High-resolution simulations of global climate, part 1: present climate

We examine simulations of today's climate performed with a global atmospheric general circulation model run at spectral truncations of T42, T170, and T239, corresponding to grid cell sizes of roughly 310 km, 75 km, and 55 km, respectively. The simulations were forced with observed sea-surface temperatures and sea-ice concentrations. The T42 simulations and initial simulations at T170 and T239 were performed using a model version that was carefully "tuned" to optimize results at T42; subsequent simulations at T170 and T239 used a model version that was partly re-tuned to improve results at T170. On the scales of a T42 grid cell and larger, nearly all quantities we examined in all the T170 and T239 simulations agree better with observations, at least in terms of spatial patterns, than in the T42 simulations. In some cases the improvements are very substantial. Improvements are seen in all-season, global domain results, and in results pertaining to most seasons and latitude bands. Increasing the model resolution from T42 introduces biases (errors in the mean) into some simulated quantities; the worst of these were removed by the partial retuning we performed at T170. This retuning has little effect on the spatial patterns of results, except in Northern Hemisphere winter at T170, where it tends to bring improvements. We discuss aspects of simulated regional climates, and their dependence on model resolution.     

The ARPEGE/IFS atmosphere model: a contribution to the French community climate modelling

A new atmospheric model has been developed jointly by Météo-France, and the European Centre for Medium-range Weather Forecasts (ECMWF) under the acronyms ARPEGE (action de recherche petite echelle grande echelle, which means research project on small and large scales) and IFS (integrated forecast system). This model includes, inter alia, an atmospheric general circulation model (GCM) which is intended by the French climate modelling community to be used for studying the anthropogenic climate impact. A preliminary version of this model has been available since 1992. This paper describes its main characteristics. Three 10-year integrations of this model having spectral horizontal resolutions of T21, T42, and T79 have been performed using prescribed monthly mean sea surface temperatures (SST) observed from 1979 until 1988. The results of these integrations are presented and compared with the observed climatology. The comparison is made for the winter (DJF) and summer (JJA) periods. It is shown that the model is capable of reproducing the observed climatology in a generally successful manner.     

High-resolution climate simulation of the last glacial maximum

The climate of the last glacial maximum (LGM) is simulated with a high-resolution atmospheric general circulation model, the NCAR CCM3 at spectral truncation of T170, corresponding to a grid cell size of roughly 75 km. The purpose of the study is to assess whether there are significant benefits from the higher resolution simulation compared to the lower resolution simulation associated with the role of topography. The LGM simulations were forced with modified CLIMAP sea ice distribution and sea surface temperatures (SST) reduced by 1°C, ice sheet topography, reduced CO₂, and 21,000 BP orbital parameters. The high-resolution model captures modern climate reasonably well, in particular the distribution of heavy precipitation in the tropical Pacific. For the ice age case, surface temperature simulated by the high-resolution model agrees better with those of proxy estimates than does the low-resolution model. Despite the fact that tropical SSTs were only 2.1°C less than the control run, there are many lowland tropical land areas 4–6°C colder than present. Comparison of T170 model results with the best constrained proxy temperature estimates (noble gas concentrations in groundwater) now yield no significant differences between model and observations. There are also significant upland temperature changes in the best resolved tropical mountain belt (the Andes). We provisionally attribute this result in part as resulting from decreased lateral mixing between ocean and land in a model with more model grid cells. A longstanding model-data discrepancy therefore appears to be resolved without invoking any unusual model physics. The response of the Asian summer monsoon can also be more clearly linked to local geography in the high-resolution model than in the low-resolution model; this distinction should enable more confident validation of climate proxy data with the high-resolution model. Elsewhere, an inferred salinity increase in the subtropical North Atlantic may have significant implications for ocean circulation changes during the LGM. A large part of the Amazon and Congo Basins are simulated to be substantially drier in the ice age—consistent with many (but not all) paleo data. These results suggest that there are considerable benefits derived from high-resolution model regarding regional climate responses, and that observationalists can now compare their results with models that resolve geography at a resolution comparable to that which the proxy data represent.     

