Canadian Ecosystem fluxes. There has been a dramatic reduction during the past few decades in the role of Canadian forests as a long term sink, despite enhanced NPP effects of CO2 fertilization. This reduction may now have reached the point where these forests have, on average, become a net source of carbon. A key factor in this rapid reversal has been the increase in wild fire losses and the outbreak of mountain pine beetle infestation in western Canadian forests. While the net emissions caused by this outbreak and related forest dieback appears to be near its peak, they will continue to be significant until at least 2020. It is unlikely that increased NPP will fully offset these emissions, especially since respiration and wildfire are also expected to increase as the climate warms65-67.
Site specific studies, however, show that there is considerable spatial and year to year variability in these fluxes. For example, multi-year analysis of fluxes in a boreal aspen stand in the northern Prairies suggest that the stand has, on average, been a small sink in recent years, but with considerable variations in response to inter-annual climate fluctuations and canopy duration/light suppression feedbacks. Surprisingly, moderate short-term droughts reduced respiration more than NPP and thus increased the NEP of the stand. Related modeling studies suggest that such stands would continue to be an average sink providing droughts last less than 3 consecutive years. Droughts of longer duration would cause the stand to become unsustainable. Studies of several Canadian black spruce/bog ecosystems, on the other hand, suggest these are neutral to weak net sources of CO2, with year to year variability in NEP dominated by changes in soil thaw seasons and water table depth. Low water tables during dry years caused increased respiration and negative NEP68-72.
Canadian peatlands have been a significant sink for carbon for many millennia, and contain more that 50% of all organic carbon in Canadian soils. About 60% of this carbon may be vulnerable to the effects of climate change. Related year-to-year fluctuations in emissions of CO2 and methane are sensitive to the effects of variations in precipitation and temperature on NPP and respiration. In areas where frozen peatlands slowly thaw or precipitation increases more than evaporation, average soil water content and methane emissions are projected to increase, but NEP is also expected to rise. One recent northern Canada peatland study indicated, for example, that average net CO2 uptake over a six year period exceeded the equivalent effect of increased methane emissions, resulting in a net equivalent carbon sink similar to that of the past few millennia. On the other hand, in regions where average peatland water tables decrease, methane emissions are expected to decrease and enhanced decay of peatland carbon will increase CO2 emissions. Although studies suggest southern peatlands may become significant net CO2 sources if warmer temperatures increase their dessication, it remains uncertain what the net overall effect on emissions will be for other Canadian peatlands73-76.
2.1.3 Ocean carbon fluxes Recent studies estimate that global ocean sinks remove about 2 to 2.3 GtC from the atmosphere each year. However, the efficiency of the oceans as carbon sinks has decreased by about 5% during the past half-century. Furthermore, fluxes vary considerably from year to year and decade to decade. During the mid 1990s, temperate Northern Hemisphere oceans were a major sink for atmospheric CO2. However, by 2002-05, the North Atlantic sink had weakened significantly, likely due to the combined effects of NAO-related circulation changes and reduced air-sea exchange due to surface CO2 buffering. Should the rate of Atlantic Ocean overturning continue to slow down as the global climate warms, there would be a further reduction in the sequestration of anthropogenic CO2 into the deep ocean through advection. However, this may be offset by increased absorption of CO2 into the cooler surface ocean waters in the North Atlantic. The Labrador Sea also has been, and continues to be a significant carbon sink. Furthermore, there is now evidence that the Canadian Basin of the Arctic Ocean is becoming an increasingly significant carbon sink as the retreat of sea ice allows increased ocean mixing and biological activity in these waters. In the Southern Hemisphere, the Southern Ocean provides a significant sink that varies from year to year. The average magnitude of the Southern Ocean sink appears to have decreased in recent decades. High latitude regions of the Southern Ocean, particularly the Ross Sea shelf waters, appear to be more significant carbon sinks than regions at lower latitudes. Some studies suggest that the decline in net Southern Ocean sinks is likely due to the effect of increasing regional wind speeds on surface mixing and regional biochemistry and that this trend is likely to continue under warmer climates. Other experts argue that other factors could offset these wind effects. Meanwhile, tropical waters and other ocean areas of the Southern Hemisphere tend to be sources of atmospheric carbon dioxide that partially offset the above sinks26,28,77-90.
