.36°F) over the United States down to negligible amounts over western areas of the former USSR and easter China. Jones et al. (1990) estimated that the apparent global warming of the land surface record attributable to urbanization is no more than about 0.05°C (0.09°F). When one considers that the land covers about three-tenths of the global surface area, the bias due to urbanization in the global temperature record is most likely to be at least an order of magnitude smaller than observed warming. A more definitive statement awaits more comprehensive studies.
Another important consideration in the compilation of hemispheric and global temperatures is the spatial coverage of observations over both land and sea. The spatial coverage of the observations since the nineteenth century has not been uniform in time or space. For example, ships of opportunity, which provide the greatest portion of the sea surface measurements, have followed preferred navigational routes and consequently leave vast portions of the ocean inadequately sampled. The effect of spatial sampling inadequacies has been addressed for the land areas by Jones et al. (1986a,b) and for the ocean areas by Bottomley et al. (1990) and Folland and Colman (1988). These analyses suggest that changes in spatial coverage have not had a serious impact on our ability to discern changes in the global thermometric record. Because of the large-scale coherence of multi-decadal climate variations, especially over the oceans, the sampling problem probably adds less than 0.1°C (0.18°F) uncertainty to the observed warming, as depicted in Figure 14.1.
Other biases and inhomogeneities in the temperature record have been directly assessed only on a regional basis, but indirectly assessed on a global basis. These include changes in observing schedules, types of instrument shelters used, and station relocations in the land-based record. Studies
by Jones et al. (1986a), Karl and Jones (1990), and Parker (1990) indicate that, although these types of biases may be quite significant on a regional scale, for hemispheric and global averages many of the biases and inhomogeneities tend to cancel.
Considering all the uncertainties in the thermometric climate record, it would be most appropriate to provide a range of most likely rates of warming from the latter half of the nineteenth century to the present. This range, as given by the IPCC (Intergovernmental Panel on Climate Change, 1990), is of the order 0.3° to 0.6°C (0.5° to 1.1°F). That is, the warming rate, calculated either as a trend from the 1860 to 1890 era to the present, or as the difference in temperature for the 1980s minus the value for the period from the 1860s to the 1890s, is between 0.3° and 0.6°C (0.5° and 1.1°F). At this time, there are no compelling reasons to dispute the range given. Whether this warming is related to the greenhouse effect is discussed in the next section.
Spatial and Temporal Characteristics
The increase in temperature has occurred somewhat differently in the northern hemisphere than in the southern hemisphere (Figure 14.1). A rapid increase in temperature occurred in the northern hemisphere during the 1920s into the 1930s, whereas the increase in the southern hemisphere during this time was much more gradual. A substantial portion of the global warming observed since the latter half of the nineteenth century occurred prior to World War II. The past half-century is characterized by relatively constant global mean temperatures except for the 0.1° to 0.2°C (0.18° to 0.36°F) steplike temperature increase that occurred around 1980. This relatively constant global mean temperature is characteristic of both hemispheres, but with some evidence of cooling in the northern hemisphere during the 1960s and 1970s. Some analyses suggest that the northern hemisphere cooling may be related to sulfate aerosol emissions (e.g., Wigley, 1991). The warmth of the 1980s has continued into the 1990s.
The global mean ocean temperature and the surface air temperature of the southern hemisphere (which is predominantly water) have risen at a rate comparable to that over land (Figure 14.3a). This is not altogether consistent with time-dependent GCM simulations with enhanced greenhouse gases. These simulations generally have a slower rate of temperature increase in the southern hemisphere and over the oceans, especially the circumpolar oceans, than in the northern hemisphere (Stouffer et al., 1989; Washington and Meehl, 1989; Manabe et al., 1990). At this time, however, the large variability in transient climate simulations and the difficulty in linking these simulations to the appropriate part of the climate record prevent unequivocal statements about the consistency or lack thereof between simulations and the climate record.
Climate models that have been allowed to reach equilibrium with 2 (sometimes 4) times preindustrial concentrations of CO2 project temperature increases in polar and high latitudes that would be larger than those at low latitudes. These are related to the extent of land and sea ice and other factors. Currently, the largest variations in temperature are found at high latitudes (Figure 14.3). Changes of temperature as depicted in Figure 14.3d are over too short a period to reflect systematic changes in the warmth of the equatorial region relative to polar regions.
