Will the tropical land biosphere dominate the climate–carbon cycle feedback during the twenty-first century?

Global warming caused by anthropogenic CO₂ emissions is expected to reduce the capability of the ocean and the land biosphere to take up carbon. This will enlarge the fraction of the CO₂ emissions remaining in the atmosphere, which in turn will reinforce future climate change. Recent model studies agree in the existence of such a positive climate–carbon cycle feedback, but the estimates of its amplitude differ by an order of magnitude, which considerably increases the uncertainty in future climate projections. Therefore we discuss, in how far a particular process or component of the carbon cycle can be identified, that potentially contributes most to the positive feedback. The discussion is based on simulations with a carbon cycle model, which is embedded in the atmosphere/ocean general circulation model ECHAM5/MPI-OM. Two simulations covering the period 1860–2100 are conducted to determine the impact of global warming on the carbon cycle. Forced by historical and future carbon dioxide emissions (following the scenario A2 of the Intergovernmental Panel on Climate Change), they reveal a noticeable positive climate–carbon cycle feedback, which is mainly driven by the tropical land biosphere. The oceans contribute much less to the positive feedback and the temperate/boreal terrestrial biosphere induces a minor negative feedback. The contrasting behavior of the tropical and temperate/boreal land biosphere is mostly attributed to opposite trends in their net primary productivity (NPP) under global warming conditions. As these findings depend on the model employed they are compared with results derived from other climate–carbon cycle models, which participated in the Coupled Climate–Carbon Cycle Model Intercomparison Project (C4MIP).

Influence of dynamic vegetation on climate change arising from increasing CO<Subscript>2</Subscript>

A dynamic global vegetation model (DGVM) is coupled to an atmospheric general circulation model (AGCM) to investigate the influence of vegetation dynamics on climate change under conditions of global warming. The model results are largely in agreement with observations and the results of previous studies in terms of the present climate, present potential vegetation, present net primary productivity (NPP), and pre-industrial carbon budgets. The equilibrium state of climate properties are compared among pre-industrial, doubled, and quadrupled atmospheric CO₂ values using DGVM–AGCM and current AGCM with fixed vegetation to evaluate the influence of dynamic vegetation change. We also separated the contributions of temperature, precipitation and CO₂ fertilization on vegetation change. The results reveal an amplification of global warming climate sensitivity by 10% due to the inclusion of dynamic vegetation. The total effects of elevated CO₂ and climate change also lead to an increase in NPP and vegetation coverage globally. The reduction of albedo associated with this greening results in enhanced global warming. Our separation analysis indicates that temperature alters vegetation at high latitudes such as Siberia or Alaska, where there is a switch from tundra to forest. On the other hand, CO₂ fertilization provides the largest contribution to greening in arid/semi-arid region. Precipitation change did not cause any drastic vegetation shift.     

Past and future changes in climate and hydrological indicators in the US Northeast

To assess the influence of global climate change at the regional scale, we examine past and future changes in key climate, hydrological, and biophysical indicators across the US Northeast (NE). We first consider the extent to which simulations of twentieth century climate from nine atmosphere-ocean general circulation models (AOGCMs) are able to reproduce observed changes in these indicators. We then evaluate projected future trends in primary climate characteristics and indicators of change, including seasonal temperatures, rainfall and drought, snow cover, soil moisture, streamflow, and changes in biometeorological indicators that depend on threshold or accumulated temperatures such as growing season, frost days, and Spring Indices (SI). Changes in indicators for which temperature-related signals have already been observed (seasonal warming patterns, advances in high-spring streamflow, decreases in snow depth, extended growing seasons, earlier bloom dates) are generally reproduced by past model simulations and are projected to continue in the future. Other indicators for which trends have not yet been observed also show projected future changes consistent with a warmer climate (shrinking snow cover, more frequent droughts, and extended low-flow periods in summer). The magnitude of temperature-driven trends in the future are generally projected to be higher under the Special Report on Emission Scenarios (SRES) mid-high (A2) and higher (A1FI) emissions scenarios than under the lower (B1) scenario. These results provide confidence regarding the direction of many regional climate trends, and highlight the fundamental role of future emissions in determining the potential magnitude of changes we can expect over the coming century.