A GCM study of Antarctic glaciation

An atmospheric general circulation model, the NCAR CCM, has been used to investigate the possible effects of two specific tectonic mechanisms on Antarctic glaciation. These two mechanisms are: (1) closing the Drake Passage (connecting South America with Antarctica), which is assumed to effectively represent an increased meridional heat transport by the ocean; and (2) changing the elevation of Antarctica. Perpetual season (summer and winter) and seasonal cycle simulations with warmer sea-surface temperatures and no sea ice prescribed for mid- to high-latitude southern oceans have been made with both present-day (high) Antarctic elevations and with low Antarctic elevations (all points 200 m). The results suggest a relatively minor role for oceanic heat transport in the formation/elimination of Antarctic glaciation. That is, under the warmer conditions inferred to have prevailed prior to the opening of the Drake Passage, conditions would still have been favorable for the maintenance of an Antarctic ice-sheet. If anything, a moderate ocean warming would promote glaciation, by increasing snowfall. Lowering the elevation of Antarctica has a larger effect on the model simulations, reducing the likelihood of glacial conditions. In the absence of snowcover, summer temperatures over Antarctica can warm considerably, leading to a monsoon-like circulation. However, it may be difficult to achieve such snow-free conditions, even with greatly increased atmospheric carbon dioxide. A tundra-like climate is the closest the model has come to representing a non-glacial climate, even when both seasurface temperatures and elevations are maximally varied.     

Simulated Antarctic precipitation and surface mass balance at the end of the twentieth and twenty-first centuries

The aim of this work is to assess potential future Antarctic surface mass balance changes, the underlying mechanisms, and the impact of these changes on global sea level. To this end, this paper presents simulations of the Antarctic climate for the end of the twentieth and twenty-first centuries. The simulations were carried out with a stretched-grid atmospheric general circulation model, allowing for high horizontal resolution (60 km) over Antarctica. It is found that the simulated present-day surface mass balance is skilful on continental scales. Errors on regional scales are moderate when observed sea surface conditions are used; more significant regional biases appear when sea surface conditions from a coupled model run are prescribed. The simulated Antarctic surface mass balance increases by 32 mm water equivalent per year in the next century, corresponding to a sea level decrease of 1.2 mm year⁻¹ by the end of the twenty-first century. This surface mass balance increase is largely due to precipitation changes, while changes in snow melt and turbulent latent surface fluxes are weak. The temperature increase leads to an increased moisture transport towards the interior of the continent because of the higher moisture holding capacity of warmer air, but changes in atmospheric dynamics, in particular off the Antarctic coast, regionally modulate this signal.     

Simulations of the Last Glacial Maximum climates using a general circulation model: prescribed versus computed sea surface temperatures

 The climate during the Last Glacial Maximum (LGM) has been simulated using the UK Universities Global Atmospheric Modelling Programme (UGAMP) general circulation model (GCM) with both prescribed sea surface temperatures (SSTs) based on the CLIMAP reconstruction and computed SSTs with a simple thermodynamic slab ocean. Consistent with the Paleoclimate Modelling Intercomparison Project (PMIP), the other boundary conditions include the large changes in ice-sheet topography and geography, a lower sea level, a lower concentration of CO₂ in the atmosphere, and a slightly different insolation pattern at the top of the atmosphere. The results are analysed in terms of changes in atmospheric circulation. Emphasis is given to the changes in surface temperatures, planetary waves, storm tracks and the associated changes in distribution of precipitation. The model responds in a similar manner to the changes in boundary conditions to previous studies in global mean statistics, but differs in its treatment of regional climates. Results also suggest that both the land ice sheets and sea ice introduce significant changes in planetary waves and transient eddy activity, which in turn affect regional climates. The computed SST simulations predict less sea ice and cooler tropical temperatures than those based on CLIMAP SSTs. It is unclear as to whether this is a model and/or a data problem, but the resulting changes in land temperatures and precipitation can be large. Snow mass budget analysis suggests that there is net ice loss along the southern edges of the Laurentide and Fennoscandian ice sheets and net ice gain over other parts of the two ice sheets. The net accumulation is mainly due to the decrease in ablation in the cold climate rather than to the changes in snowfall. The characteristics of the Greenland ice-sheet mass balance in the LGM simulations is also quite different from those in the present-day (PD) simulations. The ablation in the LGM simulations is negligible while it is a very important process in the ice mass budget in the PD simulations. Received: 10 January 1997 / Accepted: 11 December 1997     

A time-slice experiment with the ECHAM4 AGCM at high resolution: the impact of horizontal resolution on annual mean climate change

