IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis

3.1 Introduction

This chapter assesses the observed changes in surface and atmospheric climate, placing new observations and new analyses made during the past six years (since the Third Assessment Report – TAR) in the context of the previous instrumental record. In previous IPCC reports, palaeo-observations from proxy data for the pre-instrumental past and observations from the ocean and ice domains were included within the same chapter. This helped the overall assessment of the consistency among the various variables and their synthesis into a coherent picture of change. However, the amount of information became unwieldy and is now spread over Chapters 3 to 6. Nevertheless, a short synthesis and scrutiny of the consistency of all the observations is included here (see Section 3.9).

In the TAR, surface temperature trends were examined from 1860 to 2000 globally, for 1901 to 2000 as maps and for three sub-periods (1910–1945, 1946–1975 and 1976–2000). The first and third sub-periods had rising temperatures, while the second sub-period had relatively stable global mean temperatures. The 1976 divide is the date of a widely acknowledged ‘climate shift’ (e.g., Trenberth, 1990) and seems to mark a time (see Chapter 9) when global mean temperatures began a discernible upward trend that has been at least partly attributed to increases in greenhouse gas concentrations in the atmosphere (see the TAR; IPCC, 2001). The picture prior to 1976 has essentially not changed and is therefore not repeated in detail here. However, it is more convenient to document the sub-period after 1979, rather than 1976, owing to the availability of increased and improved satellite data since then (in particular Television InfraRed Observation Satellite (TIROS) Operational Vertical Sounder (TOVS) data) in association with the Global Weather Experiment (GWE) of 1979. The post-1979 period allows, for the first time, a global perspective on many fields of variables, such as precipitation, that was not previously available. For instance, the reanalyses of the global atmosphere from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR, referred to as NRA; Kalnay et al., 1996; Kistler et al., 2001) and the European Centre for Medium Range Weather Forecasts (ECMWF, referred to as ERA-40; Uppala et al., 2005) are markedly more reliable after 1979, and spurious discontinuities are present in the analysed record at the end of 1978 (Santer et al., 1999; Bengtsson et al., 2004; Bromwich and Fogt, 2004; Simmons et al., 2004; Trenberth et al., 2005a). Therefore, the availability of high-quality data has led to a focus on the post-1978 period, although physically this new regime seems to have begun in 1976/1977.

Documentation of the climate has traditionally analysed global and hemispheric means, and land and ocean means, and has presented some maps of trends. However, climate varies over all spatial and temporal scales: from the diurnal cycle to El Niño to multi-decadal and millennial variations. Atmospheric waves naturally create regions of temperature and moisture of opposite-signed departures from the zonal mean, as moist warm conditions are favoured in poleward flow while cool dry conditions occur in equatorward flow. Although there is an infinite variety of weather systems, one area of recent substantial progress is recognition that a few preferred patterns (or modes) of variability determine the main seasonal and longer-term climate anomalies (Section 3.6). These patterns arise from the differential effects on the atmosphere of land and ocean, mountains, and anomalous heating, such as occurs during El Niño events. The response is generally felt in regions far removed from the anomalous forcing through atmospheric teleconnections, associated with large-scale waves in the atmosphere. This chapter therefore documents some aspects of temperature and precipitation anomalies associated with these preferred patterns, as they are vitally important for understanding regional climate anomalies (such as observed cooling in parts of the northern North Atlantic from 1901 to 2005; see Section, Figure 3.9) and why they differ from global means. Changes in storm tracks, the jet streams, regions of preferred blocking anticyclones and changes in monsoons all occur in conjunction with these preferred patterns and other climate anomalies. Therefore the chapter not only documents changes in variables, but also changes in phenomena (such as El Niño) or patterns, in order to increase understanding of the character of change.

Extremes of climate, such as droughts and wet spells, are very important because of their large impacts on society and the environment, but they are an expression of the variability. Therefore, the nature of variability at different spatial and temporal scales is vital to our understanding of extremes. The global means of temperature and precipitation are most readily linked to global mean radiative forcing and are important because they clearly indicate if unusual change is occurring. However, the local or regional response can be complex and perhaps even counter-intuitive, such as changes in planetary waves in the atmosphere induced by global warming that result in regional cooling. As an indication of the complexity associated with temporal and spatial scales, Table 3.1 provides measures of the magnitude of natural variability of surface temperature in which climate signals are embedded. The measures used are indicators of the range: the mean range of the diurnal and annual cycles, and the estimated 5th to 95th percentiles range of anomalies. These are based on the standard deviation and assumed normal distribution, which is a reasonable approximation in many places for temperature, with the exception of continental interiors in the cold season, which have strongly negatively skewed temperature distributions owing to cold extremes. For the global mean, the variance is somewhat affected by the observed trend, which inflates this estimate of the range slightly. The comparison highlights the large diurnal cycle and daily variability. Daily variability is, however, greatly reduced by either spatial or temporal averaging that effectively averages over synoptic weather systems. Nevertheless, even continental-scale averages contain much greater variability than the global mean in association with planetary-scale waves and events such as El Niño.

Table 3.1. Typical ranges of surface temperature at different spatial and temporal scales for a sample mid-latitude mid-continental station (Boulder, Colorado; based on 80 years of data) and for monthly mean anomalies (diurnal and annual cycles removed) for the USA as a whole and the globe for the 20th century. For the diurnal and annual cycles, the monthly mean range is given, while other values are the difference between the 5th and 95th percentiles.

Temporal and Spatial Scale  Temperature Range (°C) 
Boulder diurnal cycle  

13.1 (December) to

15.1 (September)

Boulder annual cycle  23 
Boulder daily anomalies  15 
Boulder monthly anomalies  7.0 
USA monthly anomalies  3.9 
Global mean monthly anomalies  0.79 

Throughout the chapter, the authors try to consistently indicate the degree of confidence and uncertainty in trends and other results, as given by Box TS.1 in the Technical Summary. Quantitative estimates of uncertainty include: for the mean, the 5th and 95th percentiles; and for trends, statistical significance at the 0.05 (5%) significance level. This allows assessment of what is unusual. The chapter mainly uses the word ‘trend’ to designate a generally monotonic change in the level of a variable. Where numerical values are given, they are equivalent linear trends, though more complex changes in the variable will often be clear from the description. The chapter also assesses, if possible, the physical consistency among different variables, which helps to provide additional confidence in trends.