---

The greening of the McGill Paleoclimate Model. Part I: Improved land surface scheme with vegetation dynamics

The formulation of a new land surface scheme (LSS) with vegetation dynamics for coupling to the McGill Paleoclimate Model (MPM) is presented. This LSS has the following notable improvements over the old version: (1) parameterization of deciduous and evergreen trees by using the models climatology and the output of the dynamic global vegetation model, VECODE (Brovkin et al. in Ecological Modelling 101:251–261 (1997), Global Biogeochemical Cycles 16(4):1139, (2002)); (2) parameterization of tree leaf budburst and leaf drop by using the models climatology; (3) parameterization of the seasonal cycle of the grass leaf area index; (4) parameterization of the seasonal cycle of tree leaf area index by using the time-dependent growth of the leaves; (5) calculation of land surface albedo by using vegetation-related parameters, snow depth and the models climatology. The results show considerable improvement of the models simulation of the present-day climate as compared with that simulated in the original physically-based MPM. In particular, the strong seasonality of terrestrial vegetation and the associated land surface albedo variations are in good agreement with several satellite observations of these quantities. The application of this new version of the MPM (the green MPM) to Holocene millennial-scale climate changes is described in a companion paper, Part II.

---

Impact of global warming on the geobotanic zones: an experiment with a statistical�Cdynamical climate model

In this study, a zonally-averaged statistical climate model (SDM) is used to investigate the impact of global warming on the distribution of the geobotanic zones over the globe. The model includes a parameterization of the biogeophysical feedback mechanism that links the state of surface to the atmosphere (a bidirectional interaction between vegetation and climate). In the control experiment (simulation of the present-day climate) the geobotanic state is well simulated by the model, so that the distribution of the geobotanic zones over the globe shows a very good agreement with the observed ones. The impact of global warming on the distribution of the geobotanic zones is investigated considering the increase of CO₂ concentration for the B1, A2 and A1FI scenarios. The results showed that the geobotanic zones over the entire earth can be modified in future due to global warming. Expansion of subtropical desert and semi-desert zones in the Northern and Southern Hemispheres, retreat of glaciers and sea-ice, with the Arctic region being particularly affected and a reduction of the tropical rainforest and boreal forest can occur due to the increase of the greenhouse gases concentration. The effects were more pronounced in the A1FI and A2 scenarios compared with the B1 scenario. The SDM results confirm IPCC AR4 projections of future climate and are consistent with simulations of more complex GCMs, reinforcing the necessity of the mitigation of climate change associated to global warming.     

A Projection of the Effects of the Climate Change Induced by Increased CO2 on Extreme Hydrologic Events in the Western U.S.

The effects of increased atmospheric CO₂ on the frequency of extreme hydrologic events in the Western United States (WUS) for the 10-yr period of 2040–2049 are examined using dynamically downscaled regional climate change signals. For assessing the changes in the occurrence of hydrologic extremes, downscaled climate change signals in daily precipitation and runoff that are likely to indicate the occurrence of extreme events are examined. Downscaled climate change signals in the selected indicators suggest that the global warming induced by increased CO₂ is likely to increase extreme hydrologic events in the WUS. The indicators for heavy precipitation events show largest increases in the mountainous regions of the northern California Coastal Range and the Sierra Nevada. Increased cold season precipitation and increased rainfall-portion of precipitation at the expense of snowfall in the projected warmer climate result in large increases in high runoff events in the Sierra Nevada river basins that are already prone to cold season flooding in todays climate. The projected changes in the hydrologic characteristics in the WUS are mainly associated with higher freezing levels in the warmer climate and increases in the cold season water vapor influx from the Pacific Ocean.     