 The climate response to increasing levels of atmospheric greenhouse gases, prescribed according to the International Panel of Climate Change (IPCC) scenario IS92a, is studied in two model simulations. The reference simulation is a transient response experiment performed with a medium-resolution (T42) coupled general circulation model of the atmosphere and ocean (ECHAM4/OPYC) developed at the Max-Planck-Institute for Meteorology. For two 30-year “time slices”, representing the present-day climate and the future climate at the time of effective CO₂ doubling, the annual mean climate states are compared with those obtained from the high-resolution (T106) ECHAM4 model forced with monthly sea surface temperatures and sea-ice from the coupled model. The large-scale changes in temperature, zonal wind, sea-level pressure and precipitation are broadly similar. This applies, in particular, to the respective zonal means. In general, except for precipitation, the responses in the time-slice experiments are slightly weaker than those simulated in the coupled model due to a smaller effect of the horizontal resolution on the simulations of the future (warmer) period than on the simulations of the present period. On a regional scale, the impact of horizontal resolution is smaller in the Southern than in the Northern Hemisphere, where the response differences are caused mainly by changes in the positions of the stationary waves. Although the precipitation responses are broadly similar, there are few notable exceptions such as a more pronounced maximum over the equatorial oceans in the T106 experiment but a weaker response over low-latitude land areas. Differences in precipitation response are found especially in areas with strong topographical control such as South America, for example. Received: 17 January 2000 / Accepted: 7 July 2000     

A note on horizontal resolution dependence for monsoon rainfall simulations

Summary The motivation for this study came from recent results of an Atmospheric Model Inter-comparison Project (AMIP) coordinated by the Lawrence Livermore Laboratory at Livermore, California. That project included a review of seasonal monsoon simulations from 13 different atmospheric models over the world. Most of the models used a horizontal resolution of roughly 300 km. The seasonal monsoon simulations from these models varied significantly. The poor performance by these models stems in part from the use of the coarse resolution. The purpose of this note is to show that by using the same model physics and lower boundary conditions, such as snow/ice cover and sea surface temperatures, the use of the higher horizontal resolution does have a stronger positive impact on the skill of monthly rainfall when compared to a lower horizontal resolution. In this note we present the results of such a comparison between the horizontal resolutions of T42 and T170. These studies are carried out for the prescribed lower boundary specification of sea surface temperatures and snow/ice cover with the help of an Atmospheric General Circulation Model. Received August 16, 1999 Revised October 14, 1999     

The coherent large-scale circulation change between dry/wet years over central eastern China simulated by NCAR CAM5

This study evaluates the simulation of the coherent circulation structure correspond to the changes of mid-summer (July–August) rainfall over eastern China (30°–40° N, 110°–120° E) in high-resolution NCAR CAM5. Forced by historical sea surface temperatures (SSTs), the NCAR CAM5 reasonably reproduces coherent changes of temperature and large-scale circulations, corresponding to the changes in rainfall. Results show that when the rainfall decreases over eastern China, the model reproduces a remarkable warm center in the upper troposphere with an anomalous anticyclone appears above and an increase in anomalous westerlies to its north. An anomalous anticyclone also occurs in the lower troposphere, along with anomalous southerlies to its east which indicates strengthening of the East Asian summer monsoon. Both the circulation changes in the upper and lower troposphere favor a decrease in precipitation over central eastern China. There were also good correlations between the simulated upper-tropospheric temperature and other large-scale circulation changes. There are some deficiencies in the NCAR CAM5 simulations in terms of the changes in magnitude and location of the rainfall centers. However, in general, the model reasonably reproduced the coherent configuration of the large-scale circulation patterns and surface rainfall. This study further confirms that the climate variations across East Asia most likely arise from a regional response to global climate change. The well-simulated configuration by NCAR CAM5 also indicates the reliability of the model and its potential to reveal the mechanisms driving the coherent changes of the East Asian summer monsoon system.     

Past and future polar amplification of climate change: climate model intercomparisons and ice-core constraints