Because of the high natural variability of climate, it is pertinent to ask when we might expect to unequivocally link the greenhouse effect to the observed climate variations. Two recent studies address this question on a regional basis assuming validity of the IPCC projections. Briffa et al. (1990) used the variability in a 1400-year reconstructed tree-ring density record to address this question for Fennoscandia. Briffa et al. (1990) indicated that because of low-frequency climate variability, it will be decades before we will be able to confirm or contradict the IPCC projections in the Fennoscandia region. Karl et al. (1990) used the central North American observed climate record to determine when the IPCC projections could either be confirmed or contradicted for that region. Karl et al. (1990) indicated that if the models with a 4°C (7.2°F) (i.e., high) climate sensitivity to enhanced greenhouse gases are used, we probably should have already detected significant warming in this region, but for the low-sensitivity model (2°C (3.6°F) warming in central North America from preindustrial times to 2030) we probably will not be able to detect statistically significant warming rates for at least another two decades. Thus, at this time, no credible assertion that the recent temperature rise is the result of changes in greenhouse gas forcing is justified.
Changes in Extremes
Long-term changes and variations of extreme maximum and minimum temperatures have recently been assessed over three large land areas. These include the United States, the former USSR, and the People's Republic of China (PRC). Until now, the dearth of analyses about extremes has been due to the difficulty in transferring these observations to computer-compatible media. Recent work by Karl et al. (1991) indicates that the warming in the United States and the former USSR over the past several decades is solely due to an increase of nighttime temperatures (Figure 14.4). Furthermore, the range of extreme temperatures, especially during summer in the United States and the former USSR, has also decreased (Karl et al., 1991). Climate models with enhanced greenhouse gases cannot yet reliably predict the expected change in extremes. It has been demonstrated that in the
United States the rise in the minimum temperature is at least partially related to increases in cloud cover (Plantico et al., 1990).
The differential change of the maximum and minimum temperatures points out the limited amount of information that can be derived from changes in the mean and how cloud changes may already be influencing the observed climate record. Thus long-term consistent ancillary climatic data (such as vertical and horizontal cloud distribution) are needed in addition to basic temperature data.
Temperatures aloft have been measured by instruments carried by balloons and, most recently, by satellite instruments. Global records have been constructed for various atmospheric layers by Angell (1988) back to 1957 using balloon or radiosonde information. Spencer and Christy (1990) have used the space-based Microwave Sounding Unit (MSU) data to reconstruct global temperature variations back to 1979 for tropospheric layer temperatures. The global mean annual temperatures of MSU data are in excellent agreement (r2 = 0.95) with the mid-tropospheric temperatures as calculated by Angell (1988). This comparison is consistent with the less complete spatial coverage of the data used by Angell (1988), at least for the past decade.
Climate models with enhanced greenhouse gases project not only increased temperatures in the troposphere but also decreased temperatures in the stratosphere (in addition to the decrease accompanying depletion of stratospheric ozone). Mid-tropospheric temperatures since 1957 have shown an increase, and stratospheric temperatures a decrease. In addition, however, temperatures in the higher troposphere have also decreased, a circumstance that is not projected by climate models that include CO2 changes only. Furthermore, the decrease in stratospheric temperatures is primarily found in the southern hemisphere in and around Antarctica, probably related to the decrease in stratospheric ozone (Trenberth and Olson, 1989). Finally, a recent study has pointed out the danger of overinterpreting mid-tropospheric temperature trends based on short time series. Whereas previous studies showed current temperature increases compared to 1957, Elms et al. (1990) extended their examination back to 1950 and found little or no temperature change from that date.
Unlike model projections of changes in temperature with an enhanced greenhouse effect, or greenhouse warming, regional and geographic patterns
of precipitation change are not nearly as consistent from one model to the next. Nonetheless, some general characteristics are evident, such as (1) a global increase in precipitation; (2) significantly more precipitation at high latitudes and in the tropics throughout the year and during winter at mid-latitudes, but little change over the subtropics; and, of particular interest for the United States, (3) a suggestion in several but not all of the models of reduced summer and enhanced winter precipitation in the interior of North America.