Transient simulation of the last glacial inception. Part I: glacial inception as a bifurcation in the climate system

We study the mechanisms of glacial inception by using the Earth system model of intermediate complexity, CLIMBER-2, which encompasses dynamic modules of the atmosphere, ocean, biosphere and ice sheets. Ice-sheet dynamics are described by the three-dimensional polythermal ice-sheet model SICOPOLIS. We have performed transient experiments starting at the Eemiam interglacial, at 126 ky BP (126,000 years before present). The model runs for 26 kyr with time-dependent orbital and CO₂ forcings. The model simulates a rapid expansion of the area covered by inland ice in the Northern Hemisphere, predominantly over Northern America, starting at about 117 kyr BP. During the next 7 kyr, the ice volume grows gradually in the model at a rate which corresponds to a change in sea level of 10 m per millennium. We have shown that the simulated glacial inception represents a bifurcation transition in the climate system from an interglacial to a glacial state caused by the strong snow-albedo feedback. This transition occurs when summer insolation at high latitudes of the Northern Hemisphere drops below a threshold value, which is only slightly lower than modern summer insolation. By performing long-term equilibrium runs, we find that for the present-day orbital parameters at least two different equilibrium states of the climate system exist—the glacial and the interglacial; however, for the low summer insolation corresponding to 115 kyr BP, we find only one, glacial, equilibrium state, while for the high summer insolation corresponding to 126 kyr BP only an interglacial state exists in the model.

---

Decadal interactions between the western tropical Pacific and the North Atlantic Oscillation

The relationship between interdecadal variations of tropical sea surface temperature (SST) in the last 120 years and circulation anomalies related to the North Atlantic Oscillation (NAO) is investigated in this study. Using an atmospheric general circulation model (AGCM), we confirm observational evidence that variations in the SST gradient in the western tropical Pacific are related to the NAO anomalies on decadal timescale, and may be contributing to the shift towards the positive NAO phase observed in the late 20th century. The role played by the Indian Ocean-NAO teleconnection, advocated in recent studies focused on the last 50 years, is also assessed in the context of the 120-year long record. It is suggested that a positive feedback between the Pacific SST and the hemispheric circulation pattern embedding the decadal NAO signal may act to enhance the internal variability of the coupled ocean–atmosphere system, and justify the stronger teleconnection found in observational data than in SST-forced AGCM experiments.

---

Sensitivity of global warming to the pattern of tropical ocean warming

The current generations of climate models are in substantial disagreement as to the projected patterns of sea surface temperatures (SSTs) in the Tropics over the next several decades. We show that the spatial patterns of tropical ocean temperature trends have a strong influence on global mean temperature and precipitation and on global mean radiative forcing. We identify the SST patterns with the greatest influence on the global mean climate and find very different, and often opposing, sensitivities to SST changes in the tropical Indian and West Pacific Oceans. Our work stresses the need to reduce climate model biases in these sensitive regions, as they not only affect the regional climates of the nearby densely populated continents, but also have a disproportionately large effect on the global climate.

---

Future changes in cyclone climatology over Europe as inferred from a regional climate simulation

This study analyzes the cyclone climatology in regional climate model simulations of present day (1961–1990) and future (2071–2100, A2 and B2 emission scenarios) european climate conditions. The model domain covers the area from Scandinavia to Northern Africa and from the Eastern Atlantic to Russia at a horizontal grid spacing of 50 km. Compared to present day, in the A2 and B2 scenario conditions the annual average storm track intensity increases over the North-East Atlantic and decreases over Russia and the Eastern Mediterranean region. This overall change pattern is larger in the A2 than in the B2 simulations. However, the cyclone climatology change signal shows a large intermonthly variability and important differences across European regions. The largest changes are found over the North-East Atlantic, where the storm track intensity increases in winter and decreases in summer. A significant reduction of storm track intensity is found during late summer and autumn over the Mediterranean region, and from October to January over Russia. The number of cyclones decreases in future conditions throughout Europe, except over the Central Europe and Mediterranean regions in summer (where it increases). The frequency of intense cyclones and the depth of extreme cyclones increase over the North-East Atlantic, decrease over Russia and show an irregular response over the rest of the domain.