Climate model simulations available from the PMIP1, PMIP2 and CMIP (IPCC-AR4) intercomparison projects for past and future climate change simulations are examined in terms of polar temperature changes in comparison to global temperature changes and with respect to pre-industrial reference simulations. For the mid-Holocene (MH, 6,000 years ago), the models are forced by changes in the Earth’s orbital parameters. The MH PMIP1 atmosphere-only simulations conducted with sea surface temperatures fixed to modern conditions show no MH consistent response for the poles, whereas the new PMIP2 coupled atmosphere–ocean climate models systematically simulate a significant MH warming both for Greenland (but smaller than ice-core based estimates) and Antarctica (consistent with the range of ice-core based range). In both PMIP1 and PMIP2, the MH annual mean changes in global temperature are negligible, consistent with the MH orbital forcing. The simulated last glacial maximum (LGM, 21,000 years ago) to pre-industrial change in global mean temperature ranges between 3 and 7°C in PMIP1 and PMIP2 model runs, similar to the range of temperature change expected from a quadrupling of atmospheric CO₂ concentrations in the CMIP simulations. Both LGM and future climate simulations are associated with a polar amplification of climate change. The range of glacial polar amplification in Greenland is strongly dependent on the ice sheet elevation changes prescribed to the climate models. All PMIP2 simulations systematically underestimate the reconstructed glacial–interglacial Greenland temperature change, while some of the simulations do capture the reconstructed glacial–interglacial Antarctic temperature change. Uncertainties in the prescribed central ice cap elevation cannot account for the temperature change underestimation by climate models. The variety of climate model sensitivities enables the exploration of the relative changes in polar temperature with respect to changes in global temperatures. Simulated changes of polar temperatures are strongly related to changes in simulated global temperatures for both future and LGM climates, confirming that ice-core-based reconstructions provide quantitative insights on global climate changes. An erratum to this article can be found at     

大气环流的年代际变化 II.GCM数值模拟研究

类似大气环流模式比较计划（AMIP）的数值模拟,将实际观测的海表水温（SST）资料引入模式进行40多年的数值积分,得到长时间的大气环流模拟结果。分析数值模拟结果发现,无论是大气中的主要涛动和遥相关型,还是重要大气环流系统都极为清楚地存在着年代际变化特征,包括10～20年准周期振荡和可能的30年以上的准周期振荡;而且上述主要环流系统的形势及其年代际变化大都与实际观测资料所给出的结果相一致。顺便分析中国东部气候的模拟结果,年代际变化特征（包括60年代的气候突变）也很清楚,并同大气环流变化配合十分合理。结果也表明,同研究季节和年际变化一样,大气环流模式（AGCM）数值模拟也是研究大气环流和气候年代际变化的有效方法。     

The global response to Younger Dryas boundary conditions in an AGCM simulation

 Geological evidence points to a global Younger Dryas (YD) climatic oscillation during the last glacial/ present interglacial transition phase. A convincing mechanism to explain this global YD climatic oscillation is not yet available. Nevertheless, a profound understanding of the mechanism behind the YD climate would lead to a better understanding of climate variability. Therefore, the Hamburg atmospheric circulation model was used to perform four numerical experiments on the YD climate. The objective of this study is to improve the understanding of different forcings influencing climate during the last glacial/interglacial transition and to investigate to what extent the model response agrees with global geological evidence of YD climate change. The following boundary conditions were altered: sea surface conditions, ice sheets, insolation and atmospheric CO₂ concentration. Sea surface temperatures based on foraminiferal assemblages proved to produce insufficient winter cooling in the N Atlantic Ocean in two experiments. It is proposed that this discrepancy is caused by uncertainties in the reconstruction method of sea surface temperatures. Therefore, a model-derived set of Atlantic surface ocean conditions was prescribed in a subsequent simulation. However, the latter set represented an Atlantic Ocean without a thermohaline circulation, which is not in agreement with evidence from ocean cores. The global response to the boundary conditions was analysed using three variables, namely surface temperature, zonal wind speed and precipitation. The statistical significance of the changes was tested with a two-tailed t-test. Moreover, the significant responses to cooled oceans were compared with geological evidence of a YD oscillation. This comparison revealed a good match in Europe, Greenland, Atlantic Canada and the N Pacific region, explaining the YD oscillation in these regions as a response to cooled N Atlantic and N Pacific Oceans. However, the results leave the YD climate in other regions completely unexplained. This reflects either an insufficient set of boundary conditions or the important role played by feedbacks within the coupled atmosphere-ocean-ice system. These feedbacks are poorly represented in the used atmospheric model, since ice sheets and the ocean surface conditions have to be prescribed. Received: 30 July 1996 / Accepted: 12 February 1997     