There are concerns that models used to estimate ocean carbon fluxes inadequately address or are missing some of the factors that affect these fluxes. Laboratory studies suggest that one such factor may be the direct enhancement of the ocean biological pump by CO2 fertilization. Another is the regional variability of the effect of the ocean twilight zone (the limited light region between about 200 and 1000m depth) on the biological pump. Nitrogen deposition may account for one-third of the external nitrogen supply to ocean waters, and may be enhancing the ocean carbon sink by as much as 10%. However, it may also be enhancing N2O production and release from ocean surfaces. Similarly, models do not address the role of iron deposition into oceans through wind-blown dust. If this decreases due to weaker winds, the loss of related nutrients could negatively affect the biological pump and reduce ocean sinks91-94.
Other Greenhouse Gases
2.2.1 Methane Updated reconstructions of past atmospheric methane (CH4) concentrations from the extended Antarctic ice core records continue to show a strong correlation between Antarctic temperatures and CH4 concentrations throughout the past 800,000 years. Minimum values of 350 parts per billion (ppb) occurred during cold glacial stages and maximum values during the intervening interglacials reached about 800 ppb. Methane concentration changes during abrupt climate anomalies such as Dansgaard-Oeschger events (see section 5.1) also show rapid response to climate shifts, although with a lag time of about 50 years. Isotopic analyses suggest that significant contributions from human agricultural activities to the global methane budget may have started during the Little Ice Age, some 500 years ago95-100.
Figure 2. Atmospheric methane concentrations since 1979. The data show a reduction in the rate of increase in time with no significant trend in recent decades. A slight rise is apparent in 2007-08. (Adapted from NOAA data, ref. #15) By 2007, these concentrations had reached 1789 ppb – a level about 156% higher that that estimated for 1750, and without apparent precedence in the entire ice core records noted above. There has been no significant trend in methane concentrations during the past two decades. Hence, it now appears unlikely that the projected concentrations for the next few decades in the IPCC SRES scenarios will be achieved. However, there is recent evidence to suggest that concentrations began to rise again in 2007.16,98,101-103.
Year-to-year variations in methane concentrations in recent decades appear to correlate well with changes in the amount of biomass burning and with shifts in the El Niño-Southern Oscillation (ENSO) index. However, both paleo and current records of changes in methane concentrations suggest that wetlands (where biomass decay in the absence of oxygen can generate large amounts of methane rather than carbon dioxide) play a dominant role in the annual global methane budget. Much of the current wetland emissions occur in tropical regions. As permafrost slowly decays, wetlands in northern regions are expected to expand, and their methane emissions to increase. Various related studies indicate that as much as half of the annual methane release from tundra ecosystems occurs as sudden bursts during fall freeze-up. Bubbling of methane from sediments at the bottom of northern lakes can also be a large source. Much of this methane is old methane trapped in decaying permafrost along the margins of thawing lakes. Further warming of these lakes and ecosystems will have varied effects on regional emissions, with some ecosystems releasing less methane due to lower water tables, and others more as surface water increases. In general, CO2 emissions from northern ecosystems increase when methane production decreases, and vice versa. Biogeochemical model studies suggest that, at least for Alaska, the net effect, integrated over area and time, is likely to be an overall increase in northern ecosystem greenhouse gas emissions during the next century104-109.
Recent laboratory and field studies have discovered a possible new source of biological methane release that is surprising and as yet controversial. These studies indicate that many woody plants can release significant amounts of methane through as yet poorly understood oxic processes. Such emissions could represent as much as 10 to 30% of the estimated current global methane budget, and could be further enhanced by rising temperatures and by CO2 fertilization effects. Hence, oxic methane emissions could be a larger contributor in the future, creating another positive climate feedback within the climate system not yet considered in climate models60,110-114.