---

The impact of natural and anthropogenic forcings on climate and hydrology since 1550

A climate simulation of an ocean/atmosphere general circulation model driven with natural forcings alone (constant “pre-industrial” land-cover and well-mixed greenhouse gases, changing orbital, solar and volcanic forcing) has been carried out from 1492 to 2000. Another simulation driven with natural and anthropogenic forcings (changes in greenhouse gases, ozone, the direct and first indirect effect of anthropogenic sulphate aerosol and land-cover) from 1750 to 2000 has also been carried out. These simulations suggest that since 1550, in the absence of anthropogenic forcings, climate would have warmed by about 0.1 K. Simulated response is not in equilibrium with the external forcings suggesting that both climate sensitivity and the rate at which the ocean takes up heat determine the magnitude of the response to forcings since 1550. In the simulation with natural forcings climate sensitivity is similar to other simulations of HadCM3 driven with CO₂ alone. Climate sensitivity increases when anthropogenic forcings are included. The natural forcing used in our experiment increases decadal–centennial time-scale and large spatial scale climate variability, relative to internal variability, as diagnosed from a control simulation. Mean conditions in the natural simulation are cooler than in our control simulation reflecting the reduction in forcing. However, over certain regions there is significant warming, relative to control, due to an increase in forest cover. Comparing the simulation driven by anthropogenic and natural forcings with the natural-only simulation suggests that anthropogenic forcings have had a significant impact on, particularly tropical, climate since the early nineteenth century. Thus the entire instrumental temperature record may be “contaminated” by anthropogenic influences. Both the hydrological cycle and cryosphere are also affected by anthropogenic forcings. Changes in tree-cover appear to be responsible for some of the local and hydrological changes as well as an increase in northern hemisphere spring snow cover.

---

Analysis of the projected regional sea-ice changes in the Southern Ocean during the twenty-first century

Using the set of simulations performed with atmosphere-ocean general circulation models (AOGCMs) for the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR4), the projected regional distribution of sea ice for the twenty-first century has been investigated. Averaged over all those model simulations, the current climate is reasonably well reproduced. However, this averaging procedure hides the errors from individual models. Over the twentieth century, the multimodel average simulates a larger sea-ice concentration decrease around the Antarctic Peninsula compared to other regions, which is in qualitative agreement with observations. This is likely related to the positive trend in the Southern Annular Mode (SAM) index over the twentieth century, in both observations and in the multimodel average. Despite the simulated positive future trend in SAM, such a regional feature around the Antarctic Peninsula is absent in the projected sea-ice change for the end of the twenty-first century. The maximum decrease is indeed located over the central Weddell Sea and the Amundsen–Bellingshausen Seas. In most models, changes in the oceanic currents could play a role in the regional distribution of the sea ice, especially in the Ross Sea, where stronger southward currents could be responsible for a smaller sea-ice decrease during the twenty-first century. Finally, changes in the mixed layer depth can be found in some models, inducing locally strong changes in the sea-ice concentration.

Multi-century simulations with the coupled GISS–HYCOM climate model: control experiments

Multi-century climate simulations obtained with the GISS atmospheric general circulation model coupled to the hybrid-isopycnic ocean model HYCOM are described. Greenhouse gas concentrations are held fixed in these experiments to investigate the coupled model’s ability to reproduce the major features of today’s climate with minimal drift. Emphasis is placed on the realism of the oceanic general circulation and its effect on air–sea exchange processes. Several model runs using different closures for turbulent vertical exchange as well as improvements to reduce vertical numerical diffusion are compared with climate observations. As in previous studies, the Southern Ocean emerges as the Achilles Heel of the ocean model; deficiencies in its physical representation had far-reaching consequences in early experiments with the coupled model and have provided the strongest impetus for model improvement. The overarching goal of this work is to add diversity to the pool of ocean models available for climate prediction and thereby reduce biases that may stand in the way of assessing climate prediction uncertainty.