Regional climate model projections for the State of Washington

Global climate models do not have sufficient spatial resolution to represent the atmospheric and land surface processes that determine the unique regional climate of the State of Washington. Regional climate models explicitly simulate the interactions between the large-scale weather patterns simulated by a global model and the local terrain. We have performed two 100-year regional climate simulations using the Weather Research and Forecasting (WRF) model developed at the National Center for Atmospheric Research (NCAR). One simulation is forced by the NCAR Community Climate System Model version 3 (CCSM3) and the second is forced by a simulation of the Max Plank Institute, Hamburg, global model (ECHAM5). The mesoscale simulations produce regional changes in snow cover, cloudiness, and circulation patterns associated with interactions between the large-scale climate change and the regional topography and land-water contrasts. These changes substantially alter the temperature and precipitation trends over the region relative to the global model result or statistical downscaling. To illustrate this effect, we analyze the changes from the current climate (1970–1999) to the mid twenty-first century (2030–2059). Changes in seasonal-mean temperature, precipitation, and snowpack are presented. Several climatological indices of extreme daily weather are also presented: precipitation intensity, fraction of precipitation occurring in extreme daily events, heat wave frequency, growing season length, and frequency of warm nights. Despite somewhat different changes in seasonal precipitation and temperature from the two regional simulations, consistent results for changes in snowpack and extreme precipitation are found in both simulations.     

Assessing the Impacts of Eurasian Snow Conditions on Climate Predictability with a Global Climate Model

On the basis of two ensemble experiments conducted by a general atmospheric circulation model (Institute of Atmospheric Physics nine-level atmospheric general circulation model coupled with land surface model, hereinafter referred to as IAP9L_CoLM), the impacts of realistic Eurasian snow conditions on summer climate predictability were investigated. The predictive skill of sea level pressures (SLP) and middle and upper tropospheric geopotential heights at mid-high latitudes of Eurasia was enhanced when improved Eurasian snow conditions were introduced into the model. Furthermore, the model skill in reproducing the interannual variation and spatial distribution of the surface air temperature (SAT) anomalies over China was improved by applying realistic (prescribed) Eurasian snow conditions. The predictive skill of the summer precipitation in China was low; however, when realistic snow conditions were employed, the predictability increased, illustrating the effectiveness of the application of realistic Eurasian snow conditions. Overall, the results of the present study suggested that Eurasian snow conditions have a significant effect on dynamical seasonal prediction in China. When Eurasian snow conditions in the global climate model (GCM) can be more realistically represented, the predictability of summer climate over China increases.     

Sensitivity of transient eddies to climate change in the CCC general circulation model

Abstract

We use eddy life‐cycle simulations to evaluate the response of atmospheric transient eddies to a global warming caused by CO₂ doubling in the CCC general circulation model. In simulations using Northern Hemisphere winter conditions, transient waves attain larger kinetic energy and encompass a wider range of latitudes in the warmer climate. This behaviour contrasts with a previous investigation that used output from the NCAR and GFDL models. Our analysis indicates two primary factors for the difference between model responses: (1) a smaller change in the mid‐latitude temperature gradient in the CCC model, which allows (2) increased atmospheric water vapour in mid‐latitudes to catalyze a more rapidly evolving life‐cycle.     

Mid-Pliocene westerlies from PlioMIP simulations

The midlatitude westerlies are one of the major components of the global atmospheric circulation. They play an important role in midlatitude weather and climate, and are particularly significant in interpreting aeolian sediments. In this study, we analyzed the behavior and the possible mechanism involved in the change of the westerlies, mainly in terms of the jet stream position, in the mid-Pliocene warm period(3.3 to 3.0 million years ago) using simulations of 15 climate models from the Pliocene Model Intercomparison Project(Plio MIP). Compared to the reference period, the mid-Pliocene midlatitude westerlies generally shifted poleward(approximately 3.6 of latitude in the Northern Hemisphere and 1.9 of latitude in the Southern Hemisphere at 850 h Pa level) with a dipole pattern. The dipole pattern of the tropospheric zonal wind anomalies was closely related to the change of the tropospheric meridional temperature gradient as a result of thermal structure adjustment.The poleward shift of the midlatitude westerly jet corresponded to the poleward shift of the mean meridional circulation.The sea surface temperatures and sea ice may have affected the simulated temperature structure and zonal winds, causing the spread of the westerly anomalies in the mid-Pliocene between the atmosphere-only and coupled atmosphere–ocean general circulation model simulations.     