Methane is also released into the atmosphere from other natural sources, such as volcanic vents and decaying gas hydrates. New estimates based on isotopic studies suggest methane released from natural fossil sources may be much larger than previously estimated, perhaps even the second largest natural source after wetlands. However, the role of gas hydrate emissions - past, present and future – remains uncertain and controversial. Methane budget model studies suggest that the rapid rise in methane concentrations during past deglaciations is due to rapid release from high concentration sources such as hydrates. The model results also suggest that hydrate deposits under shallow ocean regions are vulnerable to decay due to future ocean warming, particularly in the Pacific Ocean. Over the next century, this could release as much as 35 MtC in the form of methane into the atmosphere each year – adding about 6% to current total estimated methane emissions. On the other hand, detailed analysis of the ice core data suggest that wetland emissions rather than hydrates were the major source of increasing atmospheric methane concentrations during both past interglacials and abrupt warm anomalies such as the Younger Dryas event99,100,115-121.
A primary sink for methane is oxidation in terrestrial soils. Although past studies have suggested this sink could be between 20 and 45 MtCH4 per year, recent model and data studies imply values in the lower end of this range. About half of this sink is in subtropical and dry tropical soils122-123.
Like CO2 and CH4, changes in nitrous oxide (N2O) atmospheric concentrations show a close correlation with changes in temperature over at least the past 650,000 years. During this time, values ranged between 200 ppb during glacial periods and 270 ppb during interglacials. By 2007, atmospheric concentrations had increased to 321 ppb, 0.25% higher than the preceding year and 19% above pre-industrial levels. Model studies suggest that agricultural lands are a primary source of N2O emissions, but that tropical rainforest soils are a significant secondary source16,100,124.
Concentrations of tropospheric ozone are estimated to have increased by 50% since pre-industrial times. The expected change in concentrations during the next century is strongly dependent on emission scenarios of ozone precursors and on how atmospheric chemical processes that create and destroy ozone evolve under changing climates. Studies suggest changes in average global concentrations could vary between a slight decrease of -6% to a significant increase of 43%. Within the lower troposphere, ozone concentrations are closely linked to local emissions and transport of precursors such as NOx. Recent studies suggest the transport of these precursors into the upper troposphere may also cause seasonal and regional variations in ozone concentrations at this level. For example, in situ production of upper tropospheric ozone over eastern North America in summer can exceed that for polluted areas below, largely because of NOx transported into the region from elsewhere125-126.
There is also evidence that naturally occurring halogen gases over the tropical oceans may be a significant sink for tropospheric ozone over the tropical Atlantic region. Since most ozone models do not include these sinks, they may have somewhat overestimated ozone concentrations and related radiative forcing in these regions127.
Halogenated gases such as CFCs, HFC, PFCs and SF6, which are primarily of human origin, have long been recognized as important greenhouse gases. These gases still have low concentrations within the atmosphere, but may be of increasing importance in the future. That is because some of these have atmospheric lifetimes on the order of millennia. Most of these are already regulated under the Montreal Protocol for ozone depleting substances or included in the gases considered under the Framework Convention on Climate Change and its protocols. However, recent investigations indicate that nitrogen triflouride (NF3) is also a potent greenhouse gas that is not as yet included in these mitigation instruments. Annual production of NF3 is currently equivalent to about 67 million tones of CO2, of which about 16% is being released into the atmosphere. Atmospheric concentrations of NF3 are rising at 11% per year.128-129.
Records of dust deposits in Antarctic ice cores show that atmospheric dust loading in the region was about 25 times larger during glacial periods than during interglacials. Since dust reduces solar insolation while in the atmosphere and enhances carbon sinks through biological activity when deposited into ocean waters, it provides an important positive climate feedback.130-131.