---

Predicting Effects of Temperature Increase on the Water Balance of Beech Forest – An Application of the ‘KAUSHA’ Model

The water balance model KAUSHA (Halldin, 1989) was applied to a 100-year-old beech (Fagus sylvatica L.) forest in northern Germany. Overall, a satisfying agreement between modelled evapotranspiration values and independent micrometeorological measurements (Bowen ratio energy balance method) could be observed, although for rainy days KAUSHA showed a tendency to overestimate evapotranspiration. The model was used to predict the effects of a climate warming on the water budgets of the forest. It is shown that a temperature increase of 2°C due to a rising CO₂ content of the atmosphere will not change the yearly totals of evapotranspiration significantly, but could have serious effects on the soil water balance during the vegetation period. Because under climate change conditions a higher amount of the available soil water has already been evaporated in winter and spring, soil water content will limit the transpiration of the trees from July to September much more strongly. Therefore, the yield of beech forest might also suffer from drought effects. It can be concluded that a better knowledge of the seasonal distribution of rainfall under climate change conditions is indispensable for predicting effects of rising temperatures and CO₂ concentrations on ecosystems.     

Assessing the Consequences of Climate Change for Food and forest Resources: A View from the IPCC

Important findings on the consequences of climate change for agriculture and forestry from the recently completed Third Assessment Report (TAR) of the Intergovernmental Panel on Climate Change (IPCC) are reviewed, with emphasis on new knowledge that emerged since the Second Assessment Report (SAR). The State-Pressure-Response-Adaptation model is used to organize the review. The major findings are:

• Constant or declining food prices are expected for at least the next 25 yr, although food security problems will persist in many developing countries as those countries deal with population increases, political crisis, poor resource endowments, and steady environmental degradation. Most economic model projections suggest that low relative food prices will extend beyond the next 25 yr, although our confidence in these projections erodes farther out into the 21st century.
• Although deforestation rates may have decreased since the early 1990s, degradation with a loss of forest productivity and biomass has occurred at large spatial scales as a result of fragmentation, non-sustainable practices and infrastructure development.
• According to United Nations estimates, approximately 23% of all forest and agricultural lands were classified as degraded over the period since World War II.
• At a worldwide scale, global change pressures (climate change, land-use practices and changes in atmospheric chemistry) are increasingly affecting the supply of goods and services from forests.
• The most realistic experiments to date – free air experiments in an irrigated environment – indicate that C₃ agricultural crops in particular respond favorably to gradually increasing atmospheric CO₂ concentrations (e.g., wheat yield increases by an average of 28%), although extrapolation of experimental results to real world production where several factors (e.g., nutrients, temperature, precipitation, and others) are likely to be limiting at one time or another remains problematic. Moreover, little is known of crop response to elevated CO₂ in the tropics, as most of the research has been conducted in the mid-latitudes.
• Research suggests that for some crops, for example rice, CO₂ benefits may decline quickly as temperatures warm beyond optimum photosynthetic levels. However, crop plant growth may benefit relatively more from CO₂ enrichment in drought conditions than in wet conditions.
• The unambiguous separation of the relative influences of elevated ambient CO₂ levels, climate change responses, and direct human influences (such as present and historical land-use change) on trees at the global and regional scales is still problematic. In some regions such as the temperate and boreal forests, climate change impacts, direct human interventions (including nitrogen-bearing pollution), and the legacy of past human activities (land-use change) appear to be more significant than CO₂ fertilization effects. This subject is, however an area of continuing scientific debate, although there does appear to be consensus that any CO₂ fertilization effect will saturate (disappear) in the coming century.
• Modeling studies suggest that any warming above current temperatures will diminish crop yields in the tropics while up to 2–3 ^°C of warming in the mid-latitudes may be tolerated by crops, especially if accompanied by increasing precipitation. The preponderance of developing countries lies in or near the tropics; this finding does not bode well for food production in those countries.
• Where direct human pressures do not mask them, there is increasing evidence of the impacts of climate change on forests associated with changes in natural disturbance regimes, growing season length, and local climatic extremes.
• Recent advances in modeling of vegetation response suggest that transient effects associated with dynamically responding ecosystems to climate change will increasingly dominate over the next century and that during these changes the global forest resource is likely to be adversely affected.
• The ability of livestock producers to adapt their herds to the physiological stress of climate change appears encouraging due to a variety of techniques for dealing with climate stress, but this issue is not well constrained, in part because of the general lack of experimentation and simulations of livestock adaptation to climate change.
• Crop and livestock farmers who have sufficient access to capital and technologies should be able to adapt their farming systems to climate change. Substantial changes in their mix of crops and livestock production may be necessary, however, as considerable costs could be involved in this process because investments in learning and gaining experience with different crops or irrigation.
• Impacts of climate change on agriculture after adaptation are estimated to result in small percentage changes in overall global income. Nations with large resource endowments (i.e., developed countries) will fare better in adapting to climate change than those with poor resource endowments (i.e., developing countries and countries in transition, especially in the tropics and subtropics) which will fare worse. This, in turn, could worsen income disparities between developed and developing countries.
• Although local forest ecosystems will be highly affected, with potentially significant local economic impacts, it is believed that, at regional and global scales, the global supply of timber and non-wood goods and services will adapt through changes in the global market place. However, there will be regional shifts in market share associated with changes in forest productivity with climate change: in contrast to the findings of the SAR, recent studies suggest that the changes will favor producers in developing countries, possibly at the expense of temperate and boreal suppliers.
• Global agricultural vulnerability is assessed by the anticipated effects of climate change on food prices. Based on the accumulated evidence of modeling studies, a global temperature rise of greater than 2.5 ^°C is likely to reverse the trend of falling real food prices. This would greatly stress food security in many developing countries.