The instrumental record related to global changes of precipitation is much more difficult to assess than the thermometric record. Nonetheless, there have been several large-scale analyses of precipitation variations over the northern and southern hemisphere land masses (Bradley et al., 1987; Diaz et al., 1989; Vinnikov et al., 1990), which have demonstrated that during the past several decades precipitation has tended to increase in northern hemisphere mid-latitudes and in the southern hemisphere. The increase in the northern hemisphere is predominantly due to an increase over the former USSR, and to a smaller degree over the United States (Figure 14.5). A decrease in precipitation over the past few decades has been noted over the subropics of the northern hemisphere. This is primarily due to the decrease observed over the Sahel, and partially due to lower monsoonal rainfall over India. Changes of precipitation over oceanic regions are largely unknown.
Overall, the land data do suggest a general increase in land-based precipitation over the past few decades. This may be consistent with model projections, but quantitative inferences are difficult to defend. Precipitation varies in space and time much more than does temperature. A much higher density of stations is needed to calculate precipitation variations than is required for temperature. Such high-density data do exist, but they are not exchanged on an international basis. Furthermore, the efficiency of the collection of precipitation (both frozen and liquid) by different rain gauges varies with wind speed and raindrop size. In many countries, collection efficiency has tended to increase as operational practices have improved, often in poorly documented ways, possibly lending to an artificial upward trend in some regions (Groisman et al., 1991). Sevruk (1982) and Folland (1988) discuss the causes of the collection error. Despite these potentially serious biases, many of the important variations apparent in the precipitation records are evident in hydrological data, such as changes in lake levels and stream flow and the severe desiccation of the Sahel.
In the recent IPCC scientific assessment (Intergovernmental Panel on Climate Change, 1990) and summary report (Jäger and Ferguson, 1991), several scenarios of changes in North American precipitation were given based on various models (a decrease in summer precipitation up to 10 percent and an increase in winter precipitation up to 15 percent). Karl et al. (1990) show that even if these projections were correct, it would take decades
before we could detect such changes. On the other hand, at the present time there is neither any evidence for increased aridity in the United States nor any systematic increase in winter to summer precipitation ratios (Karl et al., 1990). Furthermore, if the rate of change for the increase of winter to summer precipitation ratios suggested by the modeling reported by the IPCC were correct, we probably should have already detected such changes. All in all, the record cannot support persuasively a quantitative relationship between the greenhouse gas accumulation and the precipitation variations of the last century.
Improving our Climate Observations
At the present time the only reliable methods of measuring daily temperature and precipitation at the surface are the in situ measurements made at the thousands of weather stations around the world. Unfortunately, few (if any) of the stations make measurements for the purposes of detecting and monitoring climate change. The primary functions are to serve the needs of weather prediction and the weather assessment community. Such activities are quite distinct from monitoring climate change. The World Meteorological Organization is beginning to take some steps to rectify this situation (e.g., the Global Climate Baseline Data Base and the Climate Change Detection Project), but without strong support from national weather services, which currently do not operate their stations for the purpose of monitoring climate change, these programs will not succeed. That support will become important as new observing techniques and programs are brought on-line.
In the United States the National Oceanic and Atmospheric Administration (NOAA) is responsible for observing the nation's daily weather. NOAA operated 8,640 stations that monitored daily precipitation in 1967, but by 1988 the total had dropped to about 7,000. The totals for temperature monitoring dropped from 5,920 stations in 1967 to about 5,200 in 1988. The importance of these temperature and precipitation observation stations is critical, because many of them are located in rural environments, free from the urban heat island effect. This decrease in number has occurred during times of increased interest in global climate change. A dense network of stations is required to monitor the regional details of climate change, especially for precipitation, and to implement corrections for biases in the climate record. Other U.S. agencies also monitor the daily weather for a variety of purposes. At present, there is no central collection agency that can make available the full suite of weather observations made by the various U.S. agencies. This unfortunate situation should be rectified.
Furthermore, potentially new sources of bias have been and continue to be introduced into the NOAA network without an adequate means of assessing
these biases. For example, a new thermometric instrument was installed at about half of the stations in the U.S. cooperative network between 1984 and 1988. At the present time, the bias introduced by this new instrument is significant with respect to interpreting U.S. climate change (Quayle et al., 1991). Side-by-side measurements of new and old measurement systems were not required. Such tests for a period of at least one year (the annual cycle) should be considered essential for climate monitoring. In addition, this instrument did not have the ability to retain maximum and minimum temperatures on a midnight-to-midnight basis. As a result, the opportunity was lost to eliminate many of the thermometric biases that had existed in the network because observers were understandably unwilling to remain with a fixed observation schedule over the decades.