Tropospheric sulphate aerosol concentrations from human sources are estimated to have tripled relative to pre-industrial levels, while sooty aerosols have increased six-fold. Scenarios for future anthropogenic emissions indicate that tropospheric sulphate concentrations will likely be relatively stable for the next few decades, then decline to between 4 to 45% below 2000 levels by 2100. This decrease is due to an anticipated increase in the introduction of air pollution control measures in developing regions of the world. The rate of growth in black carbon concentrations in the troposphere has also declined in recent decades, likely due to increased adoption of new clean air technologies in fossil fuel combustion processes. Meanwhile, emissions of organic carbon aerosols have more than doubled, largely due to the increased use of biofuels. Observations also indicate that black carbon aerosols in urban areas, which largely come from fossil fuel combustion, are finer, smaller and have longer lifetimes than those in rural regions where biomass burning is a primary source. Volcanic eruptions such as that of Pinatubo in 1991 can also temporarily add significant amounts of sulphate from natural sources into the stratosphere, thus adding to total atmospheric loading of aerosols. Satellite data suggest that this total has declined since 1991, largely due to the settling out of the Pinatubo aerosols from the stratosphere126,132-134.
3.0 Radiative Forcing Improved estimates for net global radiation budgets suggest a total average planetary albedo of 28.9%. Surface albedo alone, in the absence of cloud, accounts for about 15%. Absorption of solar radiation within the atmosphere is now estimated at 76 W/m2, more than 10% larger than that of past estimates135.
3.1 Greenhouse Gases The rate of increase in net radiative forcing during the 20th century due to the rise in concentrations of carbon dioxide, methane and nitrous oxide is unmatched in at least the past 16,000 years. Together with the halogen gases CFC-12 and CFC-11, these gases accounted for 97% of the direct radiative forcing caused by all long-lived greenhouse gases in the atmosphere since 1979. The fraction of this attributable to CO2 alone slightly increased from about 60 to 63%. There is also a small positive indirect methane forcing of about 0.1 W/m2 caused by the oxidation of methane to stratospheric water vapour. This adds about 15-20% to the direct radiative forcing due to methane136-138.
Figure 3. Comparison of past radiative forcing from primary greenhouse gases (dark red),halocarbons and SF6 (bright red), human emissions of sulphate aerosols (green), solar forcing (dashed blue) and volcanic emissions (grey). The comparison suggests primary greenhouse gases have dominated forcing in recent decades (Joos et al., 2008, ref. #137).
The research community is now increasingly turning its attention to better quantifying the radiative forcing due to short lived greenhouse gases, which are much less well understood than the long lived gases. New satellite measurements show that the mean radiative forcing from ozone in 2006 in the region of the upper troposphere between 45°N and 45°S was 0.48 ± 0.14 W/m2, with large variations from region to region. This is within the range of results reported in the IPCC AR4. Biomass burning emissions (which include ozone precursors) may be responsible for about one-third of the global forcing from tropospheric ozone since pre-industrial times139-140.
Surface and airborne studies indicate that the impact of NOx emissions on CH4 and O3 chemistry is strongly dependent on the geographic location, season and altitude of emission. For example, emissions from aircraft that occur at higher levels of the troposphere produce greater reductions in methane concentrations and higher ozone concentrations than emissions at the earth’s surface. NOx emissions tend to decrease methane concentrations and increase ozone concentrations. Integrated over time, the radiative cooling effects of the NOx-driven CH4 depletion appears to slightly exceed the warming effects of the concurrent enhancement of ozone concentrations. Thus, NOx emissions cause a net time-integrated negative radiative forcing (cooling), and reductions in these emissions could lead to climate warming. Other studies into the co-benefits of emission reductions in various ozone precursors for air quality and climate change also show that reductions in methane emissions has the greatest impact in reducing ozone radiative forcing, while reductions in NOx emissions increased net ozone radiative forcing. These results suggest that, in order to reduce net radiative forcing from ozone, it may be better to target reductions in emissions of other ozone precursors, particularly methane, rather than NOx141-143.
There are also indirect radiative forcing effects from ozone, particularly at ground level. For example, experimental studies indicate that elevated ozone levels reduce CO2 uptake by plants, providing a positive feedback on climate. Under the IPCC SRES A2 scenario projections for ground level ozone, CO2 uptake by vegetation could be reduced by between 17 and 31%, globally. Taking this indirect effect into account doubles the effective radiative forcing due to projected increases in O355.