     

Twentieth century North Atlantic climate change. Part II: Understanding the effect of Indian Ocean warming

Ensembles of atmospheric general circulation model (AGCM) experiments are used in an effort to understand the boreal winter Northern Hemisphere (NH) extratropical climate response to the observed warming of tropical sea surface temperatures (SSTs) over the last half of the twentieth Century. Specifically, we inquire about the origins of unusual, if not unprecedented, changes in the wintertime North Atlantic and European climate that are well described by a linear trend in most indices of the North Atlantic Oscillation (NAO). The simulated NH atmospheric response to the linear trend component of tropic-wide SST change since 1950 projects strongly onto the positive polarity of the NAO and is a hemispheric pattern distinguished by decreased (increased) Arctic (middle latitude) sea level pressure. Progressive warming of the Indian Ocean is the principal contributor to this wintertime extratropical response, as shown through additional AGCM ensembles forced with only the SST trend in that sector. The Indian Ocean influence is further established through the reproducibility of results across three different models forced with identical, idealized patterns of the observed warming. Examination of the transient atmospheric adjustment to a sudden “switch-on” of an Indian Ocean SST anomaly reveals that the North Atlantic response is not consistent with linear theory and most likely involves synoptic eddy feedbacks associated with changes in the North Atlantic storm track. The tropical SST control exerted over twentieth century regional climate underlies the importance of determining the future course of tropical SST for regional climate change and its uncertainty. Better understanding of the extratropical responses to different, plausible trajectories of the tropical oceans is key to such efforts.     