The United States operates one of the most comprehensive networks in the world, but it needs to be improved in several ways:
1. Steps should be taken to ensure the continued operation of long-term stations in rural areas.
2. The surface observations of climate-related elements by various departments (e.g., Commerce, Interior, Agriculture, and Defense) should be integrated and made available for more comprehensive analyses.
3. Data should be assessed for biases as well as random errors on a routine basis, and corrections should accompany the disseminated data.
4. Electronic transfer of weather station observations should become routine operating procedure.
5. NOAA will soon be implementing an Automated Surface Observing System (ASOS). The methods of making observations and the locations of many observing stations are scheduled to undergo significant change. NOAA should provide for side-by-side operation of the new and old instruments long enough to ensure that interpretation uncertainty is not introduced.
Angell, J. K. 1988. Variations and trends in tropospheric and stratospheric global temperatures, 1958–87. Journal of Climate 1:1296–1313.
Balling, R. C., Jr., and S. B. Idso. 1989. Historical temperature trends in the United States and the effect of urban population growth. Journal of Geophysical Research 94:3359–3363.
Bottomley, M., C. K. Folland, J. Hsiung, R. E. Newell, and D. E. Parker. 1990. Global Ocean Surface Temperature Atlas (GOSTA). Joint Meteorological Office/Massachusetts Institute of Technology Project. London: U.K. Department of Energy and Environment.
Bradley, R. S., H. F. Diaz, J. K. Eischeid, P. D. Jones, P. M. Kelly, and C. M. Goodess. 1987. Precipitation fluctuations over northern hemisphere land areas since the mid-19th century. Science 237:171–175.
Briffa, K. R., T. S. Bartholin, D. Eckstein, P. D. Jones, W. Karlen, F. H. Schweingruber, and P. Zetterberg. 1990. A 1400-year tree-ring record of summer temperatures in Fennoscandia. Nature 346:434–439.
Diaz, H. F., R. S. Bradley, and J. K. Eischeid. 1989. Precipitation fluctuations over global land areas since the late 1800s. Journal of Geophysical Research 94;1195–1210.
Elms, J. D., T. R. Karl, A. McNab, and J. Christy. 1990. A preliminary analysis of the covariation of surface and midtropospheric temperatures. Proposed for the 41st International Astronautical Congress, Dresden, Germany. American Institute of Aeronautics and Astronautics, Washington, D.C.
Farmer, G., T. M. L. Wigley, P. D. Jones, and M. Salmon. 1989. Documenting and explaining recent global-mean temperature changes. Norwich, U.K.: Climatic Research Unit.
Folland, C. K. 1988. Numerical models of the raingauge exposure problem, field experiments and an improved collector design. Quarterly Journal of the Royal Meteorological Society 114:1485–1516.
Folland, C. K., and A. W. Colman. 1988. An interim analysis of the leading covariance eigenvectors of worldwide sea surface temperature anomalies for 1951–80. LRFC 20. U.K.: U.K. Meteorological Office.
Folland, C. K., and D. E. Parker. 1990. Observed variations of sea surface temperature. In Proceedings of the NATO Advanced Research Workshop on Climate-Ocean Interaction. Norwell, Mass.: Kluwer Academic Press.
Groisman, P. Ya., V. V. Koknaeva, T. A. Belokrylova, and T. R. Karl. 1991. Overcoming biases of precipitation measurement: A history of the USSR experience. Bulletin of the American Meteorological Society 72(11):1725–1733.
Hansen, J., and S. Lebedeff. 1987. Global trends of measured surface air temperature. Journal of Geophysical Research 92:13345–13372.
Hansen, J., and S. Lebedeff. 1988. Global surface temperatures: Update through 1987. Geophysical Research Letters 15:323–326.
Intergovernmental Panel on Climate Change. 1990. Climate Change: The IPCC Scientific Assessment, J. T. Houghton, G. J. Jenkins, and J. J. Ephraums, eds. New York: Cambridge University Press.
Jäger, J., and H. L. Ferguson, eds. 1991. Climate Change: Science, Impacts and Policy: Proceedings of the Second World Climate Conference. New York: Cambridge University Press.
Jones, P. D. 1988. Hemispheric surface air temperature variations: Recent trends and an update to 1987. Journal of Climate 1:654–660.