Dependence of climate forcing and response on the altitude of black carbon aerosols

Black carbon aerosols absorb solar radiation and decrease planetary albedo, and thus can contribute to climate warming. In this paper, the dependence of equilibrium climate response on the altitude of black carbon is explored using an atmospheric general circulation model coupled to a mixed layer ocean model. The simulations model aerosol direct and semi-direct effects, but not indirect effects. Aerosol concentrations are prescribed and not interactive. It is shown that climate response of black carbon is highly dependent on the altitude of the aerosol. As the altitude of black carbon increases, surface temperatures decrease; black carbon near the surface causes surface warming, whereas black carbon near the tropopause and in the stratosphere causes surface cooling. This cooling occurs despite increasing planetary absorption of sunlight (i.e. decreasing planetary albedo). We find that the trend in surface air temperature response versus the altitude of black carbon is consistent with our calculations of radiative forcing after the troposphere, stratosphere, and land surface have undergone rapid adjustment, calculated as “regressed” radiative forcing. The variation in climate response from black carbon at different altitudes occurs largely from different fast climate responses; temperature dependent feedbacks are not statistically distinguishable. Impacts of black carbon at various altitudes on the hydrological cycle are also discussed; black carbon in the lowest atmospheric layer increases precipitation despite reductions in solar radiation reaching the surface, whereas black carbon at higher altitudes decreases precipitation.     

An examination of internally generated variability in long climate simulations

General circulation model experiments designed to estimate the magnitude and structure of internally generated variability and to help understand the mechanisms underlying this variability are described. The experiments consist of three multi-century integrations of a rhomboidal 15, 9 level, version of the Center for Ocean-Land-Atmosphere Studies atmospheric general circulation model: a run with fixed sea surface temperatures and equinox solar radiation, a run with seasonally varying climatological sea surface temperatures and seasonally varying solar forcing, and a run with seasonally varying solar forcing in which the state of the ocean is predicted by a 3° by 3°, 16 vertical level, nearly global domain version of the Geophysical Fluid Dynamics Laboratory Modular Ocean Model. No flux correction is used in the coupled model integration. Selected surface fields of the three runs are compared to each other as well as to the observed climate. Statistical properties of variability on interannual time scales are compared between the runs. Evidence is presented that climate time scale variability in the simulations is produced by random weather time scale forcing due to the integrating effect of elements of the system with long memories. The importance of ocean variability for land climate variability is demonstrated and attributed to both the memory effect and coupled atmosphere-ocean instability.     

European blocking and Atlantic jet stream variability in the NCEP/NCAR reanalysis and the CMCC-CMS climate model

The relationship between atmospheric blocking over Europe and the Atlantic eddy-driven jet stream is investigated in the NCEP/NCAR Reanalysis and in a climate model. This is carried out using a bidimensional blocking index based on geopotential height and a diagnostic providing daily latitudinal position and strength of the jet stream. It is shown that European Blocking (EB) is not decoupled from the jet stream but it is mainly associated with its poleward displacements. Moreover, the whole blocking area placed on the equatorward side of the jet stream, broadly ranging from Azores up to Scandinavia, emerges as associated with poleward jet displacements. The diagnostics are hence applied to two different climate model simulations in order to evaluate the biases in the jet stream and in the blocking representation. This analysis highlights large underestimation of EB, typical feature of general circulation models. Interestingly, observed blocking and jet biases over the Euro-Atlantic area are consistent with the blocking-jet relationship observed in the NCEP/NCAR Reanalysis. Finally, the importance of sea surface temperatures (SSTs) is investigated showing that realistic SSTs can reduce the bias in the jet stream variability but not in the frequency of EB. We conclude highlighting that blocking-related diagnostics can provide more information about the Euro-Atlantic variability than diagnostics simply based on the Atlantic jet stream.     

Detection of anthropogenic climate change using an atmospheric GCM

 Atmosphere-only general circulation models are shown to be a useful tool for detecting an anthropogenic effect on climate and understanding recent climate change. Ensembles of atmospheric runs are all forced with the same observed changes in sea surface temperatures and sea-ice extents but differ in terms of the combinations of anthropogenic effects included. Therefore, our approach aims to detect the `immediate' anthropogenic impact on the atmosphere as opposed to that which has arisen via oceanic feedbacks. We have adapted two well-used detection techniques, pattern correlations and fingerprints, and both show that near-decadal changes in the patterns of zonal mean upper air temperature are well simulated, and that it is highly unlikely that the observed changes could be accounted for by sea surface temperature variations and internal variability alone. Furthermore, we show that for zonally averaged upper air temperature, internal `noise' in the atmospheric model is small enough that a signal emerges from the data even on interannual time scales; this would not be possible in a coupled ocean-atmosphere general circulation model. Finally, although anthropogenic forcings have had a significant impact on global mean land surface temperature, we find that their influence on the pattern of local deviations about this mean is so far undetectable. In order to achieve this in the future, as the signal grows, it will also be important that the response of the Northern Hemisphere mid-latitude westerly flow to changing sea surface temperatures is well simulated in climate model detection studies. Received: 3 December 1999 / Accepted: 30 October 2000