Enhanced greenhouse gas concentrations are also projected to increase the exchange of air between the stratosphere and troposphere by about 2% per decade, and cause more ozone to be transported from the stratospheric tropics to the extratropics. This helps to enhance the rate of recovery of the depleted ozone conditions within the stratosphere. Simulations that include such dynamic factors suggest full stratospheric ozone recovery by around mid-century, and a further increase in concentrations after that. That is, a ‘super recovery’ is projected. Since ozone is a greenhouse gas, this would indirectly add slightly (less than 0.1 W/m2) to the net greenhouse radiative forcing. It is projected to also cause a warming of the polar stratosphere, a weakening of the circumpolar westerlies during the spring, and a warming of the troposphere in the summer. These results differ from those simulated by the IPCC AR4 models, which did not take ozone layer recovery into account144-146.
3.2 Anthropogenic Aerosols There is now much greater confidence in estimates of the direct radiative forcing caused by changes in atmospheric aerosol concentrations. Recent model estimates suggest that this forcing, averaged globally, is strongly negative, ranging between -0.35 and -0.7 W/m2. However, it is highly variable in space and time, partly because of its dependence on meteorological conditions. Direct aerosol radiative forcing is also about twice as large in clear skies as it is in skies that include clouds. Estimates from observed data agree quite well with such model estimates over ocean areas, but are much higher over land areas. One study based on analyses of satellite data, for example, indicates that global direct radiative aerosol forcing may be as high as -0.9 W/m2 147-150.
Aerosols also affect atmospheric radiative fluxes indirectly through their impact on cloud properties and on atmospheric dynamic and hydrological processes. One factor is the influence of increased aerosol concentrations on cloud nucleation processes, and hence the number and size of cloud droplets. This also affects the total amount of precipitable water vapour in the atmosphere. Satellite data confirm that average cloud droplet radius decreases and cloud cover increases as aerosol optical thickness increases. This response may have increased global cloud cover by as much as 5%. Surface solar irradiance data provides supporting evidence for increased cloud cover. However, a number of studies now indicate that the total indirect aerosol effect may be much lower than past studies suggest, and may actually be smaller that the direct effect. Satellite based measurements, for example, suggest net average indirect cloud effects may currently be as low as -0.2W/m2. A key factor may be that these aerosols, in addition to making clouds more reflective, also change atmospheric circulation and redistribute clouds and precipitation in a manner that, in some regions, can counteract the enhanced cloud albedo. Processes can also vary by region. In the Arctic, for example, increased cloud cover can cause a positive forcing, since the reduction in reduced outgoing long-wave radiation is greater than increased reflection of incoming shortwave radiation. Here, the indirect aerosol effect can thus be positive150,151-156.
Another unusual indirect aerosol effect is that related to ship cloud trails. When mixed with moist air, the aerosols within ship exhausts can quickly form cloud droplets that, under the right humidity conditions, can spread and persist for extended time period. While the net effect of these cloud trails is negligible on a global scale (about -0.5 mW/2), it can be as high as -0.05 W/m2 in high traffic, high humidity areas such as along the west coast of North America157.
Over the past century, the net effect of rising aerosol concentrations has been a significant decrease in the amount of incoming solar radiation that reaches the Earth’s surface, often referred to as `solar dimming`. Despite improved understanding, the range of estimates for the magnitude of this dimming is still large. Although most agree that this effect is smaller than previously estimated, some still argue it could have been large enough to have more than offset the effects of greenhouse gas forcing to date. There is considerable regional variation in the magnitude and trends of solar dimming. Over Europe, it is estimated to have caused an average decline in surface incoming solar radiation between 1971 and 1986 of as much -3 W/m2/decade, with much greater declines in densely populated regions. Subsequent programs aimed at reducing aerosol emissions to improve local air quality have reversed this trend, thus causing a regional ‘solar brightening’ of about 2 W/m2/decade over the region. This has likely been a factor in the rapid regional warming in recent decades. Similar reversals in aerosol forcings have occurred over North America. However, aerosol emissions have continued to increase substantially over developing regions such as Asia. The net global effect over recent decades has been a small increase in total global aerosol cooling of -0.1W/m2. Thus the offset against rising greenhouse gases during this time period has been minimal. By 2030, continued increases in aerosol concentrations in developing regions are expected to significantly exceed any further declines in the industrialized world, potentially adding another -0.94 W/m2 to the current aerosol cooling effect158-165.