Jones, P. D., S. C. B. Raper, R. S. Bradley, H. F. Diaz, P. M. Kelly, and T. M. L. Wigley. 1986a. Northern hemisphere surface air temperature variations, 1851–1984. Journal of Climate and Applied Meteorology 25:161–179.
Jones, P. D., S. C. B. Raper, R. S. Bradley, H. F. Diaz, P. M. Kelly, and T. M. L. Wigley. 1986b. Southern hemisphere surface air temperature variations, 1851–1984. Journal of Climate and Applied Meteorology 25:1213–1230.
Jones, P. D., P. M. Kelly, G. B. Goodess, and T. R. Karl. 1989. The effect of urban warming on the northern hemisphere temperature average. Journal of Climate 2:285–290.
Jones, P. D., P. Ya. Groisman, M. Coughlan, N. Plummer, W-C. Wang, and T. R. Karl. 1990. Assessment of urbanization effects in time-series of surface air-temperature over land. Nature 347:169–172.
Karl, T. R., and P. D. Jones. 1989. Urban bias in area-averaged surface temperature trends. Bulletin of the American Meteorological Society 70:265–270.
Karl, T. R., and P. D. Jones. 1990. Reply to comments on ''Urban bias in area-averaged surface temperature trends." Bulletin of the American Meteorological Society 71:572–574.
Karl, T. R., H. F. Diaz, and G. Kukla. 1988. Urbanization: Its detection in the United States' climate record. Journal of Climate 1:1099–1123.
Karl, T. R., D. Tarpley, R. G. Quayle, H. F. Diaz, D. A. Robinson, and R. S. Bradley. 1989. The recent climate record: What it can and cannot tell us. Reviews of Geophysics 27:405–430.
Karl, T. R., R. R. Heim, Jr., and R. G. Quayle. 1990. The greenhouse effect in central North America: If not now, when? Science 251:1058–1061.
Karl, T. R., G. Kukla, V. N. Raguvayev, M. Changeny, R. G. Quayle, R. R. Heim, Jr., and D. Fasterling. 1991. Global warming: Evidence for asymmetric diurnal trends. Geophysical Research Letters 18(12):2253–2256.
Manabe, S., K. Bryan, and M. J. Spelman. 1990. Transient response of a global ocean-atmosphere model to a doubling of atmospheric carbon dioxide. Journal of Physical Oceanography 20:722–749.
Parker, D. E. 1990. Effects of changing exposure of thermometers at land stations. In Climate Change: The IPCC Scientific Assessment, J. T. Houghton, G. J. Jenkins, and J. J. Ephraums, eds. New York: Cambridge University Press.
Plantico, M., T. R. Karl, G. Kukla, and J. Gavin. 1990. Are recent changes of temperature, cloudiness, sunshine, and precipitation across the United States related to rising levels of anthropogenic greenhouse gases? Journal of Geophysical Research 95:16617–16637.
Quayle, R. G., D. R. Easterly, T. R. Karl, and P. J. Young. 1991. Effects of recent thermometer changes in the cooperative station network. Bulletin of the American Meteorological Society 72(11):1718–1723.
Sevruk, B. 1982. Methods of Correcting for Systematic Error in Point Precipitation measurement for Operation Use. Operational Hydrology Report No. 21. Geneva: World Meteorological Organization.
Spencer, R. W., and J. R. Christy. 1990. Precise monitoring of global temperature trends from satellites. Science 247:1558–1562.
Stouffer, R. J., S. manabe, and K. Bryan. 1989. Interhemispheric asymmetry in climate responses to a gradual increase of atmospheric CO2. Nature 342:660–662.
Trenberth, K. E., and J. G. Olson. 1989. Temperature trends at the South Pole and McMurdo Sound. Journal of Climate 2:1196–1206.
Vinnikov, K. Va., P. Ya. Groisman, and K. M. Lugina. 1990. The empirical data on modern global climate changes (temperature and precipitation). Journal of Climate 3:662–677.
Washington, W. M., and G. A. Meehl. 1989. Climate sensitivity due to increased CO2: Experiments with a coupled atmosphere and ocean general circulation model. Climate Dynamics 4:1–38.
Wigley, T. M. L. 1991. Could reducing fossil-fuel emissions cause global warming? Nature 349:503–506.