Effects of “realistic” land-cover change on a greenhouse-warmed African climate

The primary goal of this investigation is to focus on a realistic scenario for simulating impacts on regional African climate of future deforestation in a greenhouse-warmed world. Combined effects of plausible land-cover change and greenhouse warming are assessed by time-slice simulations with an atmospheric general circulation model (AGCM) for the middle of the twenty first century. Three time-slice integrations have been performed with the ARPEGE-Climat AGCM incorporating a zooming technique to achieve a resolution of about 100 km over Africa. A control run for the current climate is forced by observed climatological sea surface temperatures (SSTs) and the observed vegetation distribution is specified from a new vegetation database, in order to improve the geographical distribution and properties of the vegetation cover. Future SST changes are derived from a transient coupled atmosphere–ocean simulation for scenario B2 of the International Panel on Climate Change (IPCC). Future vegetation changes are specified from a simulation of scenario B2 with the Integrated Model to Assess the Global Environment (IMAGE) developed at the National Institute of Public Health and the Environment in the Netherlands (RIVM). The results show that land surface processes can locally modulate greenhouse warming effects for African climate, with reductions of surface transpiration and small increases of surface temperature. Deforestation of tropical Africa has overall only a marginal effect on precipitation because of a compensatory increase in moisture convergence. Energy budget analyses show that increases in surface temperature are produced both by increases of greenhouse gases (GHG) concentration from the increase in downward atmospheric longwave radiation, and by African tropical deforestation from the resulting reduction in transpiration. This study indicates that realistic land-use changes, though of smaller amplitude than greenhouse gas forcing, may have a small regional effect in projections of future climate.     

Regional climate model simulations of daily maximum and minimum near-surface temperatures across Europe compared with observed station data 1961–1990

Simulations of the present-day temperature climate in Europe obtained with the dynamic regional climate model HadRM3P from the Hadley Centre are evaluated. Observed daily temperature maxima (T _x) and minima (T_n) for the 1961–1990 period at 185 stations are compared with their nearest corresponding HadRM3P grid-box data. The model generally performs well over the UK and elsewhere between latitudes 50 and 55°N, with biases mostly within ±0.5 K. In other areas coherent regions with seasonal biases up to more than ±5 K exist. In some areas, biases in climatological averages are associated with even larger errors (up to more than ±15 K) in the upper/lower extreme temperature range. Both areas with systematically overestimated and underestimated intra-seasonal daily temperature variances exist, but overestimation dominates. Too hot summer T _x south of about 45°N are associated with drying soils in the model. This problem may occur further north in future integrations with a greenhouse-gas induced warming. Given the existence of errors in the simulations of the present-day climate, we recommend that results from future scenario integrations are treated with care.     

Optimal filtering for Bayesian detection and attribution of climate change

In the conventional approach to the detection of an anthropogenic or other externally forced climate change signal, optimal filters (fingerprints) are used to maximize the ratio of the observed climate change signal to the natural variability noise. If detection is successful, attribution of the observed climate change to the hypothesized forcing mechanism is carried out in a second step by comparing the observed and predicted climate change signals. In contrast, the Bayesian approach to detection and attribution makes no distinction between detection and attribution. The purpose of filtering in this case is to maximize the impact of the evidence, the observed climate change, on the prior probability that the hypothesis of an anthropogenic origin of the observed signal is true. Whereas in the conventional approach model uncertainties have no direct impact on the definition of the optimal detection fingerprint, in optimal Bayesian filtering they play a central role. The number of patterns retained is governed by the magnitude of the predicted signal relative to the model uncertainties, defined in a pattern space normalized by the natural climate variability. Although this results in some reduction of the original phase space, this is not the primary objective of Bayesian filtering, in contrast to the conventional approach, in which dimensional reduction is a necessary prerequisite for enhancing the signal-to-noise ratio. The Bayesian filtering method is illustrated for two anthropogenic forcing hypotheses: greenhouse gases alone, and a combination of greenhouse gases plus sulfate aerosols. The hypotheses are tested against 31-year trends for near-surface temperature, summer and winter diurnal temperature range, and precipitation. Between six and thirteen response patterns can be retained, as compared with the one or two response patterns normally used in the conventional approach. Strong evidence is found for the detection of an anthropogenic climate change in temperature, with some preference given to the combined forcing hypothesis. Detection of recent anthropogenic trends in diurnal temperature range and precipitation is not successful, but there remains strong net evidence for anthropogenic climate change if all data are considered jointly.

---