The surface cooling effect of sulphate aerosols and other reflective aerosols is much greater in regions with high atmospheric concentrations than in relatively unpolluted regions. This differential offset to greenhouse gas warming also affects regional atmospheric and ocean circulation and related heat transport, primarily by enhancing the rate of Atlantic ocean overturning166.
Recent updated estimates for the net radiative forcing of aircraft contrails around the world suggest a small contribution of 6 mW/m2, or less than 1 % of that due to enhanced greenhouse gas concentrations. Contrails have the greatest impact at altitudes of about 10 km, with each 1% of increase in traffic adding 0.25% to the radiative forcing. However, in areas of intense air traffic, this amount can be significantly greater. Analyses of contrails over southern England, at the entrance of the main North Atlantic air travel corridor, indicate that contrails there cause a regional annual forcing of about 0.23 W/m2. Results also suggest that effects are significantly greater in winter than during the rest of the year, and during night flights than daytime flights167-168.
While most human emissions of aerosols cause a cooling effect through their direct and indirect reflection of incoming sunlight, sooty aerosols, referred to as black carbon, do the opposite. Within the atmosphere, black carbon directly absorbs incoming solar radiation. While this heats the layers of air where their concentrations are high, it reduces net radiation reaching the surface and hence contributes to surface cooling. It also mixes with other aerosols, thus reducing their ability to reflect incoming sunlight and adding to the atmospheric warming effect. In highly polluted air masses over low latitudes, where solar radiation is most intense, this mixing of aerosols creates layers of haze, usually at altitudes below 5 km, that have become known as atmospheric brown clouds. Over India, for example, these clouds have their highest concentrations at about 2-4 km elevation, where they can cause a warming at that level of a few degrees relative to surface temperatures. This warming has likely been a factor in the increasing rate of melt of Himalayan glaciers located at these altitudes. Over adjacent oceans, the maximum warming effect is at lower elevations, thus creating horizontal temperature gradients that affect regional atmospheric circulation. The impact of these brown clouds can extend long distances beyond their source region. The positive radiative forcing they cause over the northern Maldives, for example, may be greater than that due to the local effects of increases in greenhouse gas concentrations. Their influences on large scale circulation patterns include a Hadley atmospheric circulation system that is stronger in the Northern Hemisphere and weaker in the Southern Hemisphere and changes in the Indian and Pacific Ocean circulation patterns similar to those during an ENSO event169-173.
Black carbon also causes warming by reducing surface albedo when deposited on reflective surfaces like snow and ice. In the Arctic, this can add about 0.5 W/m2 to average local radiative forcing. While some of the carbon found in the Arctic snow comes from upwind wild fires, about 80% comes from human sources. Greenland ice core records indicate that, prior to 1950, most of the carbon deposited there came from industrial activities in North America. Since then, the primary source has become Asia. The Greenland ice core data also suggest that the influence of black carbon on Arctic climate is greatest during the early summer period, when snow is still present and incoming sunlight is most intense. Between 1850 and 1950, the average impact on local forcing at this time of year is estimated at 1.13 W/m2, peaking at 3.2 W/m2 in 1906-1910. Since 1950, atmospheric concentrations of black carbon have declined, reducing related forcing to an average of 0.59 W/m2. Hence some of the early 20th century Arctic warming may already have been due to such human influences. Furthermore, when this effect is added to its atmospheric role, black carbon may be the second strongest radiative forcing mechanism in this region, after CO2169,174-176.
Figure 4. Modelled trends in spring-time radiative forcing due to carbon deposition on Arctic snow and ice over the past two centuries. Results suggest a strong positive forcing from this source in the early 19th-century (McConnell et al., 2007, ref. #176)